Author Topic: Troubleshooting a DIY constant current load  (Read 1980 times)

0 Members and 1 Guest are viewing this topic.

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Troubleshooting a DIY constant current load
« on: November 12, 2018, 08:10:39 pm »
Hello
I designed and built a constant current source (just like the one dave made) using a 25 turn pot, a shunt resistor, amosfet and an op amp.
The op amp that I had in hand was the MCP6002 (dual channel rail to rail op amp) so I thought why not use it to drive two mosfets, each with their own ballast resistor and voltage feedback.
I used a voltage divider after the 5V regulator in series with the 25 turn pot in order to get a maximum of 2.5A per channel. after building the circuit it works ... at least kind of  :palm:
one of the mosfets gets toasty hot while the other one remains stone cold no matter what. also it's extremely touchy and goes from 10mA to about 2A in about one turn of the pot!  :-// |O
I'm not sure what might be causing this or what the issue might be, so I'm gonna post schematics, PCB and the actual build here, maybe someone can help.
P.S. I forgot the 47u tantalum caps when designing the PCB so I just botch soldered them  :-DD
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1038
  • Country: ca
Re: Troubleshooting a DIY constant current load
« Reply #1 on: November 12, 2018, 08:45:49 pm »
OM, I see a few things:

1.) the opamp is probably oscillating trying to drive the gate capacitance of the MOSFET.  Place a 47 to 100 ohm resistors between the opamp output(s) and the MOSFET gate(s).

2.) connect the opamp's V+ terminal to 9V rather than 5V.  This will ensure the MOSFETs will get sufficient gate-source voltage to support the requested drain current.

3.) To reduce the touchiness of the current adjustment, you will need to place a resistor in series with the top or hot end of the potentiometer.

Cheers,

 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Troubleshooting a DIY constant current load
« Reply #2 on: November 12, 2018, 09:05:37 pm »
I try adding the gate resistor but the IRLZ44Ns are logic level fets, so 5V is sufficient for a full on state.
As I mentioned, I've already placed a 10k resistor on top of the 10k pot which is 25 turns! I can even get accurate millivolt adjustments using a multi meter. I'm not sure why the jump is so huge and rapid!
why does one of the channels carry no load tho? that just doesn't make any sense to me! it's an identical circuit with identical parts!
 

Offline aheid

  • Regular Contributor
  • *
  • Posts: 245
  • Country: no
Re: Troubleshooting a DIY constant current load
« Reply #3 on: November 12, 2018, 09:44:08 pm »
Maybe it looks worse than it is, but it looks to me like you have some spilled solder shorting the gate to ground, see attached picture, near the opamp.

Have you hooked up a multimeter probe to see what your "vref" and your current shunt voltage is when it goes from 10mA to ~2A in that one turn?
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Troubleshooting a DIY constant current load
« Reply #4 on: November 12, 2018, 09:47:31 pm »
it indeed was not shorting anything. I will need to use the arduino to measure multiple voltages at once if I want to see how Vref and Vgate and the voltage across the shunt behaves for both mosfets. let me write a quick code to do that and I will get back to you.
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Troubleshooting a DIY constant current load
« Reply #5 on: November 13, 2018, 12:34:17 am »
here is a picture of the supply voltage (used another 5V LDO) , output current and total power graph:


I circled around the places where I rotated the pot (about 1/4 of a turn) as you can see, the jumps are massive! it goes from about 200mA to about 225 then to 250 then to 500! at which point the LDO can't keep op and starts shutting down when I turn the pot down and as you can see there is a massive decrease with less than 1 turn again! from 335 to 75! I then turn it a bit more and boom! zero! after rotating it to increase the current, it jumps to full power in less than a turn again! I measured the Vref , Vgs for both fets and the the voltage shunt across both resistors and they are very well behaved. I can't give you exact measurements as I use a cheap multi meter so the numbers can be all over the place, but Vref and Vshunt were about 6mV as indicated by the multi meter (probably wrong and way off!) but they were matching which is good. Vgs for both fets was a also about high 1.7 or low 1.8V range which again is matched nicely ... I repeated the same measurements using a 12V 5A laptop charger and same story again ... I kept it at about 10.5 watts and the mosfet on the left got super hot again, while the other one stayed cold to the touch ... is it possible that the second mosfet arrived dead / died during soldering?


Edit (update):
so far I have noticed an issue:
The spacer nut I used to keep the mosfet from rocking back and forth, might have been shorting out one of the tracks (shorting source to ground), but even after lifting it up, it didn't seem to resolve the issue ... I'm not sure if this could have killed the fet or not  ??? how can I test the fet?


the op amp might also not be able to drive the gate capacitance as others have mentioned, but I have no way of testing that right now. I have to check with an oscilloscope in uni tomorrow.
« Last Edit: November 13, 2018, 12:56:45 am by OM222O »
 

Offline t1d

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: us
Re: Troubleshooting a DIY constant current load
« Reply #6 on: November 13, 2018, 02:28:23 am »
1) The circuit leaps to full on, because it is very likely that the MOSFETs are going into full saturation. There can be various causes,  which you will need to research. Research causes of oscillation in electronic load circuits. A noob mistake that I made was to starve the circuit of voltage and current. When I turned up those, things improved greatly.

2) The MOSFETs have different heats, because you do not have the load-sharing resistors suggested above. One MOSFET is doing all the work. Add those.

3) Even though the nut was touching ground, it would not ground out the MOSFET, because its mounting hole is isolated, from its internal circuit.

I just went through these same kinds of development issues. See the thread.
https://www.eevblog.com/forum/projects/electronic-load-project-ltv171-irfp250-with-kicad-files/msg1956007/#msg1956007
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11534
  • Country: my
  • reassessing directives...
Re: Troubleshooting a DIY constant current load
« Reply #7 on: November 13, 2018, 05:46:45 am »
The spacer nut I used to keep the mosfet from rocking back and forth, might have been shorting out one of the tracks (shorting source to ground), but even after lifting it up, it didn't seem to resolve the issue ... I'm not sure if this could have killed the fet or not  ??? how can I test the fet?
i bet the mosfet is toasted the very beginning you turned on the circuit thats why you experienced hot (shorted) mosfet. when you short source to gnd, shunt resistor will register 0V and opamp will go full swing up because it thinks no current is flowing, but in reality the mosfet becomes the burning coal. now with shorted fet, albeit rectified traces, shunt will register high voltage and opamp will shut off 0V to stop the current, but its no use since the fet is physically shorted. btw whats the shunt resistor value? if you have that china clone m328 transistor tester, that will make your life much easier. i suggest replacing the burning mosfet with a new one and redo the test, but beforehand make sure there is no short/fault traces again, do not underestimate this and treat your mosfet dearly otherwise electronics store seller will get to know you from your frequent visit.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Troubleshooting a DIY constant current load
« Reply #8 on: November 13, 2018, 07:33:16 am »
shunts are 1\$\Omega\$ each. I left the load on for a while and even the shunt resistor for the cold mosfet was getting hot too! is there a possibility that somehow the resistors are working in parallel and the second fet is not doing any work because of that? the gate voltage seems fine, it's neither 5 volts nor 0 ... it's withing 0.1 volt of the other fet!
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Troubleshooting a DIY constant current load
« Reply #9 on: November 13, 2018, 08:31:36 am »
yes! there indeed was a very small short across one of the resistors which made the voltages across them appear to be identical! after removing it, no longer the current jumps in less than one turn and the second mosfet starts to heat up as well! thank you all for your help!

 

Offline man_marius

  • Contributor
  • Posts: 17
  • Country: be
Re: Troubleshooting a DIY constant current load
« Reply #10 on: November 13, 2018, 09:24:33 am »
are you sure you haven't mixed up the Vn and Vp of the Op amp.  It seems to me the feedback loop should be connected to the Vp of the op amp and the Vref to Vn
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Troubleshooting a DIY constant current load
« Reply #11 on: November 13, 2018, 09:33:14 am »
are you sure you haven't mixed up the Vn and Vp of the Op amp.  It seems to me the feedback loop should be connected to the Vp of the op amp and the Vref to Vn

Feedback should go to the inverting terminal(Vn)  ;D my issue was only the small short circuit that I mentioned before.

Here is a new screenshot of the device working beautifully:
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf