Re: Re: No harm done
Hi Vinyl-Addict,
** I feel rather inept so to speak in offering the necessary data to answer your questions.**
No problem here as you did make it clear that your measurements (so far) relating to any speed-variations (long-term, or short-term drift) which had been quoted by Mark in support of his contention, were not at the *TT*, which was the crux of the matter. I appreciate that you did carry out a somewhat restricted 'dynamic' test with the KAB strobe, which is better than nothing, but this is a limitation of this method of assessment, and not in any way due to any ineptitude on your part.
**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.**
Is this a crowned pulley, here, perhaps?
**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.**
It will be interesting to hear what your findings are when you carry out this trial, and if you have a system with good overall resolution, I would be very surprised if you don't notice a difference. One trial I have not done (so far!) is to directly compare these two styles of belt, with everything else being equal, simply because I have not found the need to do this.
When I carried out some listening trials with flat belts, the belt material was quite different (cassette tape), and this is not a satisfactory way of reaching positive conclusions as the material of the belt will likely have more effect than the section used.
Maybe one day (if I live long enough!) I will do some direct comparisons, as I do expect to hear differences, and I would like to satisfy myself, here. Predicting the outcome with any accuracy, however, is beyond my current abilities and I know of no established 'authority' to refer to here.
**From an engineering standpoint, what are the differences between the 2mm roundbelt and 6mm wide by 0.5mm thick flatbelt?
**
This is not so easy (at least for me, not being omniscient!) as there are so many factors which will be relevant here. The cross-sectional area would be much the same with your choices, so I wouldn't expect much change in the overall 'stretch' of the two belts, but I would anticipate changes in the belt's flexibility, which is likely to have some effect as it passes around a relatively small dia. pulley.
What actually occurs whilst bending rectangular section materials I have found to be counter-intuitive, in that instead of the cross-sectional centre-line of the material remaining 'static' in length, I have (generally) found that the 'static' length (i.e. the part of the material which remains the same in overall length) is at around 1/3 of the radius of the bend out from the bend's axis. I am subject to some difficulty in explaining this, and I have never seen any other reference to this phenomenon, elsewhere.
However, if you consider the cross-section with a tight bend, what happens is that the 'inner' 1/3 compresses, and the 'outer' 2/3 will stretch, and not 50:50, which might be what is expected here. Accordingly, as a belt passes around a small pulley (with a tight bend of maybe 90 deg. or so) the belt's *inner surface* will shorten by about 1/3, and the *outer surface* will stretch by about 2/3.
The same principle will apply at the platter, of course, but here there is very minimal bending occurring with most 'normal' TT set-ups.
Remember that the belt is being 'driven' by contact with the pulley merely on its *inner surface*, and as the belt is passing around the motor's pulley this contact surface is continually being compressed on the lead-in side, and then stretched out again as it leaves the pulley.
Bearing in mind my remarks in a much earlier post about pulley eccentricities of merely a thou. or so being clearly detrimental to the sound, I am fairly certain that this belt stretching here will have *some* effect on the resultant sound, and I would *guess* that minimising any such effect will be beneficial overall.
With your proposed flat belt, there is a much reduced potential here, as the cross-sectional width is only 0.5 mm, compared with 2mm for the circular-section belt, which is a 4x difference.
Additionally, with the flat belt you will be applying all forces (both to, and from the platter) over a much larger 'contact area', with say a differential quite likely in the region of 10x, or even more. Suppose the part of the pulley's circumference on which the belt bears is 10mm., then you would have 60 sq.mm. surface contact with the 6mm. belt, but with only possibly 1/2mm. width of the belt 'touching' with a 2mm. circular-section, this would give you only 5sq.mm of contact area. Once again we are dealing with orders of magnitude in differences here, but to further complicate matters this will also depend on the pulley's section, too.
Here I have used a very shallow concave profile (merely enough to prevent the belt from detaching itself) so the circular-section belt's contact patch with the pulley will be minimal, but this 'sounds' different subjectively from a pulley which had a profile which contacted approx. half of the belt's cross-sectional circumference, which of course is the maximum possible here.
The latter profile was not so good sonically, though, I believe because of the 'conflicting' circumferences with which the belt was in contact on this pulley, together with the unexpected way the section of the belt stretches/shrinks as it traverses the pulley.
Also, different tensions used with exactly the same belt/pulley/platter etc., can and do affect the sound, as my trials have shown.
Then you have the different frictions and consequent heat being generated due to varying belt flexures, which will all be subject to *some* variation with alternative choices made by yourself in this area.
The surface finish of pulleys (and platters) will have some effect, as there is not a perfect interface between these parts and the (also possibly different-shaped!) belts, and see what happened in my earlier post when I burnished a motor pulley in situ on a motor.
Whether any or all of the above will affect the sonic results, and to what extent (if at all) I can only speculate, but I am not in the habit of speculation generally. I do know that this area is surprisingly 'sensitive', and to a degree which maybe 20 years ago, I would have found very hard to come to terms with.
When you start to investigate such matters, it is bit of a 'minefield', actually, which is why I am certain Mark has a few surprises in store for him if he thinks getting a motor's speed constant by measuring this, is the secret to accurately-performing and consequently good-sounding *TTs*.
What I think Mark has overlooked so far, is that the motor only 'knows' what is going on in the 'real-world' (i.e. at the platter) solely by way of this rather 'wayward' belt, which we are stuck with in this method of driving a TT. So, no matter what any motor specs. may suggest, nor any theories here might predict, nor how excellent any 'correction circuit' might be, it will never know what it is supposed to do, anyway, until it is rather too late!
When any load at the TT changes (for whatever reason) there will be an inevitable delay before this is transmitted by this 'compliant' belt to the motor, and this I believe was the reason I eventually needed to abandon the Burr Brown 'servo' circuit which someone had asked about, and which caused my initial post in this thread.
With this circuit, I initially though it was the (an!) answer, but there was something about the sound which was not quite right to my ears. However, whilst simply measuring voltages and currents at the motor didn't really show anything seriously amiss, perhaps this was not surprising in view of the intentional 'balancing' of back EMF motor effects, which the circuit was designed to achieve. It was only on using the Gyrascope, whereafter I could then 'see' what was going on in real-time, that I found I could influence these obvious 'hunting' effects by some minor changes to the circuit components' values (in accordance with the designers suggestions) but no matter what I tried, I could not eliminate them entirely. In fact, I wasted several months of 'development time' on this, and then gave up.
At that time, I had achieved a situation where I could not see nor hear any noticeable beginning-to-end-of-side speed variations, by entirely other (mostly motor choices and more basic mechanical-engineering) means.
It was, therefore, a welcome surprise (as already admitted by me!) when I discovered that my latest very robust PS which I reverted to and had improved after the 'BB' circuit fiasco, that this also handily overcame (to all practical purposes) any *short-term variations*, which had become my only remaining problem by that time.
I am aware of an extremely sophisticated computer-controlled PS in use with TTs, but cannot discuss this here. What is significant about it, and where it departs from the norm, is that it is controlled via feedback directly from the *platter*, itself, which is an entirely different ball-game, of course. In my experience, this is the only way in which any 'feedback' or servo circuit can be fully effective, and with few downsides, but this is not easy to achieve in practice, certainly for any DIYer, and is outside the scope of this discussion.
If you use a flat belt, try putting a single (180 deg.) twist in it when you cement the two ends together, which will effectively provide a 'single-sided' belt. I cannot vouch for this, personally, but more than one knowledgeable person has commented that this improves the sound, and presumably you can also try it with the 'twist' in both the leading or trailing portions of the belt, for further experimentation.
I hope some of the above will give you food for thought.
Regards,
Hi Vinyl-Addict,
** I feel rather inept so to speak in offering the necessary data to answer your questions.**
No problem here as you did make it clear that your measurements (so far) relating to any speed-variations (long-term, or short-term drift) which had been quoted by Mark in support of his contention, were not at the *TT*, which was the crux of the matter. I appreciate that you did carry out a somewhat restricted 'dynamic' test with the KAB strobe, which is better than nothing, but this is a limitation of this method of assessment, and not in any way due to any ineptitude on your part.
**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.**
Is this a crowned pulley, here, perhaps?
**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.**
It will be interesting to hear what your findings are when you carry out this trial, and if you have a system with good overall resolution, I would be very surprised if you don't notice a difference. One trial I have not done (so far!) is to directly compare these two styles of belt, with everything else being equal, simply because I have not found the need to do this.
When I carried out some listening trials with flat belts, the belt material was quite different (cassette tape), and this is not a satisfactory way of reaching positive conclusions as the material of the belt will likely have more effect than the section used.
Maybe one day (if I live long enough!) I will do some direct comparisons, as I do expect to hear differences, and I would like to satisfy myself, here. Predicting the outcome with any accuracy, however, is beyond my current abilities and I know of no established 'authority' to refer to here.
**From an engineering standpoint, what are the differences between the 2mm roundbelt and 6mm wide by 0.5mm thick flatbelt?
**This is not so easy (at least for me, not being omniscient!) as there are so many factors which will be relevant here. The cross-sectional area would be much the same with your choices, so I wouldn't expect much change in the overall 'stretch' of the two belts, but I would anticipate changes in the belt's flexibility, which is likely to have some effect as it passes around a relatively small dia. pulley.
What actually occurs whilst bending rectangular section materials I have found to be counter-intuitive, in that instead of the cross-sectional centre-line of the material remaining 'static' in length, I have (generally) found that the 'static' length (i.e. the part of the material which remains the same in overall length) is at around 1/3 of the radius of the bend out from the bend's axis. I am subject to some difficulty in explaining this, and I have never seen any other reference to this phenomenon, elsewhere.
However, if you consider the cross-section with a tight bend, what happens is that the 'inner' 1/3 compresses, and the 'outer' 2/3 will stretch, and not 50:50, which might be what is expected here. Accordingly, as a belt passes around a small pulley (with a tight bend of maybe 90 deg. or so) the belt's *inner surface* will shorten by about 1/3, and the *outer surface* will stretch by about 2/3.
The same principle will apply at the platter, of course, but here there is very minimal bending occurring with most 'normal' TT set-ups.
Remember that the belt is being 'driven' by contact with the pulley merely on its *inner surface*, and as the belt is passing around the motor's pulley this contact surface is continually being compressed on the lead-in side, and then stretched out again as it leaves the pulley.
Bearing in mind my remarks in a much earlier post about pulley eccentricities of merely a thou. or so being clearly detrimental to the sound, I am fairly certain that this belt stretching here will have *some* effect on the resultant sound, and I would *guess* that minimising any such effect will be beneficial overall.
With your proposed flat belt, there is a much reduced potential here, as the cross-sectional width is only 0.5 mm, compared with 2mm for the circular-section belt, which is a 4x difference.
Additionally, with the flat belt you will be applying all forces (both to, and from the platter) over a much larger 'contact area', with say a differential quite likely in the region of 10x, or even more. Suppose the part of the pulley's circumference on which the belt bears is 10mm., then you would have 60 sq.mm. surface contact with the 6mm. belt, but with only possibly 1/2mm. width of the belt 'touching' with a 2mm. circular-section, this would give you only 5sq.mm of contact area. Once again we are dealing with orders of magnitude in differences here, but to further complicate matters this will also depend on the pulley's section, too.
Here I have used a very shallow concave profile (merely enough to prevent the belt from detaching itself) so the circular-section belt's contact patch with the pulley will be minimal, but this 'sounds' different subjectively from a pulley which had a profile which contacted approx. half of the belt's cross-sectional circumference, which of course is the maximum possible here.
The latter profile was not so good sonically, though, I believe because of the 'conflicting' circumferences with which the belt was in contact on this pulley, together with the unexpected way the section of the belt stretches/shrinks as it traverses the pulley.
Also, different tensions used with exactly the same belt/pulley/platter etc., can and do affect the sound, as my trials have shown.
Then you have the different frictions and consequent heat being generated due to varying belt flexures, which will all be subject to *some* variation with alternative choices made by yourself in this area.
The surface finish of pulleys (and platters) will have some effect, as there is not a perfect interface between these parts and the (also possibly different-shaped!) belts, and see what happened in my earlier post when I burnished a motor pulley in situ on a motor.
Whether any or all of the above will affect the sonic results, and to what extent (if at all) I can only speculate, but I am not in the habit of speculation generally. I do know that this area is surprisingly 'sensitive', and to a degree which maybe 20 years ago, I would have found very hard to come to terms with.
