Author Topic: Op amp questions in an electronic load  (Read 3352 times)

0 Members and 1 Guest are viewing this topic.

Offline BootalitoTopic starter

  • Supporter
  • ****
  • Posts: 116
  • Country: us
    • EasyEda.com/TerryJMyers
Op amp questions in an electronic load
« on: December 18, 2017, 01:05:58 am »
I have a question about the purpose of some of the features from From Scullcom Hobby Electronics version 1 DC electronic Load schematic.  Regarding the DAC->op amp>op amp>Main Load FET below:

R4 is a 10k to ground on the output of the DAC going to the first voltage follower.  Why?
What is the purpose of U7A?  Just to be able to provide more current to the voltage divider for U7B? (mcp4725 DAC can only supply 200uA)
The purpose of U7B is to set the correct current across the shunt(just below the fet not shown).  But what is the purpose of the passives I've highlighed, R7, R8, and C19?  Is R7 to help match input bias current from R6?  I see that C19 and R8+R7 make an low-pass filter, is this just for output stability?

Link to original schematic:
https://www.youtube.com/redirect?redir_token=eZr8JP27fBXLxwHclL_oS-TY9rt8MTUxMzY0NDk5NUAxNTEzNTU4NTk1&event=video_description&v=9fsf1CgnTRk&q=http%3A%2F%2Fwww.scullcom.com%2FElectronic_DC_Load_part1.zip
« Last Edit: December 18, 2017, 01:08:05 am by Bootalito »
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: Op amp questions in an electronic load
« Reply #1 on: December 18, 2017, 01:27:32 am »
R3 looks like a termination for the DAC's output (op-amp inputs are hi-Z), and I think you're right on u7a unity gain buffer. That whole 7-part series is one of the best on any project I've ever seen. I still think he could wrap it up nicely with a 8th part (summary) and sent it to Dave for January...

I know his video's are quite long, but are you sure he left something out? He's usually super at saying everything...  :-//
 

Offline BootalitoTopic starter

  • Supporter
  • ****
  • Posts: 116
  • Country: us
    • EasyEda.com/TerryJMyers
Re: Op amp questions in an electronic load
« Reply #2 on: December 18, 2017, 01:58:25 am »
Indeed he does mention it:
, and it was for ringing.

However he never explains what the 10k pulldown resistor is on the output of the DAC does.
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: Op amp questions in an electronic load
« Reply #3 on: December 18, 2017, 02:59:17 am »
It's a termination as I see it, keeps the DAC output clean from pcb trace picking up noise.
 

Offline basinstreetdesign

  • Frequent Contributor
  • **
  • Posts: 458
  • Country: ca
Re: Op amp questions in an electronic load
« Reply #4 on: December 18, 2017, 03:40:18 am »
That DAC can supply up to 25 mA and stay in spec so I don't know what R4 is for or why the DAC needs to be buffered.

Since R7 is part of the feedback path from the output to the -ve input then C19, R8 make sure the OA get adequate feedback at higher frequencies.  The FET and R7 may not have enough bandwidth at those frequencies to keep the op-amp stable.

If R7 was to match voltage drops into the OA due to bias currents then a value closer to 2K || 18K would be better.
« Last Edit: December 18, 2017, 03:44:59 am by basinstreetdesign »
STAND BACK!  I'm going to try SCIENCE!
 

Offline tombi

  • Regular Contributor
  • *
  • Posts: 163
  • Country: au
Re: Op amp questions in an electronic load
« Reply #5 on: December 18, 2017, 05:22:20 am »
Does R7 prevent the opamp output from affecting the current through the sense resistor?

Tom
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16605
  • Country: us
  • DavidH
Re: Op amp questions in an electronic load
« Reply #6 on: December 18, 2017, 06:22:49 am »
R4 is a 10k to ground on the output of the DAC going to the first voltage follower.  Why?

They might have included R4 as a safety in case the output of the DAC went high impedance so the load current does not run away.  But if that was the case, then a resistor should have been placed from the gate to the source of the MOSFET as well.

Quote
What is the purpose of U7A?  Just to be able to provide more current to the voltage divider for U7B? (mcp4725 DAC can only supply 200uA)

That DAC should not have any trouble driving a 12k voltage divider so I do not know why U7A and R4 were included.  They selected a quad operational amplifier so maybe they just wanted to put the extra operational amplifier to use.

The voltage divider is used to match the full output range of the DAC to the more limited range of the current shunt yielding the maximum resolution.

Quote
The purpose of U7B is to set the correct current across the shunt(just below the fet not shown).  But what is the purpose of the passives I've highlighed, R7, R8, and C19?  Is R7 to help match input bias current from R6?  I see that C19 and R8+R7 make an low-pass filter, is this just for output stability?

The input bias current of the AD8630 is too low to bother matching the source resistance and the value of R7 would need to be 1.6k.  R8 and C19 are part of the frequency compensation to handle the delay through Q2 because of its high capacitance.  They require R7 to work against for them to have reasonable values; without R7, they would have to be much lower impedance to work against the shunt resistance and U7B would not be able to drive them.

The pole and zero frequency compensation indicates an effort to make well tuned design.
 

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4694
  • Country: au
  • Question Everything... Except This Statement
Re: Op amp questions in an electronic load
« Reply #7 on: December 18, 2017, 06:26:46 am »
R4 is so when you power down the DAC via I2C, its output falls to 0, as the dac pulls far more current than the op amps in general.

 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: Op amp questions in an electronic load
« Reply #8 on: December 18, 2017, 02:53:01 pm »
Buffer spoiler Alert! If you haven't watched the whole series yet... he starts out conceptual with just one MOSFET and moves on to more.

BTW, the length of the series is precisely why I think it could use a summary video using snippets from all the footage he shot. It was definitely a labor of love and maybe even deserves a "second suck of the sav" during Dave's well deserved vacation.
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2218
  • Country: mx
Re: Op amp questions in an electronic load
« Reply #9 on: December 18, 2017, 03:51:56 pm »
R8 and C19 are definitively required for stability.

Does he mention how he calculated those?
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16605
  • Country: us
  • DavidH
Re: Op amp questions in an electronic load
« Reply #10 on: December 18, 2017, 09:56:28 pm »
R4 is so when you power down the DAC via I2C, its output falls to 0, as the dac pulls far more current than the op amps in general.

So that answers that.  R4 is important for safety because the DAC:

In power down mode, the output amplifier can be configured to present a known low, medium, or high resistance output load.

High in this case is suppose to be like 500 kilohms but it is a good idea to include R4 anyway.

R8 and C19 are definitively required for stability.

Does he mention how he calculated those?

I scanned through the videos and did not find any place where he tested the stability of the feedback loop or calculate the values for the feedback network.

For something as simple as this, I usually find the feedback network values empirically through testing and substitution and then confirm them through calculations however in this case, with a known transistor capacitance, transconductance, and high source resistance from the operational amplifier in his final version, calculating the values is straightforward.  With the schematic as shown here, calculating the values will not be enough because of the undefined and nonlinear operation of the operational amplifier with such a low load impedance.  It could be done but the values would need to be confirmed empirically anyway.
 
The following users thanked this post: Inverted18650

Offline Inverted18650

  • Frequent Contributor
  • **
  • Posts: 875
  • Country: us
    • Test Gear Addicts
Re: Op amp questions in an electronic load
« Reply #11 on: January 23, 2018, 01:24:15 am »

This project is no my bench now and I am reworking it for a more traditional "300W, 10A" load. I had 5 PCB's made, so if you are going to build this, let me know. I am making all kinds of fun changes and tailoring it to my needs.

He is also working on chapter 8 in the series so expect an update soon.
« Last Edit: January 23, 2018, 01:37:41 am by Inverted18650 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf