Author Topic: Half-Bridge 0-30V, 0-20A Feedback loop problem  (Read 10628 times)

0 Members and 3 Guests are viewing this topic.

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #25 on: June 17, 2019, 06:04:53 pm »
I've done more tests. I switched off the voltage measurement point for the error amplifier on the common mode choke + 4 capacitors 1000uF caused very high loop instabilities. The loop itself was able to arouse. I did not try on the choke itself.

However, I left the voltage measurement point as in the diagrams, but additionally soldered 200nF between the comp and gnd plus plus the choke common mode choke and 4 capacitors 1000uF. The feedback loop was calculated as if the output capacity was 7000uF and the ESR was 3.57mohm.
Now the whole thing works much better. The transformer at small loads is slightly playing but not terribly loud. Certainly much lower than before.

He also adds graphs of output voltage, evaluate how it looks.

« Last Edit: June 17, 2019, 06:22:26 pm by paladyn »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #26 on: June 18, 2019, 03:15:52 am »
Trying to filter with capacitors can trade one problem for another if the roll off frequency is too low causing phase lag and that ~1.4Khz oscillation in the loop.
It is difficult to figure out the effect the 220nF has at the output of internal error amp. Its output would tend to follow the input signal then start slew limiting when the 100µA limit is reached.
You could try a low pass filter at the input.
It looks promising though.
I would like to see the implementation of the common mode choke.

That 100Hz ripple at full current is likely to be normal.
« Last Edit: June 18, 2019, 03:17:24 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #27 on: June 18, 2019, 08:21:23 am »
Direct connection of the output voltage by the common mode choke can be problematic. Mainly because the voltage measurement is single and should be rather differential. When measuring Vout with one common ground wire, I will not have a second signal to pass through the common mode choke.

Unless I understand it wrong?

As for the low-pass filter, I will try to check this option, just please point me at what frequency should this filter be? Unfortunately, the filter itself will strongly influence the calculated feedback loop coefficients.

I met with such a solution that the output voltage was read differently and additionally an offset of about 100mV was added. Have you met with such solutions?

I will also write that when I connected the oscilloscope to the COMP pin which still did not contain 200nF capacity on the oscilloscope, there was a sharp chaff, no periodic run.

Maybe in fact some disturbances penetrated the measuring circuit and the whole was excited.

I will try to check what I get on the output when I remove the common mode choke together with 4 capacitors 1000uF
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #28 on: June 18, 2019, 09:01:26 am »

The CMC is a temporary solution that should get the cleanest voltage signal from the output capacitors into the voltage sense input of the control board.

Experiment with an RC low pass filter at the voltage sense input with a roll-off about 10KHz.

"I met with such a solution that the output voltage was read differently and additionally an offset of about 100mV was added. Have you met with such solutions?"
Not yet.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #29 on: June 19, 2019, 05:58:38 am »
I checked the inverter's reactions to the ones used
RC filter on the voltage measurement input. 6.8 kohm resistor + 2.2nF capacitor. The transformer is loud and the output waveforms were much worse than at the 200nF capacity between COMP and GND. I expected this reaction, because according to what I read, this filter will enter into the calculation of the compensator coefficients.

I have also tested the influence of the common mode choke, I have not noticed anything beneficial here either. There are still loud squeaks of the transformer at different frequencies. You can see that there is no stability.

I wonder how your inverter works since virtually everything is connected on wires in the air. Life suprises...
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #30 on: June 19, 2019, 07:00:33 am »
Did you try a higher roll-off for the LPF?
You also have to try to distinguish between loop instability that's caused by incorrect compensation that caused the 1.4Khz oscillation and the other type of  instability caused by interference to the PWM comparator that causes abrupt and random changes in the drive pulses.
Mine has a trifilar CMC that feeds 12V/gnd and V sense to the control board but this was mainly to reduce interference to other sections of the power supply.
I also drive the MOSFETs with a 1.2µs Gate plateau although it causes some extra dissipation that my heatsinking can deal with.
« Last Edit: June 19, 2019, 07:56:20 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #31 on: June 19, 2019, 07:32:55 am »
Did you try a higher roll-off for the LPF?
No, I will try to check a few variants at a free time.

I still have the idea to connect the common mode choke to the differential amplifier and then give the amplifier a loop. Maybe it will help.

I do not know if I observe it well, but the higher frequency? Maybe if I moved a single field (fp0) to a higher frequency it would be better?
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #32 on: July 01, 2019, 07:49:20 am »
After a few days of break, I come back from the battlefield. I checked that on the secondary side power supply, the one that powers the SG3525 and the operational amplifier and noticed noise and ripple. That's why I'm now powering the secondary side with a separate power supply and this problem is solved.

The next test consisted in using the common mode choke at the output of the inverter along with the 4000uF capacitor package. The voltage feedback signal has been wound on the common common mode choke core with the same as the power supply. The effect? Very bad, big oscillations.

Then, for this configuration above, I used a differential amplifier so that the coupling voltage would be the difference between the output voltage i and the ground terminal. Also the lack of a positive reaction.

Another test is also error voltage read differently but this time without common mode choke in common with power lines. Also, the effect is poor.

I added a separate common mode choke to the differential amplifier independent of this at the output of the inverter, the effect also does not cause much noise and squeaking of the transformer.

The last attempt was to replace the resistor R9 in the feedback loop with a 500k potentiometer. It turns out that there is no point for which the transformer does not write or does not have any strange waveforms at the output of the inverter.


I will still try to replace the operation amplifier with another one. Maybe it will change something.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #33 on: July 01, 2019, 10:37:13 am »
After a few days of break, I come back from the battlefield. I checked that on the secondary side power supply, the one that powers the SG3525 and the operational amplifier and noticed noise and ripple. That's why I'm now powering the secondary side with a separate power supply and this problem is solved.

The next test consisted in using the common mode choke at the output of the inverter along with the 4000uF capacitor package. The voltage feedback signal has been wound on the common common mode choke core with the same as the power supply. The effect? Very bad, big oscillations.

Then, for this configuration above, I used a differential amplifier so that the coupling voltage would be the difference between the output voltage i and the ground terminal. Also the lack of a positive reaction.

Another test is also error voltage read differently but this time without common mode choke in common with power lines. Also, the effect is poor.

I added a separate common mode choke to the differential amplifier independent of this at the output of the inverter, the effect also does not cause much noise and squeaking of the transformer.

The last attempt was to replace the resistor R9 in the feedback loop with a 500k potentiometer. It turns out that there is no point for which the transformer does not write or does not have any strange waveforms at the output of the inverter.


I will still try to replace the operation amplifier with another one. Maybe it will change something.
I did some experimenting with the compensation in my bench supply to find what should be guaranteed stable albiet slow.
In your Half-Bridge_Analog_Controller-5.pdf schematic, R9,C8 would become 5.6K,1uF.  R1,C1 are open circuit.
When I do a load transient test, I see a few cycles 200Hz ringing at the output. This is close to the calculated output LC  filter resonance of 229Hz, normal and expected.

Passing the output voltage to the control circuit via a CMC definitely should not make things worse. Mine is 5 turns on a small torroidal core, about 50uH.

How fast are your MOSFETs switching?

I was planning to use IRFP460 MOSFETs . The heat-sink I used came from a plasma TV PSU and had a pair of SD20n60 MOSFETs bolted to it.
The plan was to use them while debugging the SMPS instead of risking my IRFP460 MOSFETs .
These  MOSFETs must have very low charge causing them to switch way too fast. I had to use 220Ω and 220Ω with diode at the Gate to get the Gate plateau time up to about 500ns.
I didn't bother using the IRFP460 MOSFETs.


HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #34 on: July 01, 2019, 12:18:03 pm »
The switching frequency is 40kHz, while the gate resistors have probably 12 ohms in parallel with the diode. The question is whether 220ohm is not too much? Of course, I can solder and check, but it will definitely affect the amount of heat released.

You used 5.6k and 1uF feedback, tell me how did the inverter work? Have you noticed any squeals of the transformer or noise at the output from the inverter?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #35 on: July 01, 2019, 12:32:08 pm »
The switching frequency is 40kHz, while the gate resistors have probably 12 ohms in parallel with the diode. The question is whether 220ohm is not too much? Of course, I can solder and check, but it will definitely affect the amount of heat released.

You used 5.6k and 1uF feedback, tell me how did the inverter work? Have you noticed any squeals of the transformer or noise at the output from the inverter?
220Ω would make the IRFP460  too slow. I can't say what the  IRFP460 would need to slow it down by the right amount.
With the  5.6k and 1uF in the feedback, ,mine was stable, just slow with no transformer noises except at low output voltage and no load for reasons that I explained earlier, not a real problem.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #36 on: July 02, 2019, 10:00:52 pm »
Also consider bridging out D8 in case noise is getting in after it.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #37 on: July 03, 2019, 06:30:06 am »
Yesterday I installed a new TL3474CDR operational amplifier, it is a little better, you can see the difference but still the transformer is squeaking and there are considerable distortions.

I checked the situation again with common mode choke on the output of the inverter and on the same core I wrapped 5 scrolls from Umes. Unfortunately, there are even more distortion on the output waveforms.

I'm wondering if diodes D8 and D9 are really a problem here. Maybe too big capacity? There is also the issue of choke and feedback. I wonder how much CCM or DCM work mode is important in calculating the feedback loop coefficients.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #38 on: July 03, 2019, 07:36:39 am »
Yesterday I installed a new TL3474CDR operational amplifier, it is a little better, you can see the difference but still the transformer is squeaking and there are considerable distortions.

I checked the situation again with common mode choke on the output of the inverter and on the same core I wrapped 5 scrolls from Umes. Unfortunately, there are even more distortion on the output waveforms.

I'm wondering if diodes D8 and D9 are really a problem here. Maybe too big capacity? There is also the issue of choke and feedback. I wonder how much CCM or DCM work mode is important in calculating the feedback loop coefficients.
Because you are testing the CV loop only, it would be ok to connect the output of the op-amp directly to the Comp pin 9 of the SG3525,
Are you using the simple compensation that I described earlier?
I did not think about  CCM and DCM. If there was a loop problem with DCM, it would happen only at low loads.
I would like to see  how you are applying the Common Mode Choke.

EDIT:  No more than 5V should be applied to the Comp pin 9 of the SG3525. It only needs 3V for full duty cycle.
« Last Edit: July 03, 2019, 08:11:39 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #39 on: July 03, 2019, 08:57:54 am »
I will try to do some pictures today how it looks. See how the common mode choke and the rest. You write to connect the loop amplifier to pin 9 SG3525, but what to do with the SG3525 amplifier inputs? I could simply desoldute the coupler adding diodes and solder the loop amplifier to pin 9 to leave the internal SG3525 amplifier as it is today, i.e. a voltage follower (pin 9 connected to pin 1).

As for compensation according to your description, I only tried without a C1 capacitor. Do you suggest using compensation as you did in the first post? Should there be voltage feedback? Possibly, could you suggest some element values to start with for testing?

My loop is adapted to 3V voltage.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #40 on: July 03, 2019, 09:38:59 am »
I will try to do some pictures today how it looks. See how the common mode choke and the rest. You write to connect the loop amplifier to pin 9 SG3525, but what to do with the SG3525 amplifier inputs? I could simply desoldute the coupler adding diodes and solder the loop amplifier to pin 9 to leave the internal SG3525 amplifier as it is today, i.e. a voltage follower (pin 9 connected to pin 1).

As for compensation according to your description, I only tried without a C1 capacitor. Do you suggest using compensation as you did in the first post? Should there be voltage feedback? Possibly, could you suggest some element values to start with for testing?

My loop is adapted to 3V voltage.
I didn't fully understand what you said about the SG3525 amplifier, My suggestion about driving pin 9 directly is likely a waste of time and voltage clamping is required to prevent it going above 5V. The internal error amp can source/sink only 100µA.
Do bridge out D8 though.
Yes, use the compensation as I describe earlier. Voltage feedback is still required.
Did you ever try someone else's suggestion of running the power supply in open loop by applying a fixed voltage to pin 9?
The idea is to find stable state to work from.
Actually I would prefer that the internal amplifier only to be used with the simple compensation.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #41 on: July 03, 2019, 10:23:48 am »
When writing about the internal amplifier, I thought of pin 1 and pin 2 of the SG3525 chip.
I've done a little bit already, so I'll try to make a summary.

1. Unsolder diode D8 and diode D9. The output of the TL3474CDR amplifier of the voltage loop (pin 7 in the diagram) should be connected to the pin 9 of the SG3525 system. After this procedure pin 2 SG3525 have a pull-up to Vref, pin 1 hangs in the air.

2. Unsolder C1 and R1

Can I run a test on such a configuration as above or change something?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #42 on: July 03, 2019, 10:40:09 am »
When writing about the internal amplifier, I thought of pin 1 and pin 2 of the SG3525 chip.
I've done a little bit already, so I'll try to make a summary.

1. Unsolder diode D8 and diode D9. The output of the TL3474CDR amplifier of the voltage loop (pin 7 in the diagram) should be connected to the pin 9 of the SG3525 system. After this procedure pin 2 SG3525 have a pull-up to Vref, pin 1 hangs in the air.

2. Unsolder C1 and R1

Can I run a test on such a configuration as above or change something?
Pin 1,2 of the SG3525 can be left floating or connected to something. The main thing is to put a zener diode between pin 9 and ground so that it can't go above 5V.
You can then test, If the result is still poor, try changing R9, C8 to 5.6K,1uF.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #43 on: July 04, 2019, 06:35:16 am »
Yesterday I did not have much time to test, but I changed the control loops as we agreed. The omission of the diode D8, D9 and the internal amplifier SG3525 (pins 1 and 2) did not give much. Loop configuration with 5.6k and 1uF I have not tested enough time.
I did a test where C8 is 27nF and R9 is an adjustable 500k resistor and I noticed a certain dependence. Noise occurs anyhow in the entire adjustment range, but the inverter in most cases works quietly above the load of 10A up to 20A. When the load is eg 5A and I tried to adjust the loops with the potentiometer, the converter works relatively quietly when the noise shape is sinusoidal, when the sinusoid disappears and the noise has a shape similar to the straight line the transformer begins to squeak.

I tried to see what the case with an additional 4000uF capacitor set and the CMC filter on the output is a little better than without it, but if I tried to wind the Umes measurement cable on the same CMC choke as the inverter output I can not stabilize the loop anymore. Similarly, if I use some other choke, e.g. on a yellow core, and make a few loops.

I will try to take pictures of the inverter today and carry out further tests.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #44 on: July 19, 2019, 08:40:32 am »
After a long break, I put pictures of the inverter.

I tested feedback where the output of the TL3474CDR operational amplifier is connected to pin 9 of the SG3525, and the pins of its amplifier are not connected. No positive results.
However, I was able to determine that when the operational amplifier SG3525 works as a voltage follower, the noise generated at the output is noticeably higher. When I am connected directly to pin 9, COMP is much better even when using two couplings as at the very beginning of our discussion.

After combining with the 250uH reactor I returned to the 47uH choke and again I used a type 3 compensator. However, this time I tried the brutal-force method and instead of the R1 and R9 resistors soldered the 10k and 500k potentiometers.

I observed that it is not possible to set the coupling correctly. When the loop is too fast you can see distortions of the waveform at high currents, a waveform similar to a one-wave rectifier, maybe the 47uH choke, but with a slower loop, a sawtooth shape is also obtained with a large amplitude. At low currents the squeaking occurs on the fast loop and on the slow loop.

Somewhere, I read that it is impossible to design loops for every load. I wonder if I should replace the analog loop and just do a simple PID on the STM32 without any coefficients etc.

Due to the fact that the photos are large, I will post them in several posts.
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #45 on: July 19, 2019, 08:41:03 am »
2
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #46 on: July 19, 2019, 08:43:21 am »
3
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #47 on: July 19, 2019, 08:45:50 am »
4
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #48 on: July 19, 2019, 08:47:30 am »
5
 

Offline paladynTopic starter

  • Contributor
  • Posts: 42
  • Country: pl
Re: Half-Bridge 0-30V, 0-20A Feedback loop problem
« Reply #49 on: July 19, 2019, 08:49:35 am »
6
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf