The windings of the xfmr need to be in parallel, because those are the input and output voltages you are working with. If they are in series, you would need to drive them with 24VRMS (which would require a 35-40VDC supply if running BTL amps) and the output would be 230VRMS.
Whether you use parallel or series, you must verify the polarity of the windings are correct.
In parallel: Dot to Dot and non-Dot to non-Dot.
In Series: Dot to non-Dot with the output from the free end Dot and free end non-Dot.
Whether you use parallel or series, you must verify the polarity of the windings are correct.
In parallel: Dot to Dot and non-Dot to non-Dot.
In Series: Dot to non-Dot with the output from the free end Dot and free end non-Dot.
I followed the recommendations step-by-step and verified voltages according to your post, and verified the polarity. but I'm now having a different problem.
At startup I get about 128 VAC at each primary (no motor connected, and each transformer independent). This drops to about 116VAC after the 5 second delay on the SG4. One of the transformers is only generating 114VAC but a assume this small imbalance won't cause a problem.
HOWEVER, when I connect the 3 transformers together and then to the motor (as per your post) the motor does not turn properly and I hear a clicking from one or both of the amp boards. I assume this is the over-current protection? Could I be using insufficiently sized transformers?
I've verified the connections to the motor and everything seems to be OK. I don't know the draw of the Papst motor and can't seem to find it.
Any suggestions?
My power supply is only 20V - could that be the problem?
Also, using a single amp board to drive 2 transformers is likely connecting 2 of the transformers together on one side given a common ground at the output of the amp. Is this a problem - i.e. do I need to use 3 amp boards instead of 2?
You may want to increase the attenuation between the SG4 and the amps. You should attempt to get 115VAC at the outputs when the SG4 is at maximum. The SG4 can be configured to lower the voltage after 3 seconds; the range is from maximum (no attenuation) to ~half the max voltage, in 64 steps.
Which firmware version SG4 do you have? V1.02 and later start the SG4 at minimum and ramps up to maximum over a period of ~650mS to prevent the amps from shutting down. If you have version 1.01, the output goes to maximum at startup and will cause the amps to shut down.
The xfmrs should be sized right; going larger can make things worse. What is the power rating of the Papst motor you have?
Which firmware version SG4 do you have? V1.02 and later start the SG4 at minimum and ramps up to maximum over a period of ~650mS to prevent the amps from shutting down. If you have version 1.01, the output goes to maximum at startup and will cause the amps to shut down.
The xfmrs should be sized right; going larger can make things worse. What is the power rating of the Papst motor you have?
I not sure how you are getting 13.68VRMS at the output of the amps with only 20VDC supply, unless the output is clipped and the amps are running higher gain. A 25VA xfmr with a 12V secondary will actually produce ~15VAC without a load. You should need 15VAC to drive the secondary in order to produce 115VAC at the primary side, without a load. To produce 15VAC will require a MINIMUM of 21.2VDC which will give you no head room and the amp will be at the threshold of clipping. You should have a 24VDC supply and verify that the amps are producing 15VAC without distortion, into the xfmrs with no load on the primaries.
Depending on the TDA7492 amps that you have, they should be configurable for 2 channel stereo, bridge tied load (BTL), which means they do not share a common ground on outputs and should not be tied together. Neither side of each output channel should connected to ground, or anything other than the 2 leads of the xfmr secondary; they are opposite polarity and idle at Vcc/2.
The amp can also be configured for single channel mono which is not what you want, unless you run 3 amps, one for each xfmr.
Last edited:
I have version 1.03 of the microprocessor.
I can not determine the power rating of the Papst motor. It's from an Empire 298 table.
I will verify the voltages and look at the waveform on a 'scope and report back.
I checked and the output grounds are not connected to each other or to ground, so they appear to be operating as BTL stereo, which seems to be correct.
I tried using my bench supply set at 24VDC but the controller is drawing too much current and causing the bench supply to cycle. I will see if I have another 24VDC supply.
Thanks for your continued support.
I can not determine the power rating of the Papst motor. It's from an Empire 298 table.
I will verify the voltages and look at the waveform on a 'scope and report back.
I checked and the output grounds are not connected to each other or to ground, so they appear to be operating as BTL stereo, which seems to be correct.
I tried using my bench supply set at 24VDC but the controller is drawing too much current and causing the bench supply to cycle. I will see if I have another 24VDC supply.
Thanks for your continued support.
Home Depot sells a device called a Kill-A-Watt (or similar), ~$20. If you can power the motor from a standard wall socket, measure the power consumption using the Kill-A-Watt, that should give you some idea.
The motor will draw more at start up as there is no back EMF to reduce the current demand as there is when the motor is up to speed. The class D amps (and TDA7492 in particular) have a rather low (1.8-2.1A) threshold for over current protection and there is little if any delay, so they shut down at the first sign of trouble. You might try running them in mono (PBTL mode) to get a higher threshold or use a more powerful amp i.e. a TPA-3255 which can be configured for 4x SE, 2xBTL or 1xPBTL and is rated at 170W/300W/600W.
The motor will draw more at start up as there is no back EMF to reduce the current demand as there is when the motor is up to speed. The class D amps (and TDA7492 in particular) have a rather low (1.8-2.1A) threshold for over current protection and there is little if any delay, so they shut down at the first sign of trouble. You might try running them in mono (PBTL mode) to get a higher threshold or use a more powerful amp i.e. a TPA-3255 which can be configured for 4x SE, 2xBTL or 1xPBTL and is rated at 170W/300W/600W.
I have a line splitter and clamp on ammeter so I measured another, identical Papst motor. It draws .2A at 125VAC, so 25watts. Is this the problem?
I did find a 24VDC 8.8A power supply and connected it. The motor spins but I can still hear the protection clicking on the amps.
I eliminated the drop in voltage so the output of the amps stays at about 15VAC n the secondary and this gives 130VAC on the primary. Seems higher than expected?
It seems like the next step would be to try PBTL and 3 amp boards (I have several) because the TPA3255 would be about $100 for 2 plus it needs a new PSU. As I understand it, to convert to PBTL, I connect both L & R inputs and parallel the outputs (i.e L+ and R+ and L- and R- tied together). Is this correct?
Last edited:
That should be OK.
Does the motor stay running when you are hearing the clicks? If so, maybe the amps are going into oscillations. You will need to look at the outputs with a scope.
Yes, you should get ~115VAC at the output when the SG4 is at max. Increase the attenuation between the SG4 and the input to the amps or change the gain jumpers on the amp PCB.
My mistake. The TDA-7492 is NOT capable of PBTL (the TDA-7492PE is, as are other amps).
Hi Pyramid,
Would you suggest to connect 0V of 12V PSU and either 0V of 24V PSU to Safety Earth? I use advised by you Jameco switch PSUs and both have floating GND. 24V PSU has 3 prongs (Earth IN) and 12V has L/N only. Not that I have some issue, but for V reference and stability. Up to now, I used Linear PSUs and it was out of question. Decided to go for Switch PSUs die to PSU transformer his. It is all dead-quite now and all is at ambient temperature.
Would you suggest to connect 0V of 12V PSU and either 0V of 24V PSU to Safety Earth? I use advised by you Jameco switch PSUs and both have floating GND. 24V PSU has 3 prongs (Earth IN) and 12V has L/N only. Not that I have some issue, but for V reference and stability. Up to now, I used Linear PSUs and it was out of question. Decided to go for Switch PSUs die to PSU transformer his. It is all dead-quite now and all is at ambient temperature.
Last edited:
Yes, the motor keeps running when I hear the clicks. I don't know if it's developing it's full speed or torque, however. I have a scope but am not entirely sure how to check for oscillation (or what to do if I find it).
I think I'll insert a 10K pot in place of the resistor divider network as that will give more flexibility.
Any advantage to using a single amp per phase? I could connect the same input to left and right channel and one secondary to each channel. Does that offer any advantage?
Also, the SG4 has gone into reset mode several times when power is applied (shows "fac" on the display). Does this represent a problem or give me any clues?
Also, let me go back to where I started: with the primaries and secondaries connected in series, it worked, but the transformers got fairly hot. I noted that ralphfcooke mentioned that his transformers operated at 45 degrees C - which is probably close to what mine were doing. What can be learned from this?
I think I'll insert a 10K pot in place of the resistor divider network as that will give more flexibility.
Any advantage to using a single amp per phase? I could connect the same input to left and right channel and one secondary to each channel. Does that offer any advantage?
Also, the SG4 has gone into reset mode several times when power is applied (shows "fac" on the display). Does this represent a problem or give me any clues?
Also, let me go back to where I started: with the primaries and secondaries connected in series, it worked, but the transformers got fairly hot. I noted that ralphfcooke mentioned that his transformers operated at 45 degrees C - which is probably close to what mine were doing. What can be learned from this?
Last edited:
It's probably not a bad idea to connect the minus side (ground) of the DC supply to earth ground since the SMPS "float".
If it is producing a clicking sound while running, the oscillations could be occuring in bursts. You would see this on the scope as full scale bursts of noise (high frequency) superimposed on a low frequency (60Hz) wave form. As far as what to do about it, it depends on what is causing it. The amps should have a snubber network on the outputs to squelch high frequencies from showing up on the output. You are running a fairly inductive load which will impact the phase margin of the amp. You might try adjusting the caps on the outputs to see if it makes any difference.
If it was the over current protection kicking in, the amps would shut down until reset by cycling power, so I don't think that is happening.
That may help. You will double the output capability of each amp by doing so. Just be sure to observe polarities.
If the display shows "FAC" then the uP thinks both the UP and DOWN buttons are pressed at start up. Are you using a rotary switch?
If you are powering the SG4 regulator from the same supply as the amp, a temporary voltage drop out at start up will cause the SG4 to reset. Isolate the supply to the SG4 by adding a 3A diode (anode to supply, cathode to regulator feeding the SG4) and a decent reservoir cap (1000uFd) from cathode to ground.
That will increase the impedance of both the input and outputs which may be what is needed. The output voltage may sag more under load, but if it solves the problem, it may be livable.
I connected a 3rd amp board, but am only using a single channel on each board. This did not solve the problem.
I set the SG4 so the voltage drops back to yield 115vac without motor attached. I let the SG4 start and waited till the voltage dropped and then connected the motor. Still had the problem.
I checked the output from the primaries with a scope while connected to the motor. It was a very erratic signal - I get the impression it's just caused by the cycling on/off of the amp board but could be wrong. It doesn't look like high frequency imposed on a 60Hz sine wave.
I set the SG4 so the voltage drops back to yield 115vac without motor attached. I let the SG4 start and waited till the voltage dropped and then connected the motor. Still had the problem.
I checked the output from the primaries with a scope while connected to the motor. It was a very erratic signal - I get the impression it's just caused by the cycling on/off of the amp board but could be wrong. It doesn't look like high frequency imposed on a 60Hz sine wave.
Last edited:
I understand this but only vaguely (my lack of knowledge, not your explanation) so I hope I can avoid tinkering with the caps as I don't know rather to increase or decrease them.
It's good to know that the overcurrent protection is not kicking in.
I think next step is to try using both channels of an amp board for a single transformer. If that doesn't work, then I'll go back to primaries and secondaries in series and think about installing a fan.
More to come tomorrow when things may look brighter...
Thanks for your consistent help!!!
OK, It might be solved. I'm running 3 TDA7492 amps, each with one phase from the SG4 - to both channels. Each amp channel driving a secondary of the transformer and the primaries connected in parallel. I get a bit of clicking at startup but then it quiets down. I'm running some extended testing to see about heat and will (hopefully) report success.
Thanks for all the help!
Thanks for all the help!
Yes, looks the problem is solved. Transformers get to only about 85F and amp heatsinks maybe a bit more. I will add a small DC fan when put into final case. I learned that I need to allow ventilation for the papst motor. Currently running in enclosed plinth and it gets too warm. Also currently running at 87 volts under load and will try a lower voltage.
I couldn't have done this without the patient assistance from Pyramid. Thank you.
Well, it was not solved. When I re-cased the thing, the clicking problem was back. I was getting distorted sine waves to each output. So here's what I tried:
I connected the secondaries in series and used one channel of each amp. The primaries remain connected in parallel. This eliminated the clicking sound but I was still getting a distorted sinewave when connected to the motor. I wonder if the load was too low impedance?
When I reduced the input from the SG4 to the amp boards, the distortion disappeared. I suppose I was overdriving the amps. However, the output to the motor is only about 40VAC (rms) under load. But the motor runs very stably and cool. No over heating in the transformers or amps either. So I went a step further and switched to the 45 RPM section of the pulley (Empire table with Papst motor) and reduced the frequency accordingly. It works even better. I get less noise from the motor (1300 RPM vs 1800) and no heat. I could put the primaries parallel also but I'm not sure I see a need to do so.
I've run some 1-2 hour tests and everything seems to be fine. Am I doing something that will have bad long-term effects? Or is it just the Papst motor is so oversized that this approach works.
BTW, using full 120VAC and the capacitor for phase splitting, the motor was drawing about 60 watts. A Lenco draws about 15-19 W and VPI about 5 W, I think. So maybe just over engineered?
I hope to mount a cart on the table today and the listening test will be the real crucial one.