Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

n-channel mosfet

Status
Not open for further replies.

jeff10049

New Member
Hello everyone,

I have been reading threads here for a while great board, Registered tonight to ask a question about mosfets.

I am working on a stepper motor driver I found some bad mosfets part number IRFS644A.

Also found the data sheet IRFS644A pdf, IRFS644A description, IRFS644A datasheets, IRFS644A view ::: ALLDATASHEET :::
It calls it a "advanced power mosfet" not sure what the "advanced" means I could not find much info about advanced mosfets.

The mosfets are known to fail on these boards after about 2 to 4 years. I am looking for replacement mosfets to repair it I will need 16 of them there are two boards and I want to go ahead and do them both. Hoping you guys with more experience can help me find a suitable replacement if there is a mosfet that might be a better choice "last longer" I'm interested. I just do not know enough to make the call hoping to learn some stuff here not just get free info I'm enjoying playing with electronics.

Hopefully I will have something to contribute to this board someday when I learn more.

Thanks
Jeff
 

Sceadwian

Banned
Mosfets don't go bad if they circuit they're in and their environment and usage are appropriate, they should last until long term electron drift starts blending the diffusion regions of the semi conductor materials. If these Mosfets are being run at nearly their full rated load this could very well be what's happening, if that is the case simple chose a Mosfet that's rated for two or three times the current the one you're currently using is and try to match up the rest of fet parameters to match the old fet, you may need to alter your fet drive circuitry if the gate characteristics of the new fet are different enough that's up to you to determine. You won't know if it's done any good for about 4 years though =}

As Mosaic said though it could be a waste of time, the failure could be from transients or something else entirely. I would lean towards the original fet being under rated for the usage, but that's just a guess.
 

MikeMl

Well-Known Member
Most Helpful Member
I beg to differ with Sceadwian. If the original problem is weak gate drive, and gate charge storage causing the FET to take too long to switch through the critical region where it dissipates tens to hundreds of W, replacing the existing FET with a larger one will just make the problem worse. This is what will cause a failure over time. Even though the average dissipation may within the capability of the heatsink, the repeative peak dissipation will cause thermal stresses inside the package which will eventually lead to failure.

Sounds to me that the problem could be intrinsic in the design of the stepper controller. Just because this is a commercial product doesn't guarantee that the original design was done properly.
 

crutschow

Well-Known Member
Most Helpful Member
Do you have the part number for the driver circuit or a schematic?
 

Sceadwian

Banned
Mike, we could beg to differ all day long without a schematic and components values, I did say I was just guessing, much like you are =) There is not enough information to determine what's going on.
 

jeff10049

New Member
Hello,

Thanks for all the replies, I don't have a schematic for the circuit it is for a SWF embroidery machine they keep all that stuff as proprietary information.

I can give a little more info that might be help full lots of good advice above has got me thinking. The machine has two of these boards they are identical one drives the y axis and it has one stepper motor 2 amp that board is still working.The x axis board drives two stepper motors it always fails first. The motors are 2 amps each so that would put the x axis drive fets under more load but not over loaded or at least I don't think so the data sheet says 5.6 amps for the fets that are in it now.

Also Mikeml mentioned a weak gate drive that maybe a problem I don't know how to check it, I do have a scope I need to get leads for it and see if it still works. If I understand correctly if the board was designed with a weak gate drive then increasing the current value of the fet will only make things worse?

While one this topic if the gate drive is weak could a smaller fet be used ahead of the main fet to increase gate drive voltage/current? or leave it be and just replace the fets every few years?


The motors seem to work at 30 volts checking with a fluke meter thought that might be use full info.

Mosiac, Thank you for the links very informative my mosfets did fail short circuit.

OK so on to what I did I replaced the two short circuit mosfets with fets from the local radio shack they were part# irf510 the specs were not as good on current. but this was just done as a test to see if the steeper's would work under a no load condition. However now it won't work at all? maybe those fets are just two far off spec to work the motor is still in brake mode anytime the power is on like always.
Before attempted repair It was running very rough when cold, after warm up would just shake if you tried to run it.

This is making me lose confidence in being able to fix the driver maybe someone can tell me if those irf510s just wont work at all I'm hoping thats the case and a replacement with a proper fet will get it going again. I'm still having trouble comparing specs and finding a suitable replacement I don't mind putting the exact same thing in it'll last a while but the only ones i found were in hong Kong and they want me to buy 1000.

Also this board has isolation transformers on the control so I don't think the shorted fets hurt any thing else but thats getting over my head.

Thanks again, sorry for being a pain in the ass new guy I don't want to just come in looking for someone to tell me how to do it. Looking forward to learning more and more its pretty cool to start understanding how this stuff works.

Jeff
 

Noggin

Member
I agree with ronv. The FET he linked to does look like an improvement. It has a higher resistance, but the gate charge is lower so it should switch faster.

Then again, it's possible that its switching too fast. I worked on a design recently that was switching too fast. The load was slightly inductive, so the voltage at the drain would spike to 700 volts. Put a 350v zener there to absorb the spike, it just killed the zener after a few minutes. Put a cap there to help bypass some of the current. The cap just made the instantaneous current though the FET be massive and the FET died. Put a resistor in series with the FET's gate to slow the switching. That made the miller currents increase the voltage at the gate and the gate died.

However, again, I agree with ronv. I think that FET he posted is worth a shot.
 

jeff10049

New Member
OK guys I'm back with pictures and more info, I appreciate all the help so far I have been reading night and day about mosfets and driving them I now have a pretty good understanding of them I see where you could pull your hair out trying to get everything just right.

So to get up to date on what I have done. Both the boards x and y had some shorted mosfets all three legs were shorted together on 3 mosfets on the x board I'm just going to focus on that board for now.

So I ordered the Fairchild mosfets that ronv posted I sat down and compared all the specs read all I could about what they mean I think I could make the call myself in the future thanks for pointing me in the right direction there are lots of choices out there.
I replaced all eight fets it went well came out nice, reinstalled the board same thing the stepper motors just shake when trying to move they would move if you helped them along. (I want to add that the problem is not external to the board works fine with boards from other machine).

Removed board for further inspection I'll post some pics at this point so you can see what I'm working on might help.

here it is the top side you can see the mosfets I replaced, also everything on the board past the row of transformers seems to be logic and it has diagnostics that say its working OK.
10876-fan plan 2011 097.jpg
does that make sense? If so The first transistor (Q26 in above photo)in all eight gate drives is shorted base to collector what could cause this? I verified by desoldering one end on a few of them to make sure I was not testing through another part of the circuit they were still shorted showing 0 resistance either direction on ohm meter and no voltage drop either way using diode check function the second and the last transistor in the drive check as you would expect but again all 8 first transistors shorted base to collector. could the board still work like this? but have a weak gate drive? maybe this is what was killing the mosfets? so many questions I just don't see 8 transistors all failing at exactly the same time.

So far it looks as if I need at least 8 transistors and one optocoupler to get this board working. still confused about this transistors I guess it would just loose one stage and the optocoupler would be directly driving the second transistor through the first? and still maybe work dose that sound plausible?
Trying to understand how I could loose 8 transistors and 3 fets and a optocoupler in my head transistors died off over months maybe years causing a weak gate drive, that put strain on the fets until they failed, the shorted fet took out the optocoupler what do you guys think of that scenario?

Sorry for the long post I guess my main questions are am I on the right track, do you think I have correctly id the failed components, and could all those transistors be bad and the board still work? for a while.

Thanks for all the help
Jeff
 
Last edited:

jeff10049

New Member
Hey, guys just thought Id bump this up to see if anyone has any input, for some reason my last reply showed up a couple pages back so may not have got much exposure. I think it had to do with waiting in the moderation queue because I'm new. I don't normally ever post to just to get on top hope you guys understand.

Thanks
Jeff
 

ronv

Well-Known Member
Most Helpful Member
Boy, without a schematic it is really hard to tell. Usually the opto isolaters stand between the high voltage and the rest of the logic.. But again hard to tell. It sounds like a problem that is common to all the motors.
 

jeff10049

New Member
thanks ronv,

Yes the opto isolators do stand between the high voltage and the logic I was able to determine that. This board controls one bipolar stepper motor that runs at 30 volts. so from what I'm Reading here that means the gate drive would need to be 40 volts? I am thinking thats what the series of transistors on the gate is doing driving up the gate voltage ? I ordered some new smd transistors that have a little higher current rating and matched up the rest of the specs. hopefully that will help with the failure of them shorting base to collector but I'm not sure the rated current has anything to do with the type of failure. In hindsight it seems maybe the collector base breakdown voltage may have more to do with it.. at least if i get it working it should last another 4 years or so and thats a lot of time to learn more about electronics.


thanks
jeff
 
Last edited:

Sceadwian

Banned
40 volts gate drive? Not likley. Typical mosfet gates will have a gate insulator breakdown at 20 volts. Gate drive for N fets is relative to the source voltage though.
 

jeff10049

New Member
ok I guess I did not understand when I read about fets, somewhere I thought I read that to turn on a fet needed to have gate voltage driven above source. Could you better explaine how a fet is turned on? Sorry im not getting it I thought I had that part figured out.

Thanks
Jeff
 

ronv

Well-Known Member
Most Helpful Member
Ok, I got it. So a bad opto could mess it up. For some reason I thought there were several motors.
 

jeff10049

New Member
Thanks ronv,

Yes each opto controls one side of motor winding so i can see where that would cause my problem of shaking motor I'm feeling pretty confident that if i replace the opto and all eight of the shorted transistors it will work. now I'm just trying to better understand how it works and why all those transistors driven by the opto failed.

Also I just read this https://www.electro-tech-online.com/custompdfs/2011/01/dcdc5_driver.pdf what am i missing? to me it says gate voltage must be higher than sorce but that cant be right because the fets do have a breakdown voltage of about 20, I know I'm just not reading it right or missing something.

thanks
Jeff
 

ronv

Well-Known Member
Most Helpful Member
No, your right the gate needs to be 10 volts above the source. So if you have an H-bridge configuration with the source of the NFET tied to +30 the gate needs to go to +40. But if the source is at ground it need only go to +10. It all depends on the set up.
 

jeff10049

New Member
OK so now I'm hoping that I'm wrong about the stepper operating at 30 volts but I'm pretty sure thats what i measured I must be wrong though because that would obviously kill the fet on the high side because that would mean +40 gate drive and they are rated for 30 max.
4 of the fets are at ground and 4 at I thought 30 volts I'll have to recheck that. is there a way that the fet could operate at 30 volts and be turned on safely?
I checked it at the motor with a volt meter and got 30 volts again maybe my dmm is picking up some stray signals and giving a false reading?

I'm just getting more and more confused oh well I'm determined that with some help I can figure this out.

thanks
Jeff
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Top