When you start to investigate such matters, it is bit of a 'minefield', actually, which is why I am certain Mark has a few surprises in store for him if he thinks getting a motor's speed constant by measuring this, is the secret to accurately-performing and consequently good-sounding *TTs*.
What I think Mark has overlooked so far, is that the motor only 'knows' what is going on in the 'real-world' (i.e. at the platter) solely by way of this rather 'wayward' belt, which we are stuck with in this method of driving a TT. So, no matter what any motor specs. may suggest, nor any theories here might predict, nor how excellent any 'correction circuit' might be, it will never know what it is supposed to do, anyway, until it is rather too late!
When any load at the TT changes (for whatever reason) there will be an inevitable delay before this is transmitted by this 'compliant' belt to the motor, and this I believe was the reason I eventually needed to abandon the Burr Brown 'servo' circuit which someone had asked about, and which caused my initial post in this thread.
With this circuit, I initially though it was the (an!) answer, but there was something about the sound which was not quite right to my ears. However, whilst simply measuring voltages and currents at the motor didn't really show anything seriously amiss, perhaps this was not surprising in view of the intentional 'balancing' of back EMF motor effects, which the circuit was designed to achieve. It was only on using the Gyrascope, whereafter I could then 'see' what was going on in real-time, that I found I could influence these obvious 'hunting' effects by some minor changes to the circuit components' values (in accordance with the designers suggestions) but no matter what I tried, I could not eliminate them entirely. In fact, I wasted several months of 'development time' on this, and then gave up.
At that time, I had achieved a situation where I could not see nor hear any noticeable beginning-to-end-of-side speed variations, by entirely other (mostly motor choices and more basic mechanical-engineering) means.
It was, therefore, a welcome surprise (as already admitted by me!) when I discovered that my latest very robust PS which I reverted to and had improved after the 'BB' circuit fiasco, that this also handily overcame (to all practical purposes) any *short-term variations*, which had become my only remaining problem by that time.
I am aware of an extremely sophisticated computer-controlled PS in use with TTs, but cannot discuss this here. What is significant about it, and where it departs from the norm, is that it is controlled via feedback directly from the *platter*, itself, which is an entirely different ball-game, of course. In my experience, this is the only way in which any 'feedback' or servo circuit can be fully effective, and with few downsides, but this is not easy to achieve in practice, certainly for any DIYer, and is outside the scope of this discussion.
If you use a flat belt, try putting a single (180 deg.) twist in it when you cement the two ends together, which will effectively provide a 'single-sided' belt. I cannot vouch for this, personally, but more than one knowledgeable person has commented that this improves the sound, and presumably you can also try it with the 'twist' in both the leading or trailing portions of the belt, for further experimentation.
I hope some of the above will give you food for thought.
Regards,
Hi folks,
I tried a 10mm pulley adapted from an old dual deck, its a radially split design to enable its diameter to be adjusted. This means its a bit wobbley and no good for serious use. However, with this pulley I was able to run the motor at over 3volts and low and behold the platter speed appears constant from outer to inner part of the LP! So it does seem that the speed/ torque performance was being compromised at such a low voltage (as Bob suggested). Also I think I've found someone to machine me a proper pulley, so things are looking up!
Cheers
Matt
I tried a 10mm pulley adapted from an old dual deck, its a radially split design to enable its diameter to be adjusted. This means its a bit wobbley and no good for serious use. However, with this pulley I was able to run the motor at over 3volts and low and behold the platter speed appears constant from outer to inner part of the LP! So it does seem that the speed/ torque performance was being compromised at such a low voltage (as Bob suggested). Also I think I've found someone to machine me a proper pulley, so things are looking up!
Cheers
Matt
Bingo!
Hi Matt,
I am very relieved for your sake to hear this.
However, as you might imagine, it is little surprise to me.
I am sorry I couldn't spare the time right now to make you a new bespoke pulley from scratch, myself, but if you are completely stuck, maybe at some future time I could help out here.
Regards,
Hi Matt,
I am very relieved for your sake to hear this.
However, as you might imagine, it is little surprise to me.
I am sorry I couldn't spare the time right now to make you a new bespoke pulley from scratch, myself, but if you are completely stuck, maybe at some future time I could help out here.
Regards,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.