Author Topic: Programmable AC source power stage - need insight  (Read 1436 times)

0 Members and 1 Guest are viewing this topic.

Offline homem_do_sacoTopic starter

  • Contributor
  • Posts: 15
  • Country: br
Programmable AC source power stage - need insight
« on: July 08, 2020, 04:46:05 pm »
Hi everyone.

Does anyone knows how the power stage of a product such as https://www.bkprecision.com/products/power-supplies/9801-programmable-ac-power-source.html might be laid out? I was particularly puzzled at the steep glitches it`s capable of: https://youtu.be/fWWU3lhEgB8?t=300 (it`s a portuguese language review, but one gets the idea from the video).

Alternatively, can anyone point me to the service manual of this product?

Thanks.

 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2852
  • Country: us
Re: Programmable AC source power stage - need insight
« Reply #1 on: July 08, 2020, 06:35:12 pm »
Well, this instrument is DESIGNED to simulate a number of abnormal conditions, such as short power dropouts, the use of SCR dimmers, etc.  I think that was a lot of what the video showed.

Jon
 

Offline homem_do_sacoTopic starter

  • Contributor
  • Posts: 15
  • Country: br
Re: Programmable AC source power stage - need insight
« Reply #2 on: July 08, 2020, 06:53:34 pm »
I`m puzzled in a "How do they do that?" way, not in a "Why do they do that?" way.

Specifically, would generating a low impedance, lower voltage waveform and feeding that to a step up transformer be the right way?
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Programmable AC source power stage - need insight
« Reply #3 on: July 08, 2020, 07:45:02 pm »
It is going to be something like a class-D output stage found in a switching audio power amplifier but with a current shunt and second error amplifier to support current limiting like in a constant voltage and constant current power supply.
 

Offline homem_do_sacoTopic starter

  • Contributor
  • Posts: 15
  • Country: br
Re: Programmable AC source power stage - need insight
« Reply #4 on: July 08, 2020, 08:21:08 pm »
It is going to be something like a class-D output stage found in a switching audio power amplifier but with a current shunt and second error amplifier to support current limiting like in a constant voltage and constant current power supply.


Thought about that, however, max voltage is 300V, so, that way, it must have +-300V power rails. On a DC power source, transformer taps reduce the wasted power on the transistor, but if the waveform is AC you go through all the possible voltages in each cycle.

It can be done using SCRs, but the waveform is too clean, I think, considering it also does 500Hz AC, so any low pass filtering would be above that.
 

Offline Dave

  • Super Contributor
  • ***
  • Posts: 1356
  • Country: si
  • I like to measure things.
Re: Programmable AC source power stage - need insight
« Reply #5 on: July 08, 2020, 08:53:47 pm »
You have a AC/DC SMPS on the input for generating a HV power rail (and doing the PFC) and then a class-D inverter, like David Hess suggested. You don't need bipolar rails, a bridged output topology is all there is needed.

The output can be filtered with a hefty LC filter to get rid of the inverter harmonics (likely in the 10-30kHz range, quite far from the maximum output frequency).
<fellbuendel> it's arduino, you're not supposed to know anything about what you're doing
<fellbuendel> if you knew, you wouldn't be using it
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Programmable AC source power stage - need insight
« Reply #6 on: July 08, 2020, 09:32:35 pm »
It is going to be something like a class-D output stage found in a switching audio power amplifier but with a current shunt and second error amplifier to support current limiting like in a constant voltage and constant current power supply.

Thought about that, however, max voltage is 300V, so, that way, it must have +-300V power rails. On a DC power source, transformer taps reduce the wasted power on the transistor, but if the waveform is AC you go through all the possible voltages in each cycle.

It can be done using SCRs, but the waveform is too clean, I think, considering it also does 500Hz AC, so any low pass filtering would be above that.

I am sorry if I was not clear.  When I wrote "class-D output stage" and "switching audio power amplifier", I mean one that uses pulse width modulation.  So the output circuit is a switching half-bridge followed by LC filter and powered by +/- 300 volts in your example.

