 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : MOSFET Tester/Curve Tracer Design Help!
| Author | Message | ||||
LadyN Guru  Joined: 26/01/2019 Location: United StatesPosts: 408 |
MOSFET Tester or Curve Tracer Design Help needed So I'm starting to learn about MOSFETs and there will soon be a time I need to take all my simulations forward and actually build something on a board in real life. Since I will be sourcing a lot of MOSFETs from secondary sources like eBay, Ali etc, I want to check that they meet the spec that I need them to perform at. To this end, I am thinking of designing a MOSFET Tester: now I already have the AVR-Transistortester so I can get a reasonable reading of the MOSFETs. This will rule out obvious or poor fakes or DOA parts. However, there are also "higher quality" fake MOSFETs that display expected Rds etc readings but fail in the field where they are expected to perform to spec. To catch these, I am thinking of making a MOSFET Tester or Curve Tracer (is that what the name is?) that will actually put the DUT MOSFET under real load and collect the data so I can plot it and see it. So questions: 1. Is what I am thinking of making called a MOSFET Tester or Curve Tracer? Is this something known to engineers who do this for a living (like you guys)? 2. Is this device already available for purchase from our favorite sources? 3. If not, are the following parts enough to build this device: i. two PSUs a. one to drive the gate (this is preferred to be isolated? (to make the drive circuit simple)) b. other drives the drain and can offer Id, or atleast upto the Id I want to design the system at ii. atleast 3 sets of voltage and current sensors (Vgs, Vds, Vgd, Igs, Ids, Igd) iii. atleast one temperature sensor to measure the package temperature - but where do I mount it? on the heatsink? on the package thermally coupled with some heatsink compund (ewww messy!)? iv. a uC to control and log all these readings We can assume I will be playing with N-Channel fets for now. Am I missing something? |
||||
| noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 513 |
Sounds like a great idea.. i had some p75nf75's i bought (cheap i know) and had no end of trouble with them...random blowups and shorted driver ics etc. Was a pain... especially troubleshooting... ended up just throwing the rest away... but my trusty esr/transistor tester said they were fine. Obviously not blaming it, it only has a limited testing parameters which ive found to be pretty good. I have found a thing, if the mosfets gate capacitance is heaps lower than spec, they usually s@#t... I think it works !! |
||||
mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
I can't help thinking that the Saturation Inductance tester could also incorporate: 1. Power Mosfet tester. Instead of the inductor under test use a "standard" inductor and check the Rds of a FET under test. Gate capacitance would be an add-on. I wouldn't expect this to be a full-on curve trace, there's other testers for that. 2. The high current generated would be useful for measuring milli-ohms of all sorts of items, like switches and circuit breakers, lugs and bolt connections. The DSO is the natural, best device to take the measurements, but this wouldn't exclude the idea of using the adc of a micro to sample the test and display basic results and or upload it to a PC to do real analysis. Food for thought? Cheers Andrew |
||||
| LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408 |
Never thought of that. LOVE the idea! LET'S DO IT. Challenges I see: 1. "standard" inductor: The peak di/dt this would support would limit the Id it could test the FET at correct? In that case I was thinking of using a bridge rectifier setup that connects directly to grid power. The half wave gives us the nice voltage sweep from 0 - 200V that we need (we can limit the Vds of course) Would that be a simpler design? The inductor method could allow us to run this off a smaller PSU reducing chance of death so I am happy to go that method vs. messing around with mains voltage 2. > The high current generated would be useful for measuring milli-ohms of all sorts of items, like switches and circuit breakers, lugs and bolt connections. Yes, but how to we ensure those components dont get welded during that duration? |
||||
| mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
Using mains AC, I'd rather not  Regarding welding contacts etc, the current is limited by the inductor and by a uProc switching off the current when it hits a target current, and it's operator experience, don't put 30amps through a 5 amp device. Although it might be fun to destructively test some stuff.  I've not built a saturating inductance Tester, I want to, I'd like to hear comments from Tony, who refined the circuit, the two Mike's, I know Wiseguy Mike has a design in mind, and others. For references: Tony's idea Solar Mike's Saturation Tester Cheers Andrew |
||||
| mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
I was thinking of something like this. The schematic is just a rough of my ideas. 2019-02-16_172908_iSatTester_P1.pdf 2019-02-16_172928_iSatTester_P2.pdf uProc circuit is isolated from the test rig supply - use those cheap 240vac-5v psu. Drive the fets with tlp250 with isolated psu. This circuit should allow for a DSO or for the micro to capture the data, send it to PC or to a graphical screen. The good thing about the "Blue Pill" is that it has 2 ADC's and can work in "dual" mode - meaning simultaneous sampling. Cheers Andrew |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
I don't think that a curve tracer is going to be of very much help with mosfets used as high power switches. Its only really for linear type applications where you wish to closely match characteristics that a curve tracer becomes really useful. Most fake mosfets just fail because of very poor quality control. They will probably work fine until you start really pushing the specified maximums to the absolute limit. The only way you are going to know what the limits truly are, are by testing to destruction. And that becomes a pretty pointless exercise. Cheers, Tony. |
||||
 Print this page |
