Stefanoo,
This thread has gone into different directions for design, I am glad to entertain the opinions expressed, and have learned quite a bit so far.
I am not a ME, I do have considerable experience in different disciplines of manufacturing, none of which are dedicated to quiet or slow ( I realize both are relative terms).
I would like to help, and to do so I think it is best to break this project into smaller chunks for consideration, personally I have been considering the flywheel/drive wheel and here are a few thoughts. Your proposal for a flat belt around the wheel, I had envisioned multiple O-rings slightly stretched in grooves. We could experiment with different materials, O-rings are produced in a mold, and have flash around the O.D. this could be dealt with several ways, or not at all, I do not think this would be a major concern. As for manufacturing tolerances of the molded O-ring, this is why I suggest it be slightly stretched on our flywheel, it will then conform to the grooves and have a consistent O.D. Grinding the O.D. of this configuration, or a flat belt could be accomplished, but generally the softer materials would most likely have to be frozen, in order to be machined with any accuracy. On the platter of my TT there are several grooves and having matching grooves on the flywheel would ensure a positive engagement. Different materials for O-rings are readily available, and relatively inexpensive, a more pliable material would probably compensate further for a any manufacturing defects in the ring itself.
Engineering has been said to be the art of intelligent compromise.
For me this is a DIY project, for a hobby, not an all out cost no object commercial endeavor.
That being said, I have been shopping around the shop for what is currently on hand to avoid purchasing as much as possible. I was considering the flywheel to be made of aluminum, for ease of manufacture, and availability to me, but seeing the weight estimations I have now decided steel would be the best. I have some solid carbide shafts 3/8” in diameter approximately 4” long, I also have some Teflon material to produce bushings, which we could compare to the Igus products already discussed. I have not decided on a preference for supporting the weight of this flywheel, there was discussion of a sphere, I have some ceramic bearings, I will see how well they work with a polished carbide surface. I have access to balancing equipment as well, balance for a rotating mass is usually referenced to with a service speed.
Tolerance = Component Weight [kg] x Quality Grade x 9549 / Operational Speed [RPM] (taken from ISQ Quality Grades | G40 G16 6,3 G2,5 G1 G0,4 | Rotor Types Examples)
With equipment available to me we could achieve G0.1 for separate component
The idea for removing pressure from our device when not in use, to prevent flat spots forming, is great. Before the problem was introduced, I was thinking of a thumbscrew arrangement on the side to allow adjustment of pressure, I will submit a sketch or drawing of this idea shortly. Another possible solution for this would be a lever style clamp, also with adjustment for proper pressure, DeStaco produces many inexpensive devices. To take this lever one step further a small solenoid could be added to remove the pressure at power off.
Thanks all, looking for feedback from all of you, especially on supporting the weight of the flywheel, I would like to avoid a shoulder bushing or thrust washer configuration if possible.
Ted
This thread has gone into different directions for design, I am glad to entertain the opinions expressed, and have learned quite a bit so far.
I am not a ME, I do have considerable experience in different disciplines of manufacturing, none of which are dedicated to quiet or slow ( I realize both are relative terms).
I would like to help, and to do so I think it is best to break this project into smaller chunks for consideration, personally I have been considering the flywheel/drive wheel and here are a few thoughts. Your proposal for a flat belt around the wheel, I had envisioned multiple O-rings slightly stretched in grooves. We could experiment with different materials, O-rings are produced in a mold, and have flash around the O.D. this could be dealt with several ways, or not at all, I do not think this would be a major concern. As for manufacturing tolerances of the molded O-ring, this is why I suggest it be slightly stretched on our flywheel, it will then conform to the grooves and have a consistent O.D. Grinding the O.D. of this configuration, or a flat belt could be accomplished, but generally the softer materials would most likely have to be frozen, in order to be machined with any accuracy. On the platter of my TT there are several grooves and having matching grooves on the flywheel would ensure a positive engagement. Different materials for O-rings are readily available, and relatively inexpensive, a more pliable material would probably compensate further for a any manufacturing defects in the ring itself.
Engineering has been said to be the art of intelligent compromise.
For me this is a DIY project, for a hobby, not an all out cost no object commercial endeavor.
That being said, I have been shopping around the shop for what is currently on hand to avoid purchasing as much as possible. I was considering the flywheel to be made of aluminum, for ease of manufacture, and availability to me, but seeing the weight estimations I have now decided steel would be the best. I have some solid carbide shafts 3/8” in diameter approximately 4” long, I also have some Teflon material to produce bushings, which we could compare to the Igus products already discussed. I have not decided on a preference for supporting the weight of this flywheel, there was discussion of a sphere, I have some ceramic bearings, I will see how well they work with a polished carbide surface. I have access to balancing equipment as well, balance for a rotating mass is usually referenced to with a service speed.
Tolerance = Component Weight [kg] x Quality Grade x 9549 / Operational Speed [RPM] (taken from ISQ Quality Grades | G40 G16 6,3 G2,5 G1 G0,4 | Rotor Types Examples)
With equipment available to me we could achieve G0.1 for separate component
The idea for removing pressure from our device when not in use, to prevent flat spots forming, is great. Before the problem was introduced, I was thinking of a thumbscrew arrangement on the side to allow adjustment of pressure, I will submit a sketch or drawing of this idea shortly. Another possible solution for this would be a lever style clamp, also with adjustment for proper pressure, DeStaco produces many inexpensive devices. To take this lever one step further a small solenoid could be added to remove the pressure at power off.
Thanks all, looking for feedback from all of you, especially on supporting the weight of the flywheel, I would like to avoid a shoulder bushing or thrust washer configuration if possible.
Ted
Ted your post is magnificent I really appreciate you are putting your knowledge into this thread. Yes the mechanical design should be broken in parts you and Jaimie are working officially on this. I wouldn’t suggest to have more than 2 people working on the mechanical because then it gets too hard to manage tasks and who is doing what and more heads means more ideas of doing the same thing, which if from one side it is a positive thing it can be counteracting on the other.
Balancing is I guess important so you probably have better equipment to create flywheel, if that is true, why don’t you take care of the flywheel and pulley and Jamie takes care of the enclosure for the pod and for the board for the controller?
How’s that sound?
Regarding o-ring, I am with you, it is easier to obtain o-rings with a different material sizes etc.
I still know from other’s experience that o-ring needs to be machined to have max performance, but this is a detailed that we can deal with later on.
Regarding multiple O-ring I think this is a beautiful idea as it distributes pressure a little bit. There is only a problem with that though. Having multiple rings might not work for everybody. To me it doesn’t matter my platter is flat, but for others who have grooves as well, may have different pitch.
Can you sketch your idea for setting the pressure? I was envisioning with simple flanged held by 2-3 screws that can be loosen up and slide the pulley left or right to adjust and tighten the screw. This way there wouldn’t be any mechanical mechanism hanging around which I would hate to see dangling there, springs or anything like that would in my opinion be a no go.
Balancing is I guess important so you probably have better equipment to create flywheel, if that is true, why don’t you take care of the flywheel and pulley and Jamie takes care of the enclosure for the pod and for the board for the controller?
How’s that sound?
Regarding o-ring, I am with you, it is easier to obtain o-rings with a different material sizes etc.
I still know from other’s experience that o-ring needs to be machined to have max performance, but this is a detailed that we can deal with later on.
Regarding multiple O-ring I think this is a beautiful idea as it distributes pressure a little bit. There is only a problem with that though. Having multiple rings might not work for everybody. To me it doesn’t matter my platter is flat, but for others who have grooves as well, may have different pitch.
Can you sketch your idea for setting the pressure? I was envisioning with simple flanged held by 2-3 screws that can be loosen up and slide the pulley left or right to adjust and tighten the screw. This way there wouldn’t be any mechanical mechanism hanging around which I would hate to see dangling there, springs or anything like that would in my opinion be a no go.
Glad you noticed the atypical approach in the Brinkmann Bardo. It only shows that slow start up and high mass platter are not in the private domain of belt drive system. And there is no identity crisis at all. Don't forget that many famous direct drive turntables originally started out as radio station or DJ equipment that a quick start up is a necessity but not because it's direct drive. In earlier days, many idler drive tables have very quick start up too. After all Brinkmann is known for their belt drive turntables before they launched the Bardo and Oasis. In their white paper, they clearly stated that quick start up in home stereo is not necessary so they implemented many features in their BD tables on the Bardo such as the same robust bearing and high mass platter. Kenwood did similar things in their L-07D table using coreless motor and high mass platter and not so quick start up. In Brinkmann's recent development that they admit they learned something from the Bardo and created the Sinus coreless motor to belt drive their flagship product. (Who's having identity crisis now?!) It only goes to prove each drive system can cross implement other, hence no such thing as identity crisis. What comes down to is that the stylus does NOT care what is making the platter spin. It can only "feel" the torque, compliance of the interface (be it belt, idler wheel, or magnetic force), and the amount of cogging. If all three things are matched, given the same platter, bearing, and plinth, they should sound similar or identical, regardless of drive system. Once you have a quality bearing and platter, the issue is to find a way to make it spin accurately (33.33rpm) and smoothly (no cogging) and strongly (no stylus drag), any one of the three drive systems is capable of that so it comes down to execution. I use all three and they gave me all different flavors. Just like food, I eat everything.
The Bardo use a very "soft" magnetic force, much like a low torque motor in BD a la Nottingham, to keep the platter in speed while letting the high mass flywheel effect and inertia to take care of the smooth rotation and to lessen cogging. Don't forget that many top of the line DD tables in the late 70's and early 80's never went further development because of the advent of the compact disk. All the big corporations simply stop developing turntables and analog manufacturing went sort of underground or into a cottage industry. It is much easier for a garage operation to make a belt-drive turntable than a complicated direct drive. The Bardo, Teres Certus, Monaco Grand Prix, NVS, Rockport Sirius 3, VPI Classic Direct,etc... just further this development and since they are not mass produced their price went to the stratosphere.
I never owned a JVC QL-10, which is a TT-101 as a motor unit, but I have owned many JVC DD models and many of those employ a coreless motor always sounded smooth and musical to me. I suspect your unit might need a check up or their electrolytic caps replaced. One audiophile in Australia uses that as his reference and he prefers it over his Raven TW Acoustic belt drive table. I had a Dual 701 that I took the motor out and placed it in a DIY plinth using a different tonearm and it sounded really nice. One day I noticed the sound was bright, sterile and clinical and missing the usual "bloom" or harmonic richness that I enjoyed. I went to check on the strobe and noticed some jittery motion, cogging, and it was the speed adjustment pot's wiper had intermittent contact issue. I replaced the pot and the nervous sound was gone and the smooth tone is back. Your TT-101 is more than 30 years old so it's time for a check up.
That, I totally agree.
AVWERK had his negative experience with the TT-101 and here I merely point out the positive story from another user.
Last edited:
One of the Goldmunds used JVC motors I believe. The cheapest studioette and the one above it (2 U shaped black acrylic sections placed on top of each other) we're quickly rejected against the VPI and Maplenoll Adriana. Absolute sound mag comparison
That certainly wasn't the case with their top 2 models however.
There is a considerable amount of electronics and a cap change might do the QL 10 some good, but that cheap platter has to go. I can machine a new heavy platter, but uncertain if the speed correction might over react since it wasn't tuned for a heavy platter.
I,ll stop now as this is Stefanoo thread and it's his momentum and not ours to divert.
Regards
David
That certainly wasn't the case with their top 2 models however.
There is a considerable amount of electronics and a cap change might do the QL 10 some good, but that cheap platter has to go. I can machine a new heavy platter, but uncertain if the speed correction might over react since it wasn't tuned for a heavy platter.
I,ll stop now as this is Stefanoo thread and it's his momentum and not ours to divert.
Regards
David
David...no problem I enjoy your conversation as well...I think it is very constructive! You guys know a lot of things...if you could also pour some of your knowledge on this "design" I would really appreciate it!!
I missed this a few pages back, but I'm a strong proponent of it so I'll reply.
People load turntable platters up with drag so that the motor is constantly running into a high fixed load. If you have a platter/bearing/lubrication system that applies several hundred grams of very linear drag, then any change in stylus drag due to signal variation on the record becomes utterly inconsequential.
Given that most DC motors are quite load invariant, it's a good thing to consider, then once implemented you can forget about stylus drag speed variations.
People load turntable platters up with drag so that the motor is constantly running into a high fixed load. If you have a platter/bearing/lubrication system that applies several hundred grams of very linear drag, then any change in stylus drag due to signal variation on the record becomes utterly inconsequential.
Given that most DC motors are quite load invariant, it's a good thing to consider, then once implemented you can forget about stylus drag speed variations.
Stefano,
I just did a few quick calculations regarding speeds and pulley sizes.
If we assume that the Flywheel Pulley is 100 mm in dia.
Assume that the motor drives the Flywheel at 100 mm, dia.
Assume the motor pulley is 15 mm dia.
The motor would need to run at 666 rpm.
The Maxon motor you've specified shows 4180 as the nominal speed.
To get the motor to run near it's nominal speed we would it to drive a Flywheel Pulley that is 627 mm. dia. That's 2X the platter dia.
I believe you said the motor would run somewhere around 400 to 500 rpm. Is that correct? Are you going to drive the Maxon motor slow? That would be at about 10% of the Maxon motors speed rating.
Maybe I am not understanding the idea very well. It seems odd to have a fast motor in this configuration. This is why I wanted to see the motor specification before going further in design.
Everyone please check my calculations. I tried to attach my excel file, but the attachment utility does not like the file so I just supplied the pdf image.
bfg4wd, I would be happy for you to check my work. Also, it is no problem for us to race against each other for the mechanical parts. It is hard to know how long it might take to build something when asking friends to help in the machine shops, etc. So, if you have solutions, parts, etc. before I do, I will not feel left out. I am sure the community would be happy to see some options.
Jamie
I just did a few quick calculations regarding speeds and pulley sizes.
If we assume that the Flywheel Pulley is 100 mm in dia.
Assume that the motor drives the Flywheel at 100 mm, dia.
Assume the motor pulley is 15 mm dia.
The motor would need to run at 666 rpm.
The Maxon motor you've specified shows 4180 as the nominal speed.
To get the motor to run near it's nominal speed we would it to drive a Flywheel Pulley that is 627 mm. dia. That's 2X the platter dia.
I believe you said the motor would run somewhere around 400 to 500 rpm. Is that correct? Are you going to drive the Maxon motor slow? That would be at about 10% of the Maxon motors speed rating.
Maybe I am not understanding the idea very well. It seems odd to have a fast motor in this configuration. This is why I wanted to see the motor specification before going further in design.
Everyone please check my calculations. I tried to attach my excel file, but the attachment utility does not like the file so I just supplied the pdf image.
bfg4wd, I would be happy for you to check my work. Also, it is no problem for us to race against each other for the mechanical parts. It is hard to know how long it might take to build something when asking friends to help in the machine shops, etc. So, if you have solutions, parts, etc. before I do, I will not feel left out. I am sure the community would be happy to see some options.
Jamie
Let me reply to you in a little bit...I am not at home now!
Pyramid and I have already establish set points for motor and pulleys ratio.
I will look through your post carefully and I will try to sketch dimension for everything. Maybe this coming week we can start the mechanical design.
I will probably start bread boarding the analogue control for the motor do that pyramid and I will have a base line.
directdriver,
I am really enjoying your information on the DD's. I have been fascinated by the Technics 1200 for the last few months learning about the design. It is elegant in many ways.
I think that with today's faster processors and advanced digital signal processing there might be a way to make these old DD motors really sing. When you can completely control the signal going to a motor, you can do things not possible before high speed digital processing. What Pyramid can do with a motor controller would have been impossible when these TT's were made. The amount of calculations per second that can be done now cheaply is like a supercomputer for the 70's or early 80's.
Anyway, fun to think about.
I am really enjoying your information on the DD's. I have been fascinated by the Technics 1200 for the last few months learning about the design. It is elegant in many ways.
I think that with today's faster processors and advanced digital signal processing there might be a way to make these old DD motors really sing. When you can completely control the signal going to a motor, you can do things not possible before high speed digital processing. What Pyramid can do with a motor controller would have been impossible when these TT's were made. The amount of calculations per second that can be done now cheaply is like a supercomputer for the 70's or early 80's.
Anyway, fun to think about.
The power calculation to control an AC motor is relatively simple and doesn't require a lot of calculation power.
An equally good motor control but more bulky and obviously not as flexible could have been done back then with discrete, by having a low distortion sine phase and frequency controlled and driven by low distortion amp.
A lot has been done not only on the control side in these 20-30 years but also on motors and TTs and it is the combination of all these things that lets the magic happen..
An equally good motor control but more bulky and obviously not as flexible could have been done back then with discrete, by having a low distortion sine phase and frequency controlled and driven by low distortion amp.
A lot has been done not only on the control side in these 20-30 years but also on motors and TTs and it is the combination of all these things that lets the magic happen..
All of the accuracy and resolution in my AC supply comes from the fact that I'm using DDS to generate the AC waveform instead of PWM as Music Hall and Pro-Ject have. DDS has been around since the late 80's but back then, a chip with the same capabilities of the one I'm using would have been ~$80 and housed in a 68 pin package the size of a silver dollar. The AD9833 is $9 and in a µMax10 package, it fits on the eraser of a #2 pencil. Not sure why anyone wouldn't use DDS for an AC supply?
The other thing that sets my supply apart (besides some nice software features) is the power supply and output stage, the latter which doesn't use any xfmrs or coupling caps which is why I can fit it in an extrusion the size of a pack of cigarettes. I believe my supply is the only one capable of connection to a tach as well.
For those of you who expressed an interest when it was available, the PCBs went to the assembly house and I hope to have them back by month's end. I received the metal work for the front panels last week, which I think look much better than the acrylic face plates on the prototypes.
The other thing that sets my supply apart (besides some nice software features) is the power supply and output stage, the latter which doesn't use any xfmrs or coupling caps which is why I can fit it in an extrusion the size of a pack of cigarettes. I believe my supply is the only one capable of connection to a tach as well.
For those of you who expressed an interest when it was available, the PCBs went to the assembly house and I hope to have them back by month's end. I received the metal work for the front panels last week, which I think look much better than the acrylic face plates on the prototypes.
Attachments
I could see where this would work on an AC motor as it might reduce cogging on an unloaded motor, but DC motors are not load invariant. The Maxon motor Stefano is looking at has a significant Speed/Torque gradient that is typical of these types of motors. The motor is rated at 5200 RPM NL and a max torque of 30mNm. With a negative torque slope of 34 RPM/mNm, the full load speed would drop almost 20% to 4200 RPM.
We are looking at running the motor between 700 and 945 RPM, so the 34 RPM/mNm spec will have an impact on speed stability. That is why we are including the circuitry for compensation via current monitoring. It will be jumper selectable for those who don't want to use it.
Still, I could see where viscous drag might smooth out W&F, but that's not my area of expertise.
Stefano,
I think you have oversimplified the problem and potential solutions for torque ripple or cogging torque. It's not just about calculating raw or average power, it is about potentially controlling in real time torque ripple or cogging torque. Each of the motor types, DC, AC, DC Brushless have some amount of cogging. With a DC motor the problem and solution must be mechanical since the cogging occurs with the brushes reaching zero torque momentarily. Not sure how much the controller would help. But, in AC motors of either synchronous or polyphase (DC Brushless or Induction), there is opportunity for phase shaping to reduce cogging.
Please see the article below for some recent research on using digital technologies for feed forward and feed back ripple compensation. In this study the controller updated arithmetically calculated values once per every milli-second.
http://www.nt.ntnu.no/users/skoge/prost/proceedings/cdc-ecc-2011/data/papers/0428.pdf
I'd like to hear how you could do this discretely? I would also like to hear how you could do this with technology from 30 years ago?
If speed control is so simple and you have all the answers, why are we bothering?
I think you have oversimplified the problem and potential solutions for torque ripple or cogging torque. It's not just about calculating raw or average power, it is about potentially controlling in real time torque ripple or cogging torque. Each of the motor types, DC, AC, DC Brushless have some amount of cogging. With a DC motor the problem and solution must be mechanical since the cogging occurs with the brushes reaching zero torque momentarily. Not sure how much the controller would help. But, in AC motors of either synchronous or polyphase (DC Brushless or Induction), there is opportunity for phase shaping to reduce cogging.
Please see the article below for some recent research on using digital technologies for feed forward and feed back ripple compensation. In this study the controller updated arithmetically calculated values once per every milli-second.
http://www.nt.ntnu.no/users/skoge/prost/proceedings/cdc-ecc-2011/data/papers/0428.pdf
I'd like to hear how you could do this discretely? I would also like to hear how you could do this with technology from 30 years ago?
If speed control is so simple and you have all the answers, why are we bothering?
The idler system IS the mechanical drag that offsets needle drag as long as you have tight coupling. A loose belt here would defeat its purpose.
The high speed motor mass, idler mass and platter weight constitute a mechanical leverage advantage that helps bury outside resistance, in this case the needle when it hits a big bass transient or otherwise.
It was designed when cartridges had 3 times the downforce that they have now.
The idler is akin to your car approaching a hill and you downshift to compensate. As you go up the hill, it doesn't,t feel like a load since you have a mechanical advantage by shifting to a lower gear and higher engine speeds
Done right and everything else just so, the idler looks like it has real merit.
Getting everything to work in unison quietly is the mountain to climb.
Regards
David
The high speed motor mass, idler mass and platter weight constitute a mechanical leverage advantage that helps bury outside resistance, in this case the needle when it hits a big bass transient or otherwise.
It was designed when cartridges had 3 times the downforce that they have now.
The idler is akin to your car approaching a hill and you downshift to compensate. As you go up the hill, it doesn't,t feel like a load since you have a mechanical advantage by shifting to a lower gear and higher engine speeds
Done right and everything else just so, the idler looks like it has real merit.
Getting everything to work in unison quietly is the mountain to climb.
Regards
David
Last edited:
David thanks for your post. Yes the idler was initially thought for the reason you mentioned.
I know the mechanical side won't be easy to get right.
Pyramid and I are trying to get the controller done right as much as possible but how to design the mechanical goes beyond my set of skill.
I have given the initial design concept and I rely on the mechanical guys to get it done. I will try to provide all measurements tonight!
I think we need somebody on here that helps to give ideas on material choice and type of process required to get things done right.
What I am not seeing is a discussion on how to implement idler bearing materials for sleeve and so so forth! I think these are the type of discussions we need right now, am I wrong?
I know the mechanical side won't be easy to get right.
Pyramid and I are trying to get the controller done right as much as possible but how to design the mechanical goes beyond my set of skill.
I have given the initial design concept and I rely on the mechanical guys to get it done. I will try to provide all measurements tonight!
I think we need somebody on here that helps to give ideas on material choice and type of process required to get things done right.
What I am not seeing is a discussion on how to implement idler bearing materials for sleeve and so so forth! I think these are the type of discussions we need right now, am I wrong?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.