Special provisions also have to be made for 4 quadrant operation with reactive loads because power can flow backwards into the power supply.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: Programmable AC source power stage - need insight
« Reply #7 on: July 08, 2020, 10:12:38 pm »
About ten years ago I designed and built an AC source for a small company.  The source was to provide about 1 kW at 150 VAC at up to 500 Hz.  It was cheaper and faster to use a commercial servo amplifier than to design one even though I had designed one over 20 years before.  I chose this unit: https://www.a-m-c.com/product/dpmaniu-030a800/  It could drive three phase or DC motors and had various feedback options but I operated it as a simple voltage amplifier.   It has a power bandwidth of about 1 kHz that was fine for this application.  I added an LC filter to the output to reduce noise.  I also added circuitry monitor the output voltage and current after the output filter, multiply them together and determine the actual real power delivered to the load.

In my case, the user provided the input signal that was generated by a personal computer with a sound card and some software.  I think B&K has done something similar and used a simple Arbitrary Waveform Generator inside their unit to supply the test waveforms.

There isn't a great amount of imformation on the design of AC Test Sources, however, as others have said, Class-D audio amplifiers (and servo drives) would use the same basic circuits and techniques.  PWM Audio amplifiers have greater power bandwidths so have higher switching frequencies and provide better filtering.  Here's a link to a tutorial: http://www.irf.com/product-info/audio/classdtutorial.pdf

Does this answer any of your questions?
« Last Edit: July 08, 2020, 10:15:41 pm by duak »
 

Offline homem_do_sacoTopic starter

  • Contributor
  • Posts: 15
  • Country: br
Re: Programmable AC source power stage - need insight
« Reply #8 on: July 09, 2020, 04:57:57 pm »
Thank you guys.

I had the misconception that harmonics of the synthesized waveform would show up unless filtered, and the PDF, David`s explanations and some studying took care of my doubts.


Special provisions also have to be made for 4 quadrant operation with reactive loads because power can flow backwards into the power supply.


Like a crowbar if voltage or current gets over some limit, for example?

Other than that, the waveform error amplifier would take care of loads with lower power factor, wouldn`t it?
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: Programmable AC source power stage - need insight
« Reply #9 on: July 09, 2020, 05:57:09 pm »
If needed, a circuit similar to the crowbar is used to dissipate excess energy - see file.  In the servo or motor drive world it's called a dynamic brake.  It detects when the DC bus voltage to the output stage is higher than a predefined threshold and connects a power resistor across the DC bus to reduce the voltage.  When the voltage is lower than the threshold, the resistor is disconnected.

The error amplifier tries to minimize the error regardless of the load whether it is resistive, reactive, complex, open circuit or zero ohms.  The power factor should not matter providing the voltage and current are within the V-I limits of the output stage.  Unlike a linear power output stage, a properly designed PWM output stage can drive virtually any load within its operating range.  This usually means there are inductors in series with the output connections to limit peak current.  These inductors can also be part of the LC output  filter if needed to reduce PWM switching artifacts.
« Last Edit: July 09, 2020, 06:00:00 pm by duak »
 

Offline Dave

  • Super Contributor
  • ***
  • Posts: 1356
  • Country: si
  • I like to measure things.
Re: Programmable AC source power stage - need insight
« Reply #10 on: July 10, 2020, 06:10:44 am »
Crowbars are intended to protect a DC power supply in the event of a fault that might damage the supply. It sacrifices a fuse to protect the rest of the supply.
This is intended as a last resort protection, not for normal operation. Having a reactive load on an AC supply is normal operation.

There's no need to waste any energy, you can just push the excess back into the mains network.
<fellbuendel> it's arduino, you're not supposed to know anything about what you're doing
<fellbuendel> if you knew, you wouldn't be using it
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Programmable AC source power stage - need insight
« Reply #11 on: July 10, 2020, 09:55:45 am »
Special provisions also have to be made for 4 quadrant operation with reactive loads because power can flow backwards into the power supply.

Like a crowbar if voltage or current gets over some limit, for example?

Other than that, the waveform error amplifier would take care of loads with lower power factor, wouldn`t it?

During half of the AC cycle when powering a reactive load, power flows back into the power supply.  This will increase ripple on the DC bus which powers the half bridge.  In some cases, some way to limit the maximum voltage on the DC bus may be needed.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf