pixpop said:
Your second point goes to the heart of the matter. The speed torque curve of a DC motor is a very nice straight line, so the speed / torque ratio does not change with torque (or speed). That's why the manufacturers quote a speed / torque constant.
BTW even if this was not the case, adding an extra load would actually make it worse. This is what happens with induction motors - as you load them they slip down their speed / torque curve and the speed / torque ratio gets worse with increased torque.
Hi,
Having looked back over this thread quickly, there are a few comments which I hope might be helpful to some posters, but I am not prepared to indulge anyone in any arguments about what I say.
It is clear that there are several influencing factors to take into account here. e.g. Speed variation across a record, variation due to drag caused by heavily modulated passages, and, (IMHO) the most important of all, how the records actually sound when being played.
Accordingly, there are many different ways of tackling these issues, and individual designers will inevitably end up with different choices/compromises when they finalise their designs.
It seems that Mark has made different choices than I have, which is fine by me, but from what he says I now do not consider that he has *measured* the results as carefully as I have done, when making his design choices. I don't know anything about his listening trials, of course.
I would be very interested to learn how Vinyl Addict actually "measured about 0.1% drift" which he mentions now.
I am not sure I understand all that Mark said in response to Pixpop, recently. Whilst I don't condone deliberately loading up any motor unnecessarily, if I understand what is meant by both posters here, I don't agree that Pixpop is entirely wrong in his thoughts. Whether it relates to 'reserves' of available power, or whatever, I don't know, but maybe it is analagous to the following.
Until recently, my wife had a 1300 cc car, which I never enjoyed driving much in comparison with my own 3 litre turbo, which had masses of torque. Going along on the flat there was little difference, of course, but when a hill (an unwanted additional load) came up, the lower-torque 1300 flagged and slowed down noticeably, but this did not occur with my own car which did not even notice the difference.
Whatever additional 'drag' is caused and by whatever means, if it is infinitesimal compared with the 'normal forces' involved in maintaining the TT's speed at other times, even if it suddenly and unwantedly doubles, it will still be less consequential in effect, in my opinion.
e.g. If the *change* in drag across a record is say a reduction of x2 (1/2) from start to end, then 2X 0.1% (of the 'normal' forces involved) would be equivalent to an increase in speed of also 0.1%.(i.e. it is doubled). However, if the stylus drag represents a geater proportion of the other forces involved, say 1%, then this similar 2x change across the record will then amount to a 1% increase in the speed by the end of the record.
In this thread alone, Dice45 says he uses a simple LM317 regulator (not sure of voltages used here, though), and Planet IX an LT1083 at 11v DC, and both state that they do not have any speed variations across the record, although maybe they do have some stylus drag (or some other) concerns. I use similar voltages to Dice45, and I know of several TT manufacturers who also use voltages in this region.
My own regulator (which is more complex than I would prefer) has no feedback, or whatever, and I don't have problems in either long-term speed variations, nor any due to shorter-term record 'modulation-induced' changes. I don't believe it is possible to measure these effects any better than I have already done, and (together with many other listeners) I am quite content with the resultant sound at all times. It doesn't vary, cold or warm, night or day, week to week, record to record, to any extent which is capable of being measured, nor heard. The extremely low tempco components used in the final PS version, gave an improvement (and multiplied the costs, several-fold!) here.
Regrettably, I am subject to an NDA with a TT manufacturer so I am precluded from giving away any further details, but many of the principles which I have found to be significant have already been posted by me here.
It is up to individual DIYers to decide on the best methods to adopt for their own situations, and of course this will be influenced by the components available to them, but, as I said before, I am wary of anyone who boldly states "This is the only way" etc.
On some individual points raised before:
Ball-bearing-v-sleeve-bearing motors. Here I would go for sleeve-bearing every time, as ball bearing motors 'rumble' and this appears to be inherent in their construction. I have not tried every available motor, though, but I have made several direct comparisons with a medical stethoscope. Incidentally, placing the handle end of a screwdriver gently against the inner part of your ear, with the point on any mechanical device which is operating, works quite effectively as a crude stethoscope, as the vibrations are transmitted directly.
Mains-related strobes are not good enough for accurate speed-checking of TTs (IMO) when making PS design decisions, as Pixpop has already found out. They may be acceptable for general use, though, and are admittedly cheap.
The Burr-Brown 'servo' circuit which Pixpop queried is not good enough for this application, as I mentioned in post #8, and for the reasons discussed. I wasted several months attempting to refine this design, and went back to straightforward regulators.
DC motors can be fine in use, but maybe more attention to details like PS etc., is needed when using these. I would not revert to using AC motors as these have some shortcomings which are harder (sometimes impossible) to overcome. Unless you have a bespoke AC supply where the frequency can be varied, it is not possible to adjust the speed with much accuracy, and of all the AC driven TTs I tested (mentioned earlier) not one ran at the correct speed at any time! Also, I did notice, first of all by listening, and then when I had the Gyrasope (but not before) through measuring/sight-tests, that they all varied in speed when playing certain passages on records.
I don't know much about Origin Live's motors , nor their PS, but they are obliged to use voltages etc. to suit the existing OEM platter diameters on their many different aftermarket equipment kits, but it does not follow that their choices are optimum, here. Any DIYer starting from scratch does not have this limitation, and I have read several comments that some (especially DC) motors can be 'heard' right across a room!
I cannot hear my motor even when my ear is a couple of inches away from the motor, or possibly a very faint 'rustling' sound from the belt being rapidly wound/unwound tightly around a small pulley, which disappears when the belt is detached. To all intents and purposes it is absolutely silent.
Matt.
If you are really stuck, I could machine down a pulley of yours if you wish, as I have the necessary equipment, and I machine my own pulleys and other TT parts, when developing these.
I hope this helps.
Having looked back over this thread quickly, there are a few comments which I hope might be helpful to some posters, but I am not prepared to indulge anyone in any arguments about what I say.
It is clear that there are several influencing factors to take into account here. e.g. Speed variation across a record, variation due to drag caused by heavily modulated passages, and, (IMHO) the most important of all, how the records actually sound when being played.
Accordingly, there are many different ways of tackling these issues, and individual designers will inevitably end up with different choices/compromises when they finalise their designs.
It seems that Mark has made different choices than I have, which is fine by me, but from what he says I now do not consider that he has *measured* the results as carefully as I have done, when making his design choices. I don't know anything about his listening trials, of course.
I would be very interested to learn how Vinyl Addict actually "measured about 0.1% drift" which he mentions now.
I am not sure I understand all that Mark said in response to Pixpop, recently. Whilst I don't condone deliberately loading up any motor unnecessarily, if I understand what is meant by both posters here, I don't agree that Pixpop is entirely wrong in his thoughts. Whether it relates to 'reserves' of available power, or whatever, I don't know, but maybe it is analagous to the following.
Until recently, my wife had a 1300 cc car, which I never enjoyed driving much in comparison with my own 3 litre turbo, which had masses of torque. Going along on the flat there was little difference, of course, but when a hill (an unwanted additional load) came up, the lower-torque 1300 flagged and slowed down noticeably, but this did not occur with my own car which did not even notice the difference.
Whatever additional 'drag' is caused and by whatever means, if it is infinitesimal compared with the 'normal forces' involved in maintaining the TT's speed at other times, even if it suddenly and unwantedly doubles, it will still be less consequential in effect, in my opinion.
e.g. If the *change* in drag across a record is say a reduction of x2 (1/2) from start to end, then 2X 0.1% (of the 'normal' forces involved) would be equivalent to an increase in speed of also 0.1%.(i.e. it is doubled). However, if the stylus drag represents a geater proportion of the other forces involved, say 1%, then this similar 2x change across the record will then amount to a 1% increase in the speed by the end of the record.
In this thread alone, Dice45 says he uses a simple LM317 regulator (not sure of voltages used here, though), and Planet IX an LT1083 at 11v DC, and both state that they do not have any speed variations across the record, although maybe they do have some stylus drag (or some other) concerns. I use similar voltages to Dice45, and I know of several TT manufacturers who also use voltages in this region.
My own regulator (which is more complex than I would prefer) has no feedback, or whatever, and I don't have problems in either long-term speed variations, nor any due to shorter-term record 'modulation-induced' changes. I don't believe it is possible to measure these effects any better than I have already done, and (together with many other listeners) I am quite content with the resultant sound at all times. It doesn't vary, cold or warm, night or day, week to week, record to record, to any extent which is capable of being measured, nor heard. The extremely low tempco components used in the final PS version, gave an improvement (and multiplied the costs, several-fold!) here.
Regrettably, I am subject to an NDA with a TT manufacturer so I am precluded from giving away any further details, but many of the principles which I have found to be significant have already been posted by me here.
It is up to individual DIYers to decide on the best methods to adopt for their own situations, and of course this will be influenced by the components available to them, but, as I said before, I am wary of anyone who boldly states "This is the only way" etc.
On some individual points raised before:
Ball-bearing-v-sleeve-bearing motors. Here I would go for sleeve-bearing every time, as ball bearing motors 'rumble' and this appears to be inherent in their construction. I have not tried every available motor, though, but I have made several direct comparisons with a medical stethoscope. Incidentally, placing the handle end of a screwdriver gently against the inner part of your ear, with the point on any mechanical device which is operating, works quite effectively as a crude stethoscope, as the vibrations are transmitted directly.
Mains-related strobes are not good enough for accurate speed-checking of TTs (IMO) when making PS design decisions, as Pixpop has already found out. They may be acceptable for general use, though, and are admittedly cheap.
The Burr-Brown 'servo' circuit which Pixpop queried is not good enough for this application, as I mentioned in post #8, and for the reasons discussed. I wasted several months attempting to refine this design, and went back to straightforward regulators.
DC motors can be fine in use, but maybe more attention to details like PS etc., is needed when using these. I would not revert to using AC motors as these have some shortcomings which are harder (sometimes impossible) to overcome. Unless you have a bespoke AC supply where the frequency can be varied, it is not possible to adjust the speed with much accuracy, and of all the AC driven TTs I tested (mentioned earlier) not one ran at the correct speed at any time! Also, I did notice, first of all by listening, and then when I had the Gyrasope (but not before) through measuring/sight-tests, that they all varied in speed when playing certain passages on records.
I don't know much about Origin Live's motors , nor their PS, but they are obliged to use voltages etc. to suit the existing OEM platter diameters on their many different aftermarket equipment kits, but it does not follow that their choices are optimum, here. Any DIYer starting from scratch does not have this limitation, and I have read several comments that some (especially DC) motors can be 'heard' right across a room!
I cannot hear my motor even when my ear is a couple of inches away from the motor, or possibly a very faint 'rustling' sound from the belt being rapidly wound/unwound tightly around a small pulley, which disappears when the belt is detached. To all intents and purposes it is absolutely silent.
Matt.
If you are really stuck, I could machine down a pulley of yours if you wish, as I have the necessary equipment, and I machine my own pulleys and other TT parts, when developing these.
I hope this helps.
Bob,
thanks for a very reasonable post. I've learned a great deal from this thread, cleared up some wrong assumptions of mine. This is good, because I'm currently working on my own turntable project, and want to avoid going down any blind alleys. I apologize to Matt if any of my questions have taken the thread too far from his topic.
I was shocked to learn that he experiences this speed variation across the record, and, as I'm sure Matt does, would love to know the explanation. When thinking about it, the only way I can make sense of it is to realize that the change in drag across the record must be a significant proportion of the total drag. If it was only a tiny proportion, then it would have only a tiny effect. This is what led me to ask about increasing the drag. I'm not suggesting this as a solution, although I believe there are TTs that deliberately introduce some drag. I mention it only in order to clarify my understanding about what's going on.
I think Mark's description of the speed/torque curve is only valid if the voltage across the motor is constant. This is clear when you realize that the speed/torque curve is drawn from the zero speed/maximum torque point, to the maximum speed/zero torque point (AKA unloaded speed). If you just think for a moment about any DC motor connected to a voltage source, you will agree that if you increase the voltage then the speed will increase. This changes the speed/torque curve, because you now have a new unloaded speed. Similarly, if you think of the same motor stalled, and then increase the voltage, the torque will increase. Once again, a different speed/torque curve, because the stall torque has changed.
If Mark's circuit (IR compensation) holds the speed perfectly stable when the load changes, then it must change the voltage. I agree with Mark that a constant voltage controller will suffer speed change with load change. I differ from Mark, because I think that the amount of speed change will depend on the steady state torque, which in turn depends on the steady state frictional load.
This is further complicated because we are currently talking about simple linear regulators, and these exhibit a small voltage drop with increasing load. Also, we're not all talking about the same means of measuring speed.
Matt, I hope you keep at it, and find the solution. I'd love to know (rather than merely speculate) what's going on with your TT.
thanks for a very reasonable post. I've learned a great deal from this thread, cleared up some wrong assumptions of mine. This is good, because I'm currently working on my own turntable project, and want to avoid going down any blind alleys. I apologize to Matt if any of my questions have taken the thread too far from his topic.
I was shocked to learn that he experiences this speed variation across the record, and, as I'm sure Matt does, would love to know the explanation. When thinking about it, the only way I can make sense of it is to realize that the change in drag across the record must be a significant proportion of the total drag. If it was only a tiny proportion, then it would have only a tiny effect. This is what led me to ask about increasing the drag. I'm not suggesting this as a solution, although I believe there are TTs that deliberately introduce some drag. I mention it only in order to clarify my understanding about what's going on.
I think Mark's description of the speed/torque curve is only valid if the voltage across the motor is constant. This is clear when you realize that the speed/torque curve is drawn from the zero speed/maximum torque point, to the maximum speed/zero torque point (AKA unloaded speed). If you just think for a moment about any DC motor connected to a voltage source, you will agree that if you increase the voltage then the speed will increase. This changes the speed/torque curve, because you now have a new unloaded speed. Similarly, if you think of the same motor stalled, and then increase the voltage, the torque will increase. Once again, a different speed/torque curve, because the stall torque has changed.
If Mark's circuit (IR compensation) holds the speed perfectly stable when the load changes, then it must change the voltage. I agree with Mark that a constant voltage controller will suffer speed change with load change. I differ from Mark, because I think that the amount of speed change will depend on the steady state torque, which in turn depends on the steady state frictional load.
This is further complicated because we are currently talking about simple linear regulators, and these exhibit a small voltage drop with increasing load. Also, we're not all talking about the same means of measuring speed.
Matt, I hope you keep at it, and find the solution. I'd love to know (rather than merely speculate) what's going on with your TT.
Hi Pixpop,
Like yourself, I was amazed (and I think I said so at that time) that Matt had the problems he did, and especially to the extent that he experienced them.
Over many years, I have not seen this degree of problem myself, and I recall asking another poster (maybe Planet IX?) to clarify his concerns, as until this point in this thread I thought the main concerns were more over the short-term record-induced variations.
What I am personally concerned about, though, is not getting into a sparring match with anyone over these opinions, as I see this far too much on this Forum, and it is simply not necessary, nor becoming of any gentlemen. Life is too short, at least mine is, being well into my 60s!
However, when the results of my carefully investigated experiments which I have tried to share (for what they are worth) with anyone who is interested, are summarily 'shot down' with suggestions that I haven't measured something properly, It is sometimes a little difficult to counter this without ruffling any feathers.
Oversimplified statements are often not very helpful, and you are quite right (IMO) with regard to the point you made here. If any unwanted effects are minimal *in the overall scheme of things*, it cannot matter much (or as much) if they also vary sporadically, whatever any theory, or whatever else, might suggest.
For this reason, I have found it best in designing circuits/choice of parameters to 'bias' things so that these potential abberations are minimised, in *relative terms* wherever possible, and hence my comments in earlier posts. One of the other significant matters which I am sure I also mentioned somewhere before, is the also-unwanted 'noise' from any electronic circuits, which is generally so low in level, it is not intuitively a problem. However, I found that it does have a significant effect on the sound, no matter what any theory suggests, and so I cannot ignore this. This unwelcome but inevitable noise is not related linearly to the output voltages used, but is more governed by the circuits chosen, and to a large extent, the active and passive devices used therein. It is, therefore, reasonably 'static' in level, once the circuit has been finalised.
It can be minimised (at a cost!) but it still remains, even if at a very low level. Whatever this level may be, when related to say 10 volts DC, it will be (generally) 10x less significant or 'damaging' than when using say 1 volt to drive the motor, and you cannot escape from this. And yet I see no comments from anyone about this whenever I read about others' opinions, and this, just like the choice of wire used to drive the motor, does matter as I have satisfied myself.
Measuring/sight-tests etc. did not confirm to me what I had actually *heard* when I relied on conventional means of assessing TTs, unfortunately. Using test records, strobes, or whatever only worked to give an overall indication of these sporadic record-induced changes, mainly because the test records did not have varying levels of 'cuts' on them, and anyway, the frequency meters at my disposal all have a sample-rate delay, so, often any very short-term and constantly varying frequency-changes were simply missed in the counting and display process. An example is one meter has 'measurement times' of 0.1 sec., 1 sec., and even 10sec., depending on the frequency scale chosen. Due to this counting process, the necessary information I wished to see, was just not 'grabbed' and duly displayed, unfortunately.
It wasn't until I could also *see* in 'real-time' the effect of these changes, by way of the Gyrascope, that I really got on top of this, and made some further progress. Maybe this could explain why I would naturally feel a bit put-out when someone else more-or-less says I am not measuring anything properly, so I cannot be right with my chosen direction or choice of PS.
As soon as I could see what was going on, I immediately abandoned the BB circuit you referred to which (IIRC) adjusted the output voltage via back-EMF changes, or similar, as it was 'hunting' at *all* times, and never rock-steady. It was a great disappointment, but nevertheless I couldn't 'tame' it, no matter how I tried, and no matter how much better (theoretically) such a feed-back arrangement might be. I concluded that the designer's acceptance over his circuit falling short of an ideal feed-back system was an accurate description, but that *in this application* unfortunately I disagreed that it was simply better than nothing.
It had been my intention to then consider other DC circuit ideas, possibly including a separate tacho coil on the motor, considering whether a negative impedance PS might help etc., but now I could see what was happening in real time with a much more straightforward regulator which I first went back to, it just didn't appear to be necessary. Setting the speed is like falling off a log with this excellent strobe device, and with a 20 turn bulk-foil potentiometer with a tempco of merely 15 parts per million to adjust the PS output voltage (this is over a total adjustment range of merely 1.5 volts) it didn't take very long.
As I mentioned much earlier, no-one could have been more surprised than myself when I realised how stable my latest attempt turned out to be. I worried about motor temperature changes, I wondered about long-term variations, and whatever else was appropriate, but there was nothing to be concerned about. With a Gyrascope, it is an easy trick to make a wire pointer and stick this to the TT with Blu-Tack, so that it's end is right by the flashing leds which ordinarily appear to be stationary, if the TT's speed is correct. Any very slight changes in speed, even over a long period, can readily be seen by the moving leds appearing to 'creep' during motion.
25 minutes later (when I first tried this out for 'real') and at the end of the record's first side, there was not a single mm of relative movement from what my eyes told me. I watched the set-up with amazement for the entire time, and simply couldn't believe what I was seeing.
This was repeatable, any number of times, so how accurately is this being controlled percentage-wise across an entire side of a record? Apparently zero error in 832.5 (33.3 x 25) x 628.6 ( 22/7 x 200 [average mid. dia of LP?] ) = 523310 mm, which the stylus had travelled in that time. Doubtless someone will immediately correct me if my calculations are wrong here, but hopefully the essence of this point is taken.
If it had 'strayed' by 1mm (which I consider it did not) this looks to be an accuracy of 0.000002%, and presently I don't wish to better this!
Blowing or strong draughts on the TT had some overall adverse effect in these tests, as I mentioned earlier, and as I later discovered.
I am not in the habit of deliberately (nor accidentally) misleading other people, and "there are many ways to skin a cat" as the saying goes. There will undoubtedly be differences of opinion as to how any solution is arrived at, but my own experiences do not result from looking at motor specs, nor any postulations here. They are the results of many hours of pratical trials, over very many years, and, where something is not successful (the BB circuit, for example) I will happily share this opinion/conclusion with others to save them from going down the same blind alleys, as you have put it.
Regards,
Like yourself, I was amazed (and I think I said so at that time) that Matt had the problems he did, and especially to the extent that he experienced them.
Over many years, I have not seen this degree of problem myself, and I recall asking another poster (maybe Planet IX?) to clarify his concerns, as until this point in this thread I thought the main concerns were more over the short-term record-induced variations.
What I am personally concerned about, though, is not getting into a sparring match with anyone over these opinions, as I see this far too much on this Forum, and it is simply not necessary, nor becoming of any gentlemen. Life is too short, at least mine is, being well into my 60s!
However, when the results of my carefully investigated experiments which I have tried to share (for what they are worth) with anyone who is interested, are summarily 'shot down' with suggestions that I haven't measured something properly, It is sometimes a little difficult to counter this without ruffling any feathers.
Oversimplified statements are often not very helpful, and you are quite right (IMO) with regard to the point you made here. If any unwanted effects are minimal *in the overall scheme of things*, it cannot matter much (or as much) if they also vary sporadically, whatever any theory, or whatever else, might suggest.
For this reason, I have found it best in designing circuits/choice of parameters to 'bias' things so that these potential abberations are minimised, in *relative terms* wherever possible, and hence my comments in earlier posts. One of the other significant matters which I am sure I also mentioned somewhere before, is the also-unwanted 'noise' from any electronic circuits, which is generally so low in level, it is not intuitively a problem. However, I found that it does have a significant effect on the sound, no matter what any theory suggests, and so I cannot ignore this. This unwelcome but inevitable noise is not related linearly to the output voltages used, but is more governed by the circuits chosen, and to a large extent, the active and passive devices used therein. It is, therefore, reasonably 'static' in level, once the circuit has been finalised.
It can be minimised (at a cost!) but it still remains, even if at a very low level. Whatever this level may be, when related to say 10 volts DC, it will be (generally) 10x less significant or 'damaging' than when using say 1 volt to drive the motor, and you cannot escape from this. And yet I see no comments from anyone about this whenever I read about others' opinions, and this, just like the choice of wire used to drive the motor, does matter as I have satisfied myself.
Measuring/sight-tests etc. did not confirm to me what I had actually *heard* when I relied on conventional means of assessing TTs, unfortunately. Using test records, strobes, or whatever only worked to give an overall indication of these sporadic record-induced changes, mainly because the test records did not have varying levels of 'cuts' on them, and anyway, the frequency meters at my disposal all have a sample-rate delay, so, often any very short-term and constantly varying frequency-changes were simply missed in the counting and display process. An example is one meter has 'measurement times' of 0.1 sec., 1 sec., and even 10sec., depending on the frequency scale chosen. Due to this counting process, the necessary information I wished to see, was just not 'grabbed' and duly displayed, unfortunately.
It wasn't until I could also *see* in 'real-time' the effect of these changes, by way of the Gyrascope, that I really got on top of this, and made some further progress. Maybe this could explain why I would naturally feel a bit put-out when someone else more-or-less says I am not measuring anything properly, so I cannot be right with my chosen direction or choice of PS.
As soon as I could see what was going on, I immediately abandoned the BB circuit you referred to which (IIRC) adjusted the output voltage via back-EMF changes, or similar, as it was 'hunting' at *all* times, and never rock-steady. It was a great disappointment, but nevertheless I couldn't 'tame' it, no matter how I tried, and no matter how much better (theoretically) such a feed-back arrangement might be. I concluded that the designer's acceptance over his circuit falling short of an ideal feed-back system was an accurate description, but that *in this application* unfortunately I disagreed that it was simply better than nothing.
It had been my intention to then consider other DC circuit ideas, possibly including a separate tacho coil on the motor, considering whether a negative impedance PS might help etc., but now I could see what was happening in real time with a much more straightforward regulator which I first went back to, it just didn't appear to be necessary. Setting the speed is like falling off a log with this excellent strobe device, and with a 20 turn bulk-foil potentiometer with a tempco of merely 15 parts per million to adjust the PS output voltage (this is over a total adjustment range of merely 1.5 volts) it didn't take very long.
As I mentioned much earlier, no-one could have been more surprised than myself when I realised how stable my latest attempt turned out to be. I worried about motor temperature changes, I wondered about long-term variations, and whatever else was appropriate, but there was nothing to be concerned about. With a Gyrascope, it is an easy trick to make a wire pointer and stick this to the TT with Blu-Tack, so that it's end is right by the flashing leds which ordinarily appear to be stationary, if the TT's speed is correct. Any very slight changes in speed, even over a long period, can readily be seen by the moving leds appearing to 'creep' during motion.
25 minutes later (when I first tried this out for 'real') and at the end of the record's first side, there was not a single mm of relative movement from what my eyes told me. I watched the set-up with amazement for the entire time, and simply couldn't believe what I was seeing.
This was repeatable, any number of times, so how accurately is this being controlled percentage-wise across an entire side of a record? Apparently zero error in 832.5 (33.3 x 25) x 628.6 ( 22/7 x 200 [average mid. dia of LP?] ) = 523310 mm, which the stylus had travelled in that time. Doubtless someone will immediately correct me if my calculations are wrong here, but hopefully the essence of this point is taken.
If it had 'strayed' by 1mm (which I consider it did not) this looks to be an accuracy of 0.000002%, and presently I don't wish to better this!
Blowing or strong draughts on the TT had some overall adverse effect in these tests, as I mentioned earlier, and as I later discovered.
I am not in the habit of deliberately (nor accidentally) misleading other people, and "there are many ways to skin a cat" as the saying goes. There will undoubtedly be differences of opinion as to how any solution is arrived at, but my own experiences do not result from looking at motor specs, nor any postulations here. They are the results of many hours of pratical trials, over very many years, and, where something is not successful (the BB circuit, for example) I will happily share this opinion/conclusion with others to save them from going down the same blind alleys, as you have put it.
Regards,
Hi Bob
Thanks for the kind offer to re-machine my Linn pulley, unfortunately it has a hollow centre. So I do need one made from scratch. I agree that a higher drive voltage is desirable but from my experience I think some kind of load compensation is necessary.
Mark, despite my implications, I'm not ready to give up dc just yet as I've only just invested in a new motor! The circuit you made for Vinyl Addict sounds like what I'm after. I wonder whether you are willing to share it? Thanks for the information so far.
Pixpop, your encouragement is appreciated and I have found all the contributions to the thread informative, although somewahat frustrating overall as it seems going dc is more tricky than I had anticipated!
Cheers for now
Matt
Thanks for the kind offer to re-machine my Linn pulley, unfortunately it has a hollow centre. So I do need one made from scratch. I agree that a higher drive voltage is desirable but from my experience I think some kind of load compensation is necessary.
Mark, despite my implications, I'm not ready to give up dc just yet as I've only just invested in a new motor! The circuit you made for Vinyl Addict sounds like what I'm after. I wonder whether you are willing to share it? Thanks for the information so far.
Pixpop, your encouragement is appreciated and I have found all the contributions to the thread informative, although somewahat frustrating overall as it seems going dc is more tricky than I had anticipated!
Cheers for now
Matt
Bobken
I find your posts interesting but I cannot believe your claim for "perfect' speed stability, it defies the laws of physics. Where is the extra energy to compensate the drag coming from?
Also I'd be interested to know the torque / speed constant on the motor you use.
Bobken and Pixpop
you haven't grasped my point about speed / torque constants. It is the slope of the line that matters and this does not change with load. It is a fixed parameter related only to the winding and magnet geometry of the motor. Again, I'm not making this up, it's there in the manufacturers data sheets.
Pixpop
Yes my controller changes voltage with current, that's what negative impedance means.
Matt
Full plans will be in Bas Hornemann's DIY mag, I sent the last draft and the schematics yesterday. The controller will set you back about 50 quid.
Forgive the brevity but vintage starts today so I'm very busy. I may not be able to respond to any further posts. Good luck with your motor Matt.
I find your posts interesting but I cannot believe your claim for "perfect' speed stability, it defies the laws of physics. Where is the extra energy to compensate the drag coming from?
Also I'd be interested to know the torque / speed constant on the motor you use.
Bobken and Pixpop
you haven't grasped my point about speed / torque constants. It is the slope of the line that matters and this does not change with load. It is a fixed parameter related only to the winding and magnet geometry of the motor. Again, I'm not making this up, it's there in the manufacturers data sheets.
Pixpop
Yes my controller changes voltage with current, that's what negative impedance means.
Matt
Full plans will be in Bas Hornemann's DIY mag, I sent the last draft and the schematics yesterday. The controller will set you back about 50 quid.
Forgive the brevity but vintage starts today so I'm very busy. I may not be able to respond to any further posts. Good luck with your motor Matt.
Mark, I wasn't thinking you made anything up. I believe you, it's just counter to my intuition.
I've done more research, and found good articles on the Maxon site. They of course agree with you that the change in speed for a given change in torque does not depend on voltage.
Maxon info
It's hard to accept, but I guess I have to accept it. The amount of the change is, as you suggest, dependent only on the construction of the motor. They suggest that to obtain the smallest possible change in speed with change in torque, one needs to select the motor with the smallest speed/torque constant.
If one looks at the different Maxon motor specs, you'll see that the speed/torque gradient varies wildly between models.
At the rate I'm learning new stuff, I'm soon to be a bloody genius
I've done more research, and found good articles on the Maxon site. They of course agree with you that the change in speed for a given change in torque does not depend on voltage.
Maxon info
It's hard to accept, but I guess I have to accept it. The amount of the change is, as you suggest, dependent only on the construction of the motor. They suggest that to obtain the smallest possible change in speed with change in torque, one needs to select the motor with the smallest speed/torque constant.
If one looks at the different Maxon motor specs, you'll see that the speed/torque gradient varies wildly between models.
At the rate I'm learning new stuff, I'm soon to be a bloody genius
Hi Mark,
Lets face it, we both know there is no such thing as absolute perfection in any engineering, and I have said so several times earlier, I am sure. I also alluded to the fact that my present set-up *must* vary to some degree (in post #62), but that I have now minimised any variations to the extent that I cannot now hear this, nor can I measure it. I admit, perhaps somewhat tongue-in-cheek I suggested 0.000002%, or whatever, which is not zero, but in my book and relative to all these other claims I have read about, I don't know how to improve on this. Other uncontrollable extraneous factors dominate now.
I would dearly like to claim some super-human abilities, or perhaps suggest there is some mystique here, but this would not be honest.
I am certainly not claiming to defy gravity, to disprove Einstein's Theory, nor to have invented a perpetual motion machine, or whatever else!
What I have succeeded in doing by careful engineering practices, a great deal of practical experimentation/measurements, and perhaps a bit of good fortune (which provided the icing on the cake) with my latest version of the PS, is to reduce the entire stylus drag effects to zero, in all practical respects.
DPS have apparently done something similar, also using a very high torque motor (albeit in their case a Berger synch. AC motor was chosen) whereafter they say "cartridge drag becomes negligible".
You cannot divorce motors nor a PS from the entire set-up, which includes the TT itself, its bearing, the method of drive etc., and several other factors, and I have been working on this concept 'as a whole' for very many years. Thinking 'outside of the box' and not slavishly following everyone else in considering the solutions to good-sounding TTs, is sometimes a good thing, but I will say no more on this.
In my experience, cartridge drag (however it is caused) is actually very low, although, maybe as quantifying it would not help me eliminate it, I have not yet succeeded in actually measuring it to satisfy myself over this. It is precisely for this reason that I, and apparently several others, find it so hard to believe that Matt (for example) is having so much of a problem with this, but clearly he is.
In an earlier post, I stated that I can grab my belt and pinch it quite hard between my fingers before any noticeable speed reduction occurs, but no-one seemed to pick up on this. Again, I'll say no more, except that such a braking force as this, must surely exceed any stylus drag by several orders of magnitude.
When you reach this rather happy state of affairs, the relatively minor matter of dealing with concerns relating to the effects from playing highly-modulated passages, which has been the essence of my entire postings in this thread, then takes on some significance. From my experiments, I depart from what you (and maybe many others) appear to believe regarding the two obvious (not only!) different possibilities for the causes of stylus drag.
I consider the 'modulation' variations to be of far greater magnitude/significance than any changes which can occur across a record, which are merely due to the reduced 'surface speed' commensurate with the record's reducing radius, and therefore the circumference which the stylus needs to navigate in one revolution.
Another point is that I do not always assume that manufacturers know everything there is to know about their products, nor slavishly follow their stated parameters/specifications without question. Many is the time (mostly in other areas of engineering) when I have pointed out to a manufacturer that he does not fully understand certain features/parameters of his product, or that his stated parameters/specs. (when carefully measured) are not correct, unfortunately. Just because he makes it, it doesn't necessarily follow that he knows everything about it, that he is always entirely open and honest about it, nor that he has spent maybe years in diligent experimentation with it, probably because he is rather understandably not always motivated to do this.
I cannot divulge details of the motor I use, and it is a world apart from what most other designers appear to be using, but torque/ speed constants (if they really are accurately described) don't make me lose any sleep, as I am not hung up on them like some are.
Having been born towards the end of WW11, when there was nothing 'worth having' available for many ensuing years so anything one needed had to be made by one's self, tought me to be innovative and I became a 'hands-on' and very inquisitive kind of person, out of necessity. Accordingly, I don't blindly accept everything at face value (until I have checked it out personally) and spec. sheets are OK as a general guideline, but you do need to ensure that any parameters on which you duly rely, are actually reliable.
I am very sorry to see you apparently excusing yourself from further participation in this thread. As you jumped into the thread with such bold and categoric statements over others not measuring the speed of TTs properly, I was hoping that you might share some of your secrets in this respect, to help the rest of us.
Personally, having gone to such lengths to explain my own methods of accurately checking TT speeds under all 'real-life' circumstances, I hoped you would return the favour and explain just how you verify your own results and reach the conclusions you have over the accuracy of your designs. I would be especially interested to know how the 0.1% speed variation (*at the TT*) mentioned by you, was actually *measured* on Vinyl Addict's truly excellent-looking creation.
Regards,
Regret several spelling Edits!
Lets face it, we both know there is no such thing as absolute perfection in any engineering, and I have said so several times earlier, I am sure. I also alluded to the fact that my present set-up *must* vary to some degree (in post #62), but that I have now minimised any variations to the extent that I cannot now hear this, nor can I measure it. I admit, perhaps somewhat tongue-in-cheek I suggested 0.000002%, or whatever, which is not zero, but in my book and relative to all these other claims I have read about, I don't know how to improve on this. Other uncontrollable extraneous factors dominate now.
I would dearly like to claim some super-human abilities, or perhaps suggest there is some mystique here, but this would not be honest.
I am certainly not claiming to defy gravity, to disprove Einstein's Theory, nor to have invented a perpetual motion machine, or whatever else!
What I have succeeded in doing by careful engineering practices, a great deal of practical experimentation/measurements, and perhaps a bit of good fortune (which provided the icing on the cake) with my latest version of the PS, is to reduce the entire stylus drag effects to zero, in all practical respects.
DPS have apparently done something similar, also using a very high torque motor (albeit in their case a Berger synch. AC motor was chosen) whereafter they say "cartridge drag becomes negligible".
You cannot divorce motors nor a PS from the entire set-up, which includes the TT itself, its bearing, the method of drive etc., and several other factors, and I have been working on this concept 'as a whole' for very many years. Thinking 'outside of the box' and not slavishly following everyone else in considering the solutions to good-sounding TTs, is sometimes a good thing, but I will say no more on this.
In my experience, cartridge drag (however it is caused) is actually very low, although, maybe as quantifying it would not help me eliminate it, I have not yet succeeded in actually measuring it to satisfy myself over this. It is precisely for this reason that I, and apparently several others, find it so hard to believe that Matt (for example) is having so much of a problem with this, but clearly he is.
In an earlier post, I stated that I can grab my belt and pinch it quite hard between my fingers before any noticeable speed reduction occurs, but no-one seemed to pick up on this. Again, I'll say no more, except that such a braking force as this, must surely exceed any stylus drag by several orders of magnitude.
When you reach this rather happy state of affairs, the relatively minor matter of dealing with concerns relating to the effects from playing highly-modulated passages, which has been the essence of my entire postings in this thread, then takes on some significance. From my experiments, I depart from what you (and maybe many others) appear to believe regarding the two obvious (not only!) different possibilities for the causes of stylus drag.
I consider the 'modulation' variations to be of far greater magnitude/significance than any changes which can occur across a record, which are merely due to the reduced 'surface speed' commensurate with the record's reducing radius, and therefore the circumference which the stylus needs to navigate in one revolution.
Another point is that I do not always assume that manufacturers know everything there is to know about their products, nor slavishly follow their stated parameters/specifications without question. Many is the time (mostly in other areas of engineering) when I have pointed out to a manufacturer that he does not fully understand certain features/parameters of his product, or that his stated parameters/specs. (when carefully measured) are not correct, unfortunately. Just because he makes it, it doesn't necessarily follow that he knows everything about it, that he is always entirely open and honest about it, nor that he has spent maybe years in diligent experimentation with it, probably because he is rather understandably not always motivated to do this.
I cannot divulge details of the motor I use, and it is a world apart from what most other designers appear to be using, but torque/ speed constants (if they really are accurately described) don't make me lose any sleep, as I am not hung up on them like some are.
Having been born towards the end of WW11, when there was nothing 'worth having' available for many ensuing years so anything one needed had to be made by one's self, tought me to be innovative and I became a 'hands-on' and very inquisitive kind of person, out of necessity. Accordingly, I don't blindly accept everything at face value (until I have checked it out personally) and spec. sheets are OK as a general guideline, but you do need to ensure that any parameters on which you duly rely, are actually reliable.
I am very sorry to see you apparently excusing yourself from further participation in this thread. As you jumped into the thread with such bold and categoric statements over others not measuring the speed of TTs properly, I was hoping that you might share some of your secrets in this respect, to help the rest of us.
Personally, having gone to such lengths to explain my own methods of accurately checking TT speeds under all 'real-life' circumstances, I hoped you would return the favour and explain just how you verify your own results and reach the conclusions you have over the accuracy of your designs. I would be especially interested to know how the 0.1% speed variation (*at the TT*) mentioned by you, was actually *measured* on Vinyl Addict's truly excellent-looking creation.
Regards,
Regret several spelling Edits!
I think I can answer your question.Bobken said:
I measured the speed using a KAB strobe. I also measured the voltage fluctuation and calculated drift based on the reading on a digital voltmeter.
We also checked the performance of the motor on the scope FWIW. I'm not an EE but the person I had perform the measurements is. He said the motor performed well within 0.1% tolerance zone.
Hi Vinyl-Addict,
Thanks very much for joining in here.
May I say that your superb creation is one of the finest looking TTs I have ever seen, and I have seen quite a few over the years. You must be very proud of your work here.
The machined finish looks to be exemplary, and I wonder if you needed to resort to diamond turning tools to obtain this?
In about 1960, I had the (short) privilege of operating the first diamond milling machine to be seen in the UK. This was imported from Switzerland by the Dennison Watch-Case Company, and the finish straight from the tool was so good that no further polishing/finishing process could improve on it.
I especially like your additional underslung brass bob-weights, which not only seriously enhance the appearance, but if my experiences are anything to go by, this will help a lot with reducing record-modulation 'braking effects', too. I have found these short-term constantly-changing variations to be harder to quantify (capture) and eliminate, and the best way to overcome this appears to be with 'brute-force' engineering. i.e. increase the rotational mass (especially right at the periphery, where it has the greatest effect, just like you have done.)
Although others have criticised heavy-weight platters, the only downsides I have found relate to longer 'start-up' times to full speed, and possibly marginally greater main bearing wear. Neither of these is problematical for me, and there is no discernible effect (that I can hear) which is adverse to the overall sound.
Interestingly, I have ended up with a platter arrangement which is almost like a twin to yours, including the gold-plated brass weights, and the more mass I added here, the better and more stable the sound became.
I confess I am rather puzzled by your explanations over the *TT's speed variations*, though, with this PS.
Most of these comments relate to the *motor's* performance, and because of the elastic nature of (most) drive belts I found that there is not a sufficiently tight relationship between motor and platter to be able to rely on measuring how the motor performs. I see you are trying out a flat belt, and if this is made of less stretchy material like the suggested vinyl cassette tapes (which I have tried briefly, but didn't like what happened to the sound, unfortunately) this would of course provide a 'tighter' coupling between platter and motor, all other things being equal.
I have also experimented with duplex or dual belts for a while, but these seem to have certain downsides, sonically, regrettably.
Anyway, what I cannot figure out (unless the KAB strobe has completely changed in its construction) is how you can use this device to carry out any 'real-time' tests across a record to assess any speed changes seen at the TT. Regrettably, as with all other similar devices, I felt obliged to abandon this idea when I researched it a while ago, in favour of another (6x more expensive, unfortunately!) tool, as otherwise any record placed on top of such a device will mask seeing any worthwhile results.
IIRC, the KAB permitted setting speeds 'statically' with a short grooved portion on the strobe, but this does not allow any of the 'dynamic' speed checks which matter most here, and which is what speed stability is all about in this case.
How can you compare speeds say at the start of a conventional 12" record, with the speed at the end of the side?
Also, and more importantly in my experience, how can you assess any short-term speed changes due to the wildly-differing groove modulations which appear on (most) records?
The only way of doing this to any worthwhile effect is if you can see what is going on whilst actually playing a 'normal' record (from start to finish) and I simply don't understand how this can be done with a strobe like the KAB.
I am certain many DIYers would like to know how you do this, as unnecessary expenditure on these ridiculously-priced 'tools' as I felt obliged to finally employ, can only usually be justified (as in my case) if there is some commercial 'return'.
Regards,
More spelling edits!
Thanks very much for joining in here.
May I say that your superb creation is one of the finest looking TTs I have ever seen, and I have seen quite a few over the years. You must be very proud of your work here.
The machined finish looks to be exemplary, and I wonder if you needed to resort to diamond turning tools to obtain this?
In about 1960, I had the (short) privilege of operating the first diamond milling machine to be seen in the UK. This was imported from Switzerland by the Dennison Watch-Case Company, and the finish straight from the tool was so good that no further polishing/finishing process could improve on it.
I especially like your additional underslung brass bob-weights, which not only seriously enhance the appearance, but if my experiences are anything to go by, this will help a lot with reducing record-modulation 'braking effects', too. I have found these short-term constantly-changing variations to be harder to quantify (capture) and eliminate, and the best way to overcome this appears to be with 'brute-force' engineering. i.e. increase the rotational mass (especially right at the periphery, where it has the greatest effect, just like you have done.)
Although others have criticised heavy-weight platters, the only downsides I have found relate to longer 'start-up' times to full speed, and possibly marginally greater main bearing wear. Neither of these is problematical for me, and there is no discernible effect (that I can hear) which is adverse to the overall sound.
Interestingly, I have ended up with a platter arrangement which is almost like a twin to yours, including the gold-plated brass weights, and the more mass I added here, the better and more stable the sound became.
I confess I am rather puzzled by your explanations over the *TT's speed variations*, though, with this PS.
Most of these comments relate to the *motor's* performance, and because of the elastic nature of (most) drive belts I found that there is not a sufficiently tight relationship between motor and platter to be able to rely on measuring how the motor performs. I see you are trying out a flat belt, and if this is made of less stretchy material like the suggested vinyl cassette tapes (which I have tried briefly, but didn't like what happened to the sound, unfortunately) this would of course provide a 'tighter' coupling between platter and motor, all other things being equal.
I have also experimented with duplex or dual belts for a while, but these seem to have certain downsides, sonically, regrettably.
Anyway, what I cannot figure out (unless the KAB strobe has completely changed in its construction) is how you can use this device to carry out any 'real-time' tests across a record to assess any speed changes seen at the TT. Regrettably, as with all other similar devices, I felt obliged to abandon this idea when I researched it a while ago, in favour of another (6x more expensive, unfortunately!) tool, as otherwise any record placed on top of such a device will mask seeing any worthwhile results.
IIRC, the KAB permitted setting speeds 'statically' with a short grooved portion on the strobe, but this does not allow any of the 'dynamic' speed checks which matter most here, and which is what speed stability is all about in this case.
How can you compare speeds say at the start of a conventional 12" record, with the speed at the end of the side?
Also, and more importantly in my experience, how can you assess any short-term speed changes due to the wildly-differing groove modulations which appear on (most) records?
The only way of doing this to any worthwhile effect is if you can see what is going on whilst actually playing a 'normal' record (from start to finish) and I simply don't understand how this can be done with a strobe like the KAB.
I am certain many DIYers would like to know how you do this, as unnecessary expenditure on these ridiculously-priced 'tools' as I felt obliged to finally employ, can only usually be justified (as in my case) if there is some commercial 'return'.
Regards,
More spelling edits!
Bobken said:
Thank you Bob for the kind remarks regarding my turntable. I used standard carbide turning tools to machine the components. I am fortunate to have access to the necessary machine tools to perform such a project.
I am not educated in any way to comment on the engineering efforts of Mark's DC drive. I only posted what I did to measure the speed of the platter.
Initially I had an EE friend measure the output voltage of the drive to measure drift. Regarding the KAB strobe, I did measure the speed with a record on the platter and the KAB strobe on top of the record. I then placed the stylus in the groove and watched the strobe as the record was playing. I could only do this for a short period since eventually the stylus would come in contact with the strobe disc.
In Mark's defense he has asked me to perform measurements and I have not been able to do this as of yet because of workload.
I wish I knew although I'm sure I could machine a smaller disc with graduations timed at the correct intervals so I could measure at the end of a record. Can you share your methods for measuring? Sorry if you have already explained this.
Hi Guys
I can finally do something other than worrying about pH and sulphur levels in sparkling base for five minutesso I thought I'd post a response. I don't have time to address all the issues raised in Bobken's posts.
Bob if I amended your original claim to "My (Bobken's) motor gives unmeasureable speed variation using a linear power supply" would we be able to agree on that?
I would then amend my original statement to say " Anyone who claims apparently stable speed has either a measurement technique with a resolution above 0.1% speed variability or a motor with an effective / speed torque constant less than 0.2 rpm / mNm (assuming 300rpm drive eg 9:1 torque ratio, the torque ratio and the required speed / torque constant scale together)"
We seem to be in agreement that a low effective speed / torque constant is a good thing. Note that I said "effective speed / torque constant". If you refer back to my original post, I showed that the effective speed / torque constant for a motor is given by the ratio between speed constant and torque constant, multiplied by the sum of the winding resistance and the supply resistance. The manufacturer's quoted speed / torque constant is exactly the constant given by assuming a drive resistance of zero. Where we differ is in the way we achieve the low speed / torque constant.
I advocate using a small motor and reducing its effective speed / torque constant by trimming the drives output resistance so it mirrors the winding resistance. I came to this through an analysis of the theory and my practical results appear to confirm the theory.
I haven't tried your approach of using a motor with a low speed / torque constant as per manufacturer's specification so I can make no comment on it.
I can finally do something other than worrying about pH and sulphur levels in sparkling base for five minutesso I thought I'd post a response. I don't have time to address all the issues raised in Bobken's posts.
Bob if I amended your original claim to "My (Bobken's) motor gives unmeasureable speed variation using a linear power supply" would we be able to agree on that?
I would then amend my original statement to say " Anyone who claims apparently stable speed has either a measurement technique with a resolution above 0.1% speed variability or a motor with an effective / speed torque constant less than 0.2 rpm / mNm (assuming 300rpm drive eg 9:1 torque ratio, the torque ratio and the required speed / torque constant scale together)"
We seem to be in agreement that a low effective speed / torque constant is a good thing. Note that I said "effective speed / torque constant". If you refer back to my original post, I showed that the effective speed / torque constant for a motor is given by the ratio between speed constant and torque constant, multiplied by the sum of the winding resistance and the supply resistance. The manufacturer's quoted speed / torque constant is exactly the constant given by assuming a drive resistance of zero. Where we differ is in the way we achieve the low speed / torque constant.
I advocate using a small motor and reducing its effective speed / torque constant by trimming the drives output resistance so it mirrors the winding resistance. I came to this through an analysis of the theory and my practical results appear to confirm the theory.
I haven't tried your approach of using a motor with a low speed / torque constant as per manufacturer's specification so I can make no comment on it.
Hi Vinyl-Addict,
Thanks for this response.
I am sorry you felt the need to use the word "defense" relating to Mark here, as I bear him no ill will, and I am certainly not deliberately attempting to attack him.
I simply found myself in a difficult situation, where I happen to be confident of my facts, and for whatever reason, Mark chose to rather openly 'walk all over' these.
As Mark appears to have re-joined the discussion, I will respond to him on the speed variation issues, as it was he who brought them up.
If you are interested in my methods of assessing speed variations of TTs 'in normal use' (as opposed to perhaps idling along unloaded) then I have written quite a lot on this thread, already. I don't think it would help anyone for me to repeat that, and I cannot add anything worthwhile to what I have said.
I say again, your TT is truly excellent in appearance, and do I hope that with Mark's (or anyone elses) PS, you experience some wonderful sounds from it, to match its appearance. Such an effort as you have already made, surely deserves an excellent sound, and this, after all, is what it is all about.
Regards,
Thanks for this response.
I am sorry you felt the need to use the word "defense" relating to Mark here, as I bear him no ill will, and I am certainly not deliberately attempting to attack him.
I simply found myself in a difficult situation, where I happen to be confident of my facts, and for whatever reason, Mark chose to rather openly 'walk all over' these.
As Mark appears to have re-joined the discussion, I will respond to him on the speed variation issues, as it was he who brought them up.
If you are interested in my methods of assessing speed variations of TTs 'in normal use' (as opposed to perhaps idling along unloaded) then I have written quite a lot on this thread, already. I don't think it would help anyone for me to repeat that, and I cannot add anything worthwhile to what I have said.
I say again, your TT is truly excellent in appearance, and do I hope that with Mark's (or anyone elses) PS, you experience some wonderful sounds from it, to match its appearance. Such an effort as you have already made, surely deserves an excellent sound, and this, after all, is what it is all about.
Regards,
Hi Mark,
I am glad to see you haven't abandoned the thread at this point, but, from what you last said, this did seem very likely. Also, I apologise to other posters if this matter is becoming boring and unnecessarily protracted, and, as I said before, I didn't wish to indulge in any 'sparring matches' nor be dragged into any protracted arguments/disagreements.
Please also bear in mind my intention in my own posts was merely to help others who may not all have had the benefit of many years experience in this field, and I have been attempting to suggest practical and hopefully readily-understood ways of overcoming some of the problems which have been exposed here. At no time have I ever said anyone *must* follow what I say here, nor that this is the *only* method of achieving success, and I have no wish to 'sell' them any ideas nor concepts .
I have made it very clear that there is more than one solution to the unwanted speed-variation effects which have befallen some people, but I will only usually comment on what I have actually experienced for myself, and suggest some 'generalisations' which I hope may be of interest to other members.
With respect, I didn't at any time make an "original claim" along the lines of what you now infer, so I don't feel inclined to "amend", as you put it, nor modify anything which I have said before.
Instead, typically what I have said is "I could not reliably see any repeatable speed variations when using my current PS when comparing results at the start or end (or anywhere in between) of these tests." However, I have also suggested several times that nothing in life nor in engineering is ever absolutely perfect, and that there *must* also be some variations in my set-up, albeit that I cannot now hear nor measure them.
I am not entirely divorced from theory, as it happens, even if I do choose to 'explore' it and sometimes question its limitations, occasionally.
A couple of other posters have said that with their quite basic '3-terminal' regulators, they also don't experience *any* such 'beginning-to-end-of-side' problems, so I confess to having been somewhat dismayed when you categorically stated that we cannot be "measuring speed properly". This was followed up by your subsequently saying that you *calculate* under certain circumstances a (almost inevitable?) speed variation of 1% at the motor (which is colossal in my book) but which does accord with several posters' experiences, unfortunately for them. You also later implied, when I asked "Could you kindly tell us how you *verify* your own results empirically", that the only *measurement* relating to the TT's speed variations (and at the TT is all that matters here, may I observe) with your own PS and which are apparently 0.1% (or less?), was actually *measured* by Vinyl-Addict.
Vinyl-Addict has now confirmed that most of these "measurements" were made at the *motor* (and I don't know how his EE acquaintance can possibly have learned much about very minor speed variations of a *DC* motor under varying loads by looking at anything with a 'scope, as I certainly haven't managed to do this) but this is explained by his honest acceptance that he is not au fait with these procedures, anyway. Unless one is dealing with direct-drive motors, it doesn't matter too much (and within reason) what the motor does because of the 'loose coupling' provided by the belt, so long as the *platter* performs well. However, and whilst it is quite clear that 'unstable' motors/PS are quite useless here, even very good measurements at the motor do not satisfactorily predict what will happen at the platter, as I have found. What Vinyl-Addict has made clear though, is that he has not (yet) even been able to perform the most essential tests here which would establish or verify how his TT performs with anyone's PS, under 'real-life' conditions, which is the very essence of what (most of) this thread has been about.
Frankly, I am not quite sure how helpful most of this has been to others, as it seems a bit like it is an 'advert' for your own solution, regardless, and there was some obviously implied criticism of several others' (quite effective) approaches to these problems. Maybe not everyone will wish to build your design, though, and I have found that it is not the only way of solving these issues.
There are so many other matters I could go into detail over, but I am beginning to lose interest myself as I didn't come here to learn anything special, mainly because I am presently satisfied with my own results. These results *I have* empirically verified (very carefully) and under all known 'real-life' conditions, and I have gone to great lengths to explain precisely how I have done this, together with what I have assessed, and the conclusions I reached, more recently using the (claimed) worlds most accurate 'tool' for this purpose.
Briefly, for example, why do you assume an apparent 'restriction' of somewhere near to " 300 RPM drive, or a 9:1 torque ratio" which you mention here?
Have you never heard of idler wheels and/or secondary belts being used, which broadens one's scope here enormously? These can and do also assist with divorcing any motor vibration from the platter, and can allow vastly differing motor speeds and DC voltages to be used.
Contrary to what "you advocate" I have sometimes found small (especially DC) motors to be less reliable and often noisy, and there have been a host of reports about them being heard right across a room! Several makers specs. here are appalling, and seldom did I see anything worthwhile about any 'side-loading' effects caused by stretchy belts being used as most of do, nor inherent friction changes with speed variations. Has this improved since I last investigated these areas?
There is not *only* one way of dealing with back EMF effects emanating from DC motors.
Why even restrict one's self to using one motor, for that matter? Some quite acceptable motors are not frighteningly expensive. At one time, I had three on the go, with some very interesting results, but my comments on this thread have been restricted to using single motors so as not to cause any confusion.
These are just a few examples, and the list is almost endless when one starts to think outside of the box which unfortunately appears to trammel less resourceful and less ingenious designers' minds.
I have already acknowledged that there are some benefits to be seen with 'feedback' (both in amplifiers, and electro-mechanical devices, such as here), but there are inevitably some down-sides too, unfortunately. Every designer's solution is likely to be a carefully-chosen set of compromises, and mine seem to be very different from yours, not surprisingly. Here, what matters to most listeners is how the entire set-up *sounds* when a record is being played, and not what the motor manufacturers' spec. sheets might indicate, nor what any theory would suggest, or whatever else, so I just don't get hung up on these.
Regarding your remarks relating to an earlier point of mine (and Pixpop's, IIRC) I suggest that it is you who are not '"getting the point", rather than us. Maybe this example will illustrate what is meant here.
I have a large and a small bird's feather in my hands, and an empty matchbox on my bench.
With the large (wing) feather I can move the matchbox , overcoming its friction, and more importantly its potential energy, but the small (breast) feather will not budge it, as it simply isn't strong enough to apply the necesssary force. There is a profound difference here by way of a result, and here extremely weak forces (although different in magnitude between the 2 feathers) have a totally different influence on the also *low mass* box, as one feather moves it, whereas the other does not.
If I try the same trick with the same feathers on a house-brick, neither will budge the brick at all, due to its higher potential energy (and some friction) of its greater mass.
The only difference here is the much greater potential energy (kinetic energy with a rotating TT) and some additional friction associated with the 'massier' brick. The forces exerted by the two feathers are exactly the same as when they are used to push against the matchbox, but you can no longer differentiate between these two forces. (i.e. there is little 'practical' effect, or difference.)
Using different feathers (which consequently apply different forces) is akin to the small variations in drag caused by the stylus, say from the start to the end of a record. The higher kinetic energy of my deliberately-chosen much massier TT is akin to the brick which is (to all practical purposes) unaffected by either of these different forces. This is simply because there are 'orders of magnitude' variations between the forces relative to the feathers and the forces relative to the brick, unlike those with the feathers in relation to the matchbox, which are very much closer in magnitude.
Notice here I don't say there is *no* effect, either with the brick or the massier TT. In relative terms, it is simply reduced to insignificance, as another TT manufacturer has apparently discovered to his benefit.
***I'm told by the system, I must post the remainder of this in another post as there is a limit to the characters.***
I am glad to see you haven't abandoned the thread at this point, but, from what you last said, this did seem very likely. Also, I apologise to other posters if this matter is becoming boring and unnecessarily protracted, and, as I said before, I didn't wish to indulge in any 'sparring matches' nor be dragged into any protracted arguments/disagreements.
Please also bear in mind my intention in my own posts was merely to help others who may not all have had the benefit of many years experience in this field, and I have been attempting to suggest practical and hopefully readily-understood ways of overcoming some of the problems which have been exposed here. At no time have I ever said anyone *must* follow what I say here, nor that this is the *only* method of achieving success, and I have no wish to 'sell' them any ideas nor concepts .
I have made it very clear that there is more than one solution to the unwanted speed-variation effects which have befallen some people, but I will only usually comment on what I have actually experienced for myself, and suggest some 'generalisations' which I hope may be of interest to other members.
With respect, I didn't at any time make an "original claim" along the lines of what you now infer, so I don't feel inclined to "amend", as you put it, nor modify anything which I have said before.
Instead, typically what I have said is "I could not reliably see any repeatable speed variations when using my current PS when comparing results at the start or end (or anywhere in between) of these tests." However, I have also suggested several times that nothing in life nor in engineering is ever absolutely perfect, and that there *must* also be some variations in my set-up, albeit that I cannot now hear nor measure them.
I am not entirely divorced from theory, as it happens, even if I do choose to 'explore' it and sometimes question its limitations, occasionally.
A couple of other posters have said that with their quite basic '3-terminal' regulators, they also don't experience *any* such 'beginning-to-end-of-side' problems, so I confess to having been somewhat dismayed when you categorically stated that we cannot be "measuring speed properly". This was followed up by your subsequently saying that you *calculate* under certain circumstances a (almost inevitable?) speed variation of 1% at the motor (which is colossal in my book) but which does accord with several posters' experiences, unfortunately for them. You also later implied, when I asked "Could you kindly tell us how you *verify* your own results empirically", that the only *measurement* relating to the TT's speed variations (and at the TT is all that matters here, may I observe) with your own PS and which are apparently 0.1% (or less?), was actually *measured* by Vinyl-Addict.
Vinyl-Addict has now confirmed that most of these "measurements" were made at the *motor* (and I don't know how his EE acquaintance can possibly have learned much about very minor speed variations of a *DC* motor under varying loads by looking at anything with a 'scope, as I certainly haven't managed to do this) but this is explained by his honest acceptance that he is not au fait with these procedures, anyway. Unless one is dealing with direct-drive motors, it doesn't matter too much (and within reason) what the motor does because of the 'loose coupling' provided by the belt, so long as the *platter* performs well. However, and whilst it is quite clear that 'unstable' motors/PS are quite useless here, even very good measurements at the motor do not satisfactorily predict what will happen at the platter, as I have found. What Vinyl-Addict has made clear though, is that he has not (yet) even been able to perform the most essential tests here which would establish or verify how his TT performs with anyone's PS, under 'real-life' conditions, which is the very essence of what (most of) this thread has been about.
Frankly, I am not quite sure how helpful most of this has been to others, as it seems a bit like it is an 'advert' for your own solution, regardless, and there was some obviously implied criticism of several others' (quite effective) approaches to these problems. Maybe not everyone will wish to build your design, though, and I have found that it is not the only way of solving these issues.
There are so many other matters I could go into detail over, but I am beginning to lose interest myself as I didn't come here to learn anything special, mainly because I am presently satisfied with my own results. These results *I have* empirically verified (very carefully) and under all known 'real-life' conditions, and I have gone to great lengths to explain precisely how I have done this, together with what I have assessed, and the conclusions I reached, more recently using the (claimed) worlds most accurate 'tool' for this purpose.
Briefly, for example, why do you assume an apparent 'restriction' of somewhere near to " 300 RPM drive, or a 9:1 torque ratio" which you mention here?
Have you never heard of idler wheels and/or secondary belts being used, which broadens one's scope here enormously? These can and do also assist with divorcing any motor vibration from the platter, and can allow vastly differing motor speeds and DC voltages to be used.
Contrary to what "you advocate" I have sometimes found small (especially DC) motors to be less reliable and often noisy, and there have been a host of reports about them being heard right across a room! Several makers specs. here are appalling, and seldom did I see anything worthwhile about any 'side-loading' effects caused by stretchy belts being used as most of do, nor inherent friction changes with speed variations. Has this improved since I last investigated these areas?
There is not *only* one way of dealing with back EMF effects emanating from DC motors.
Why even restrict one's self to using one motor, for that matter? Some quite acceptable motors are not frighteningly expensive. At one time, I had three on the go, with some very interesting results, but my comments on this thread have been restricted to using single motors so as not to cause any confusion.
These are just a few examples, and the list is almost endless when one starts to think outside of the box which unfortunately appears to trammel less resourceful and less ingenious designers' minds.
I have already acknowledged that there are some benefits to be seen with 'feedback' (both in amplifiers, and electro-mechanical devices, such as here), but there are inevitably some down-sides too, unfortunately. Every designer's solution is likely to be a carefully-chosen set of compromises, and mine seem to be very different from yours, not surprisingly. Here, what matters to most listeners is how the entire set-up *sounds* when a record is being played, and not what the motor manufacturers' spec. sheets might indicate, nor what any theory would suggest, or whatever else, so I just don't get hung up on these.
Regarding your remarks relating to an earlier point of mine (and Pixpop's, IIRC) I suggest that it is you who are not '"getting the point", rather than us. Maybe this example will illustrate what is meant here.
I have a large and a small bird's feather in my hands, and an empty matchbox on my bench.
With the large (wing) feather I can move the matchbox , overcoming its friction, and more importantly its potential energy, but the small (breast) feather will not budge it, as it simply isn't strong enough to apply the necesssary force. There is a profound difference here by way of a result, and here extremely weak forces (although different in magnitude between the 2 feathers) have a totally different influence on the also *low mass* box, as one feather moves it, whereas the other does not.
If I try the same trick with the same feathers on a house-brick, neither will budge the brick at all, due to its higher potential energy (and some friction) of its greater mass.
The only difference here is the much greater potential energy (kinetic energy with a rotating TT) and some additional friction associated with the 'massier' brick. The forces exerted by the two feathers are exactly the same as when they are used to push against the matchbox, but you can no longer differentiate between these two forces. (i.e. there is little 'practical' effect, or difference.)
Using different feathers (which consequently apply different forces) is akin to the small variations in drag caused by the stylus, say from the start to the end of a record. The higher kinetic energy of my deliberately-chosen much massier TT is akin to the brick which is (to all practical purposes) unaffected by either of these different forces. This is simply because there are 'orders of magnitude' variations between the forces relative to the feathers and the forces relative to the brick, unlike those with the feathers in relation to the matchbox, which are very much closer in magnitude.
Notice here I don't say there is *no* effect, either with the brick or the massier TT. In relative terms, it is simply reduced to insignificance, as another TT manufacturer has apparently discovered to his benefit.
***I'm told by the system, I must post the remainder of this in another post as there is a limit to the characters.***
As is often the way when solving problems, having had different backgrounds we have trodden very different paths, it seems, but I would never categorically state that my way is the only way, nor cast doubts on the activities of others without first knowing much about their work. Making assumptions, and consequently sweeping statements, especially if they are of a critical nature, is not always wise in my experience, and it doesn't make many friends. I won't (usually!) fall out with anyone over it though, as it is not worth it.
I certainly don't pretend to know *all* about this sphere of audio 'expertise', but I will always keep an open mind and, by doing so, I learn something new every day as a bonus.
Finally, I am sure your PS design is fine, even if it has not perhaps yet been proven to the extent which mine has (approx. 2 yrs. in the field), and I do hope that it will be of benefit to other DIYers, if and when they construct it, as I openly encourage them to do. I mean you no disrespect in saying this , but if you feel that you already know all there is to know about this very complex area, whether it be merely related to motors, themselves, or to the much more important combinations of motor/belt/bearing/platter-mass etc., then I fear that there are few surprises in store for you in the future.
It is sometimes to my regret that I cannot give any more details of my own set-up, but that is just the way it is. I have tried to post a lot of clues, however.
Regards,
I certainly don't pretend to know *all* about this sphere of audio 'expertise', but I will always keep an open mind and, by doing so, I learn something new every day as a bonus.
Finally, I am sure your PS design is fine, even if it has not perhaps yet been proven to the extent which mine has (approx. 2 yrs. in the field), and I do hope that it will be of benefit to other DIYers, if and when they construct it, as I openly encourage them to do. I mean you no disrespect in saying this , but if you feel that you already know all there is to know about this very complex area, whether it be merely related to motors, themselves, or to the much more important combinations of motor/belt/bearing/platter-mass etc., then I fear that there are few surprises in store for you in the future.
It is sometimes to my regret that I cannot give any more details of my own set-up, but that is just the way it is. I have tried to post a lot of clues, however.
Regards,
Bobken said:
Hi Bob, I guess I misled you by mentioning the scope. He merely used the scope to look at vibration I believe and complex impedance. I can assure you he is quite capable of analyzing the performance of a motor, AC or DC. I guess I should have kept my mouth shut because it seems that my comments are misleading the more technical minded because of my ignorance.
I can also assure you that regardless of the 0.1% variation I measured both on the voltmeter and the strobe that I am very happy with its performance.
I will leave it to the technical lads from now on, sorry for the confusion.
No harm done
Hi Vinyl-Addict,
There's no need for any apology, and actually you didn't really mislead me to any extent, anyway. I guessed you were rather 'out of your depth' here, which is quite understandable, and I realised there was no deliberate untoward intent on your part.
However, it would have been presumptuous and ill-mannered for me to do what I have seen all too often on the Forum, and simply say you are just talking out of your ***, or whatever, and that is not my way.
The background leading up to your involvement was admittedly unfortunate, but having been categorically told that I (and some others) couldn't be measuring our *TTs* properly, and your *TT* was given as the *only* example of measuring Mark's own results, I suppose it was inevitable. It is perhaps slightly ironic, that this reference has turned out to be somewhat unsubstantiated for now.
I am glad you are happy with the performance of Mark's PS, and remember that at no time have I attempted to cast any doubts on its effectiveness, nor anything else. Indeed I would be foolish to do so, as I have not 'heard' it in use.
Unfortunately, the same cannot truthfully be said for the converse, but sometimes that is the way of things in life.
I have merely observed that this might not be the only way of dealing with the issues, and I have not had good experiences with circuits which vary their outputs according to the load, due to the unavoidable delays which all 'feedback' is subject to. As mentioned in an earlier post of mine, this is more due to the 'loose-coupling' between motor and platter, which results in 'playing catch-up' continually, and not as a result of the associated electronics, which do react very quickly, although not instantaneously.
Actually, I was glad to have had the opportunity of praising your efforts, as without Mark's earlier reference to yourself here I would probably not have seen the other thread which showed how superb your TT looks.
I know how much work must have gone into a creation like that, and having been involved in similar activities for so many years, myself, I know it is unusual for mechanical engineers to also be aesthetically so astute. There are quite a few good industrial designers around, and some good machinests too, but to have achieved what you have done entirely on your own, would take an awful lot of beating!
You might find one or two of my comments to be of use to you at some point in the future, and don't forget that it is what happens at the *TT*, itself, rather than merely at the motor, which will provide the best sound. The two are inextricably linked, of course, but some people tend to overlook these important facts, and can get hung up on one particular parameter, sometimes at the expense of another.
Because of the relatively elastic nature of (most) drive belts, short-term 'abberations' can/will occur whether we like it or not with this means of transmission, and I therefore stand by my often-expressed views that increasing a TT's effective rotational mass, can only be a good thing, and my long-term listening tests have always confirmed this.
Regards,
Hi Vinyl-Addict,
There's no need for any apology, and actually you didn't really mislead me to any extent, anyway. I guessed you were rather 'out of your depth' here, which is quite understandable, and I realised there was no deliberate untoward intent on your part.
However, it would have been presumptuous and ill-mannered for me to do what I have seen all too often on the Forum, and simply say you are just talking out of your ***, or whatever, and that is not my way.
The background leading up to your involvement was admittedly unfortunate, but having been categorically told that I (and some others) couldn't be measuring our *TTs* properly, and your *TT* was given as the *only* example of measuring Mark's own results, I suppose it was inevitable. It is perhaps slightly ironic, that this reference has turned out to be somewhat unsubstantiated for now.
I am glad you are happy with the performance of Mark's PS, and remember that at no time have I attempted to cast any doubts on its effectiveness, nor anything else. Indeed I would be foolish to do so, as I have not 'heard' it in use.
Unfortunately, the same cannot truthfully be said for the converse, but sometimes that is the way of things in life.
I have merely observed that this might not be the only way of dealing with the issues, and I have not had good experiences with circuits which vary their outputs according to the load, due to the unavoidable delays which all 'feedback' is subject to. As mentioned in an earlier post of mine, this is more due to the 'loose-coupling' between motor and platter, which results in 'playing catch-up' continually, and not as a result of the associated electronics, which do react very quickly, although not instantaneously.
Actually, I was glad to have had the opportunity of praising your efforts, as without Mark's earlier reference to yourself here I would probably not have seen the other thread which showed how superb your TT looks.
I know how much work must have gone into a creation like that, and having been involved in similar activities for so many years, myself, I know it is unusual for mechanical engineers to also be aesthetically so astute. There are quite a few good industrial designers around, and some good machinests too, but to have achieved what you have done entirely on your own, would take an awful lot of beating!
You might find one or two of my comments to be of use to you at some point in the future, and don't forget that it is what happens at the *TT*, itself, rather than merely at the motor, which will provide the best sound. The two are inextricably linked, of course, but some people tend to overlook these important facts, and can get hung up on one particular parameter, sometimes at the expense of another.
Because of the relatively elastic nature of (most) drive belts, short-term 'abberations' can/will occur whether we like it or not with this means of transmission, and I therefore stand by my often-expressed views that increasing a TT's effective rotational mass, can only be a good thing, and my long-term listening tests have always confirmed this.
Regards,
Re: No harm done
I have machined a new pulley to incorporate a flat belt. I had to fabricate a special cutting tool to center the belt on the pulley.
It works perfectly and I hope to do some listening to see if there is any sonic improvement over the round 2mm belt I currently use.
From an engineering standpoint, what are the differences between the 2mm roundbelt and 6mm wide by 0.5mm thick flatbelt?
Hi Bob, You have been nothing but a gentleman in responding to my posts. I feel rather inept so to speak in offering the necessary data to answer your questions. As you have often said, it's what happens at the turntable itself that matters. I am very happy with the results I am hearing and I am really the only one that "matters",LOL.Bobken said:Hi Vinyl-Addict,
There's no need for any apology, and actually you didn't really mislead me to any extent, anyway. I guessed you were rather 'out of your depth' here, which is quite understandable, and I realised there was no deliberate untoward intent on your part.
Regards,
I have machined a new pulley to incorporate a flat belt. I had to fabricate a special cutting tool to center the belt on the pulley.
It works perfectly and I hope to do some listening to see if there is any sonic improvement over the round 2mm belt I currently use.
From an engineering standpoint, what are the differences between the 2mm roundbelt and 6mm wide by 0.5mm thick flatbelt?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.