Electronic Load Build Log

[hammy]

Thats a fine piece of work! 

[robrenz]

+1

[shailesh3t]

Hi i would like to know how did you match the output current to the setting point
am too building a dummy load using an arduino had it working but i have some oscillation on the mosfet i think
thanks \
shailesh

[jaxbird]

Thats a fine piece of work! 

+1

Thanks

I think I'll implement the USB/UART commands as something simple like this:

Command:   Description:

@MC#      Set Current Mode (Value is mA)
@MR#      Set Resistance Mode (Value is Ohm)
@MP#      Set Power Mode (Value is W)
@MN#      Set Random Noise Mode (Value is mA)
@START#      Enable load
@STOP#      Disable Load
@SV0000.00#   Set Value (mA, Ohm, W depending on mode)
@SC00.00#   Set cut off voltage (Value is V)
@READ#      Returns mode, state and readings
@HELP#      Returns list of commands


I did have to slightly revise my ground routing in a few places. I guess I was being a bit optimistic when I did the single layer board. Precision improved with some star like grounding using a few extra wires.

[jaxbird]

Hi i would like to know how did you match the output current to the setting point
am too building a dummy load using an arduino had it working but i have some oscillation on the mosfet i think
thanks \
shailesh

Hello, I am not 100% sure I understand the question, but this is how I do sampling and adjusting the duty cycle.

I attenuate the input voltage so it fit's within the range of the ADC e.g. 30V => 3V, mine is max 25V input. I've set the attenuation of the PWM to give me a max of ~6000mA and then with a 0.1R shunt I know the voltage over that resistor will be 100uV per mA, then I amplify the voltage over the shunt to match the scale of the 2nd ADC for current measurements.

The sampling, averaging and duty cycle adjustments are done on a timer interrupt that runs at 1kHz. The samples are averaged using 20 running averages.

The duty cycle is adjusted by a line of code something like this:

Code: [Select]

targetPwm += ((targetCurrent - lastCurrent) * 0.005);
This gives a nice self adjusting duty cycle without oscillation or overshoot.

Note: targetPwm is not the duty cycle, it is the on time.

If you have problems with oscillation, check that you have a configuration of the driving opamp and MOSFET similar to the one I posted in my partial schematic in one of the first posts. (Couple of resistors and a small capacitor) Power MOSFETs have significant capacitance on the gate.

[dr_p]

Nice, happy to see you putting all that equipment to good use.

Minor suggestion: you could print (regular 2D) the front panel text and graphic labels on a transparent foil (mirrored on the back of it). Transparent paper is available in inkjet or laser variety. You then mask the LCD position with a temporary rectangular sticker and spray it all with white paint (also on the back side). Remove the sticker and you have a white background with text on top and a window for the LCD. Just glue it in place or cut it to go under the edges of the enclosure.

Also, FWIW you can mount LCDs and other stuff directly on the enclosure's metalic/plastic front panel with sunken screws that get covered by the previously mentioned transparent panel.

[shailesh3t]

Thanks  jaxbird for the explanation,
i will try to make a video and post it in a new post , else its a very nice one u made ,
mine am using INA 219 to read the current and voltage out of the shunt, and am Using dac 4725 to set the voltage on op amps followed by the mosfet,

[jaxbird]

Nice, happy to see you putting all that equipment to good use.

Sure, no point in building a small CNC and 3D printer if I wasn't going to use them

Minor suggestion: you could print (regular 2D) the front panel text and graphic labels on a transparent foil (mirrored on the back of it). Transparent paper is available in inkjet or laser variety. You then mask the LCD position with a temporary rectangular sticker and spray it all with white paint (also on the back side). Remove the sticker and you have a white background with text on top and a window for the LCD. Just glue it in place or cut it to go under the edges of the enclosure.

Also, FWIW you can mount LCDs and other stuff directly on the enclosure's metalic/plastic front panel with sunken screws that get covered by the previously mentioned transparent panel.

There are many ways to do this, this one I did the 3D printed front because the raw front didn't look so great with the display sticking out from behind and everything. I did consider adding some text, but I don't really need it. I could 3D print some letters and use hot air to melt them on there, but I like the look as it is.

In my experience the best way I can make nice front plates is to use a piece of sign material (the stuff with a different color layer on top), then CNC engrave the text. That looks really great. I have also done a few really fancy ones in painted acrylic and then engrave and add some LEDs and the artwork/text will light up when switched on.

[jaxbird]

Thanks  jaxbird for the explanation,
i will try to make a video and post it in a new post , else its a very nice one u made ,
mine am using INA 219 to read the current and voltage out of the shunt, and am Using dac 4725 to set the voltage on op amps followed by the mosfet,

You can do the same kind of continuous feedback adjustment with a DAC. Anyway, looking forward to see your build.

