|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.28.84.51
In Reply to: RE: Another DIY 3 phase variable frequency drive for Papst motor update. posted by coffee-phil on July 20, 2016 at 21:33:49
Thanks Phil, I think there is an error on both the phase generator and the tach. However, all things are relative, and since I can change the frequency 1/10hz at a time, I should be able to dial things in. 1/10hz should equal about 3 rpm on the motor.
I really don't care if I have to run the signal generator at 60.7 hz (or whatever) to actually get the correct speed. I do seem to get repeatable results when I power the motor and/or the generator on and off again, same with the tach. The plan is, once I get the TT built, to get something like the KAB SpeedStrobe to set the speed right at 33 1/3 and 45 rpm, then write down the correct settings for both the voltage and hz settings for the generator. I really don't care what the settings on the generator are, as long as I can get correct and consistent speed and voltage on the TT itself. So far it seems that is a realistic expectation.
Also I plan on running an external idler drive (single idler) on a ROK belt drive platter, as the belt drive platter is heavier than the internal rim drive platter of the idler drive ROK. The outer rim of the ROK belt drive platter is machined true for the belt, hence external idler drive. I have a mint belt drive platter and bearing that I'm planing on sending to Applied Fidelity, and get the "Full Monty" work over.
That means I'm going to have to machine down the capstan on the Papst motor, so there will probably be a slight error there as well. So the VFD should be able to be tweaked to get the speed where it should be. The way I see it the 1/10hz increments should translate to .0555 rpm @ ~ 33.333plater speed. I think that is more accurate than the fluctuations of 120V 60hz line frequency.
BTW, did you ever try those triple mica Voshkod tubes in your Little Bear? My bud bought a Little Bear, and we are running them in his, we think it sounds better than with the Chinese tubes. Here's a link to the signal generator.
twystd
Follow Ups:
Hi Twystd,
Thanks for the link. You are correct in that absolute frequency accuracy is not important as long as the stability and resolution is sufficient. In the end you will be adjusting the frequency to get the speed which you want.
I may just order this thing. I have also found others fully packaged which go up to several MHz and would be usable as lab instrument.
Is the unit which you have user friendly? I just see little buttons. Can you direct enter frequencies or do you have to scroll?
May I ask why when you have a belt drive platter and belt drive capstan machined into you motor shaft you are going to an idler drive?
I still have not gotten to the Little Bear. It still has the bogus RIAA feedback network in it. I may just go back to my 40 dB network which worked OK. Ideally I'll finish Lipschitz' paper and calculate the values according to his equations. It is still on my list of things to do. Thanks for the tubes by the way. I will definitely try them after I get the LB up and running again.
Phil
The generator is not that user friendly. The button on the left is a multiplier it scrolls from *.1, *1, *10, *100 for the frequency, then it repeats. The button on the right scrolls up, the button next to it scrolls down. The button second from left is the function button, and scrolls from output level, to A phase angle , then B phase angle, then to C phase angle, then to filter on or off, then repeats. So I set the A phase angle to 0, B phase angle to 120, and C phase angle to 240, and leave it that way, you don't have to reenter. When you power off, it also remembers the last frequency entered. Output voltage goes from E=0 to E=60. The output voltage level has to be reset every time, starts at max output after power off.
The power supplies I use have a standby mode, where only the low voltage supplies output. That allows me to power up the generator under standby mode, set the parameters on the generator, then I can power up the amps and thus the motor. Altogether, I can set it up and change speeds in less than 30 seconds. A little clunky, but you get used to it.
Like I say this is a no frills project. If they send you any documentation (highly unlikely) it will be in the worst Chingrish imaginable. Play with it, you'll get it. The auction site has rudimentary instructions on how to wire it.
The reason I don't want to go belt drive is I prefer idler drive for this application. I like that it is a more positive drive system, and I don't have to deal with belt slip, and or any belt stretch issues. Also this platter is going to be very heavy (more about that in later posts), and I'll want a very positive drive. With a heavy platter, and the torque available from running the Papst motor on 3 phase, I just feel better about it. The downside of idler drive, is that it can cause rumble, but I have a mad plan to deal with that. I'll get into that later too, when I build the plinth.
twystd
Hi Twystd,What you describe is tolerable. With the multiplier and the fact that it remembers, the scrolling should not be too egregious. To bad they don't store the level setting as well but what you did is a good workaround. Another approach which I may do if the outputs don't clip at the max setting is deal with the level with analog pots off board.
With what you said and what is on the auction I'm pretty sure I can use this thing.
Thanks,
Phil
Edits: 07/21/16
The nice thing about using the digital output to get your voltage right is that once you get the range of the voltages you need, they are repeatable. I guess if you only need 1 voltage that a pot may be more convenient. In my case, I will need at least a couple of output levels available, so that's why I went the route I did.
twystd
Well a pot would work, of course you'd have a 3 gang one to deal with. There is also a possible impedance issue to deal with. The generator has a DC offset. The way I approached the problem was to use DC blocking caps on the three outputs. I also had a problem getting the output levels attenuated to get the generator outputs (too much voltage) in the range that worked with my amps, and the eventual output voltage to the motor. I was worried that the impedance of the amps, the attenuating resistors, and the caps could impose phase shift, and possibly affect the efficiency of the motor, due to power factor problems.The way I went about that is to figure the amps input impedance (3k3) plus the attenuating resistor (5k6) for a total of 8k9 impedance each phase. Since the lowest frequency I was going to pass was 50hz, I went down a decade to avoid phase shift, and solved the cap value for 5hz. That comes out to about 3.58uf, so I'm using 3.3uf caps on the outputs. Not sure all that is really necessary, but decided to do it anyway, doesn't seem to hurt.
twystd
Edits: 07/21/16
Hi Twystd,
If the amplitude of three phases program together it won't be too much of a pain.
I think I'll just buy the three phase generator and start to play. At ~$38 I'm not risking much.
I'm a little afraid to push the motor to the voltage required to get to 78.26 and 80 RPM with the 33 1/3 RPM capstan. I'll try to make a push on cover for the higher speeds.
Thanks again and keep us posted on your progress.
Phil
Yeah, the generator does change the voltage output of all three phases at the same time. I don't understand what you mean by voltage for 78-80 RPM, that should remain the same. The RPMs are another thing, that motor is going pretty fast at only 81 hz, which should be about right for 45 RPM records. I'm only interested in 33 1/3 and 45 for my TT speeds. The tach should be here tomorrow.
twystd
twystd-
You have to think about the motor as both a motor and a generator (it is in fact, constructed very similar to an AC generator with a permanent magnet spinning inside of field coils). As the motor spins, it generates a back EMF (voltage) that opposes the drive voltage, effectively increasing the motor winding impedance. If you keep the voltage constant, at some speed the back EMF will equal the applied driving signal and the net effect will be zero current flowing in the motor (actually, it would stall before they became equal, but you get the idea). 78 RPM is 2.34 x 33 RPM so the back EMF will be similarly higher. On my 3 phase prototype, I monitor this and change the output voltage for any change in frequency so the power remains nearly constant. The drive voltage at 5 RPM platter speed is 2-3VPP, and at 78 RPM, nearly 24VPP.
Thanks for that. Since I'm seeing an increase in voltage when I change from 60hz to 81hz that is due to back EMF, and when I decrease the generator voltage to get 120V on the motor at 81hz I am actually decreasing the ratio of driving EMF to back EMF. I assume this would decrease torque at 81hz @120v, when compared to 60hz @ 120v. Do I have this right?
Now I'm thinking that maybe I should machine the capstan to run @50hz @ 33 1/3, and 67.5hz for 45. My transformer cores are good to 50hz. Do you think that would be a good idea?
twystd
The increase in voltage you are seeing is most likely a reduction in the voltage sag of your output. You can't really directly measure the back EMF, but it will manifest itself as an increase in effective impedance of the motor coils as the motor spins faster. Decreasing the voltage output at 81Hz will reduce the torque, but the torque will decrease with the higher speed even with the voltage held constant as there is less current flowing in the coils as the effective impedance increases.
I believe Mark Kelly posted somewhere that the approximate impedance of the motor coils = (coil resistance x forward voltage)/(forward voltage-back EMF). At zero speed, the output of the amp sees the DC resistance of the coils; as the motor increases in speed, the effective impedance will increase. If you want the motor to consume the same power at all speeds (and maintain torque) you will have to increase voltage with speed to compensate.
Going to 50Hz will require more power than at 60Hz. There's really no reason to do this unless you have problems with higher voltage at 81Hz.
Just when I thought I had a grasp on all this, going to have to re-think. I'm pretty sure this VFD will meet my needs, but a little more experimenting and testing is in order. Thanks again for the heads up, don't hesitate to let me know if I'm barking up the wrong tree. As you can probably tell, I'm completely self taught, so any help is greatly appreciated.
twystd
I have what I think is the same Chinese 3phase signal generator, and I've found that if you hold down the far left key after making changes the unit saves all the settings.
I'm in the fortunate position of building my setup for a Papst motor driving a Thorens TD124, so I don't need variable frequencies. Unfortunately my Papst motor isn't the one designed specifically for the TD124, and has a different shaft diameter, so I will have to machine my own drive pulley, not too much of an issue as I anticipate completing the finish after attaching the pulley to the shaft.
Thanks for the tip about the generator. There are other benefits with VFD other than changing speeds. I've discussed the accuracy required for machining a capstan, and there is time and trouble (basicly trial and error), which can be expensive in a machine shop. The 1/10hz adjustability theorectically gets you the same accuracy as .00015 on the capstan so especially helpful with capstan changes. There are the other benifits such as lower noise, higher torque, and lower temperatures, when running on true 3 phase as well.
Not saying you need to do it, but thought I'd point it out.
Twystd
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: