My Electronic Load

[Steven Waring]

Hi everyone,

Just wanted to share my first functional prototype of a design that wasn't just some blinking LEDs, which I think are really awesome at any rate.

I am starting with electronics design a bit late in life, I come from a software design background and now have the freedom to branch out to my other childhood interests.

I decided to take on a constant current load in part due to need, pointless frugality, general interest and not to mention the sheer wealth of information that has been generated by Dave and this community.

There are many thing I know are missing and things I want to add, first and foremost input protection for reverse voltage followed by a rudimentary voltmeter and ammeter for live display.

My hope is to eventually get the specs handling up to 100V at 1.5A or up to 150W at lower voltages.  I'm currently using two Bourns power resistors (1ohm each, in parallel) being switched by an IRL540NPBF power mosfet.  The opamp is an LM358n from an Amazon lot, getting an offset voltage (I think that's the term) that leads to a minimum 22mA draw so I feel like opamps and their tuning are going to be my next deep focus.

I am attaching some photos of the current prototype, initial schematic, as well as some bonus of the amazing equipment I am lucky to have on loan right (see if you can guess which devices I actually own )

Any kind of feedback would be great.  Thanks, and greetings.

P.S. Hi-Res in case anyone cares: https://imgur.com/a/4mI6JpJ

[dardosordi]

Congratulations on first project. I’m also a software guy that used to love electronics as a child.

From the schematic looks like the load could oscillate and might need a cap to compensate the loop. I’ve no idea (yet) how to calculate that but there are a few posts around the forum with good examples and plenty of very helpful and knowledgeable members that could help out.

[Etesla]

I'm also a beginner, but my understanding is that this kind of analog electronic load actually uses the transistor in its 'active' region. This means that the transistor itself (irf540n), not the big power resistors you have, dissipates the vast majority of the power. The thing works by effectively changing the transistors resistance to maintain the set current.

This means that if you are using your electronic load to load a 12V power supply at 1 amp, the two resistors you have would dissipate 1^2*.5 = .5 watts, and the transistor would dissipate (12-.5)*1 = 11.5 watts. This means your power resistors are hugely oversized power wise, and your mosfet requires a very beefy heatsink. Also, the opAmp will likely saturate at maybe 1.5V or something, so the minimum current this thing would be able to regulate would be something like 3A. This would be fixed by adding a negative supply rail, or getting a bit more creative with your single supply opAmp.

I recommend installing LTSpice to test this sort of circuit out. It's an invaluable tool, and it's pretty much bang on for circuits as simple as this.

My understanding is that no 'switching' is happening at all, but then again I might be out of my league

[Steven Waring]

Thank you dardosordi!  Really appreciate it.  I'm pretty sure it is oscillating, at least to a degree.  While running in the amp range I think it might be to small for the meters to pick up but when I am in the mA range the least significant digit doesn't always settle going +/- 1mA.  I will definitely investigate this further.

Etesla, I do believe you are right that describing the action as switching is incorrect, perhaps driving would be a good term, as it doesn't really make it all the way on or off.  I very deliberately overrated everything I could with what I had in stock and expect around 15 Watts per resistor in the end.  I think the heatsink/fan combo that the mosfet is mounted on is meant for quite warm core2duo generation processors, but I bought it quite a while ago before I was even aware that they could be more well defined.  In a later stage I was going to add another mosfet+resistor group in parallel to spread the heat out as well.

Also, the opAmp will likely saturate at maybe 1.5V or something, so the minimum current this thing would be able to regulate would be something like 3A. This would be fixed by adding a negative supply rail, or getting a bit more creative with your single supply opAmp.

This is interesting, I can't recall if I am familiar with opamp saturation.  I will have to do some reading.  My tacklife supply goes to 5 amps so I might try running that through tomorrow to get some real world experience to go with it.

I do have LTSpice installed and have run very simple manual dc voltage variation without the mc but it hasn't really become natural for me yet.  I think the last time I gave up I was trying to get the opamp parameters in correctly.

I have played with the idea of running a negative line for the opamp and recently built a little two transformer +/- supply with an smd negative regulator to practice my smd soldering as well.  For this specific application I have played with tuning using the positive and negative rails of the Keysight supply.  The best absolutes (I should have been taking notes on the specifics) I came up with were +11.6 and -1.something.  Does that sound about right?  I was looking into ways to generate the negative rail while still being able to use a single supply.  Voltage inverters like TC7660HCPA, perhaps combined with a negative regulator or an IC that does both.  Haven't gotten that far yet though, not enough time in the world.

Thanks for the input!

[Vovk_Z]

I downloaded but can't open your *.pdf unfortunately to say something.
I just want to say that it is a little unclear for me - are you using switching type of load? That is not good. Classic DC electronic load use Mosfet in active region.

[MarkF]

Using the normal op-amp and MOSFET symbols would make the circuit much easier to read.

Is all the input (from Arduino) filtering really necessary?

Why all the extra resistors on the output?

   

A simpler circuit with Arduino buffer and 10:1 voltage divider:
   With 5V output from Arduino and a 0.2Ω sense resistor, you will have a 2.5A load current.

   

I used this 5W sense resistor:
   https://www.mouser.com/ProductDetail/Vishay-Dale/LVR05R2000FE73?qs=sGAEpiMZZMtbXrIkmrvidDnVsdYYat5oyVAQwF7oirU%3D

[MarkF]

Consider using an IRFP250 MOSFET instead.
The TO-247 package is almost twice as big for better heat dissipation.

[Steven Waring]

Using the normal op-amp and MOSFET symbols would make the circuit much easier to read.

Is all the input (from Arduino) filtering really necessary?

Why all the extra resistors on the output?

A simpler circuit with Arduino buffer and 10:1 voltage divider:
   With 5V output from Arduino and a 0.2Ω sense resistor, you will have a 2.5A load current.

I used this 5W sense resistor:
   https://www.mouser.com/ProductDetail/Vishay-Dale/LVR05R2000FE73?qs=sGAEpiMZZMtbXrIkmrvidDnVsdYYat5oyVAQwF7oirU%3D

The input filtering was me experimenting with filtering in general.  I just experimented with it until the output looked smooth. 

The extra resistors are starting points for feedback to the Arduino, just not in use yet. 

This is very much a work in progress.  It looks like in the simplified circuit you posted that the filtering is happening on the input and out of the second opamp stage, is that correct?

I think I am going to put in a parts order and get some more appropriate resistors and mosfets before I go much further on this design.  I do like the looks of IRFP250MPBF. 

The schematic could definitely benefit from better symbols.  I have been using the part symbols available from searches like this - https://componentsearchengine.com/search.html?searchString=IRFP250&source=1048575&language=en&country=GB

I took some screenshots of the output to the gate, they are attached.  I think my next move will be to probe the load I purchased and compare against different circuits.  It will probably be a good way to visualise the issues my inexperience brings.  I will definitely be building your recommendation.

[Steven Waring]

I downloaded but can't open your *.pdf unfortunately to say something.
I just want to say that it is a little unclear for me - are you using switching type of load? That is not good. Classic DC electronic load use Mosfet in active region.

I believe I used the term switching incorrectly, it was just what was in my head at the moment.  I am attaching a .png of the schematic if you wish to look.

[MarkF]

You might try Diptrace for your schematics and PCB layout needs.
I have been using it for many years and find it very intuitive to use.

   https://diptrace.com/download/download-diptrace/

[MarkF]

The input filtering was me experimenting with filtering in general.  I just experimented with it until the output looked smooth. 

The extra resistors are starting points for feedback to the Arduino, just not in use yet. 

This is very much a work in progress.  It looks like in the simplified circuit you posted that the filtering is happening on the input and out of the second opamp stage, is that correct?

I think I am going to put in a parts order and get some more appropriate resistors and mosfets before I go much further on this design.  I do like the looks of IRFP250MPBF.

I only have a very minimal filter on the set value.

A little thinking out loud for an Arduino interface:



I have seen the IRFP250 used in some electronic load products.

[MarkF]

The oscillations in your screen captures are most likely due to your direct connections to the MOSFET.

The values I choose are simply because I already used them in the 10:1 voltage divider.
Start with 100Ω and 1KΩ resistors and a 10nF capacitor.
The Gate resistor probably needs to be at least 50Ω and
the capacitor needs to be at least 3 or 4nF.