If I build another one I would probably make a few changes like using a small transformer for power supply, use some better opamps and be a bit more careful with my ground. Probably I would also use 2 MOSFETS instead of just one. And of course I should really isolate the USB from the device. But overall I'm pretty happy with the performance and features, it is a very useful little device.

[shailesh3t]

Jaxbird , would be Nice to have a video of your Dummy load in action , was trying to have some pictures of mine my camera battery went down, have been able to take some pictures will try to upload it
regards
shailesh

[shailesh3t]

Thanks  jaxbird for the explanation,
i will try to make a video and post it in a new post , else its a very nice one u made ,
mine am using INA 219 to read the current and voltage out of the shunt, and am Using dac 4725 to set the voltage on op amps followed by the mosfet,

You can do the same kind of continuous feedback adjustment with a DAC. Anyway, looking forward to see your build.

If I build another one I would probably make a few changes like using a small transformer for power supply, use some better opamps and be a bit more careful with my ground. Probably I would also use 2 MOSFETS instead of just one. And of course I should really isolate the USB from the device. But overall I'm pretty happy with the performance and features, it is a very useful little device.

i have posted the link for the dummy load i built, https://www.eevblog.com/forum/projects/dummy-load-digitally-controlled-by-arduino-328/

[jaxbird]

Jaxbird , would be Nice to have a video of your Dummy load in action , was trying to have some pictures of mine my camera battery went down, have been able to take some pictures will try to upload it
regards
shailesh

Nah, I think I'll just stick to the old fashioned text and pictures for now

i have posted the link for the dummy load i built, https://www.eevblog.com/forum/projects/dummy-load-digitally-controlled-by-arduino-328/

Great, I'll have a look

[jaxbird]

Got the USB interface implemented, I can now control everything the load does from a PC, plus read all the values. That was really the reason I built it, I want to be able to easily test some programmable/variable buck/boost converters.

I have sometimes used LogView to monitor my battery chargers, so I had a look at adding support for LogView.

They have something they call OpenFormat, it's a bit quirky, but it works, just need to send values like this:

Code: [Select]

$1;1;0;1133;318;0;0;0;0;0;20;21;0;16<cr><lf>
Details of the format can be found here: http://www.logview.info/cms/d_formatbeschreibung.phtml (mix of German and English  ) .. My German is getting a bit rusty, but google translate does a decent enough job

Anyway, here a capture of discharging a battery using the Load:

[shailesh3t]

Its very Nice the Graph, you have a smooth value read out,
hmm which mosfet are you using in your design?
regards
shailesh

[jaxbird]

Its very Nice the Graph, you have a smooth value read out,
hmm which mosfet are you using in your design?
regards
shailesh

Thanks .. actually I think the software is reducing my resolution a bit.

I watched your video and I think Dave's example was a bit too simplified, it's best if you drive your MOSFET something like this: (Sorry for really bad drawing  )



My MOSFET is an IRL2203 (lowish voltage, high amps), but it's not particularly selected for this project, it's just what I had in my parts drawers.

[shailesh3t]

Jaxbird, will try the dummy load cct u have done and see how it works
n will revert back
Else today is my last day of holiday as will soon join another work so wont have time with projects now, and got a small baby to look after just turned 11 months.

else i have tried it with the cct you have made and its better now in most range of the current ,  i added one component at a time to see its effect and the 1 k was the one who reduced most of the  noise, could you tell me what did the 1 k resistor do ?
thx
Shailesh

[jaxbird]

Jaxbird, will try the dummy load cct u have done and see how it works
n will revert back
Else today is my last day of holiday as will soon join another work so wont have time with projects now, and got a small baby to look after just turned 11 months.

else i have tried it with the cct you have made and its better now in most range of the current ,  i added one component at a time to see its effect and the 1 k was the one who reduced most of the  noise, could you tell me what did the 1 k resistor do ?
thx
Shailesh

Hello Shailesh,

You'll need all 3 components for the proper effect, the problem is that opamps cannot supply enough current on the output to drectly deal with the capacitance on the gate of a high power MOSFET. So by adding these resistors and capacitor, you can slow down the feedback and avoid oscillation. If you still have problems, you can try adjust the values until you achieve a good result.

Anyway, best of luck on both the new job and your baby

[shailesh3t]

Hello Jaxbird,
have joined th work and it seems to be going fine now, else will try with other values and see the results n THX
n regards
shailesh

[jaxbird]

Hello Jaxbird,
have joined th work and it seems to be going fine now, else will try with other values and see the results n THX
n regards
shailesh

That is great  you should be able to make a nice load device with the parts you have

[brabus]

Wow, that's really well made! My congrats!

I really find the idea of using a P4 heatsink... pretty simple and strong! Well done!

You know what? I may take inspiration from this, and improve my active load... thank you!