Continue to Site

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.

Mma welding system ( heat problems)

Status
Not open for further replies.

BGAmodz

Member
So i got recently a MMA portable welder , the problem with it is that the system shutsdown after a short time of usage .

I opened it and checked if something is wrong with it , and everything seems to work fine , Fan , bridge rectifiers , actually the unit turns on normally and also soldering intensity can be varied .
Next i checked the MOSFETs there is 12 MOSFETs 6 facing 6 , one of them was hard to move from the radiator's aluminium and also from the fixation screws's metal plate , an other one i noticed that the writing on it turned from white to yellow .

Could those particular MOSFETs be heating up so fast because they are getting bad or just due to some lake of heat dissipation ????
 
Last edited:
Next i checked the MOSFETs there is 12 MOSFETs 6 facing 6 , and two of them were hard to move from the radiator's aluminium and also from the fixation screws's metal plate ,

Not clear what that means. The transistors (mosfets or IGBT's) should not be easy to move from the heat sink. If they are not well attached to the heat sink, it may not be doing its job.

What is the duty cycle rating for your welder? Brand and model number?

John
 
That welder has a 25% duty cycle rating. That can mean 2.5 minutes out of 10 minutes or 1.25 minutes out of 5 minutes at rated capacity (supposedly). Are you sure you are not exceeding that duty cycle? What current are you running?

I am still concerned about you comment on the heat sink. Are all of the mosfets firmly attached to it? If not, the loose ones may be over heating.

John
 
Can you send a picture of the transistor to heat sink area. There should be no movement there. Something should press the transistor to the heat sink.
 
That welder has a 25% duty cycle rating. That can mean 2.5 minutes out of 10 minutes or 1.25 minutes out of 5 minutes at rated capacity (supposedly). Are you sure you are not exceeding that duty cycle? What current are you running?

I am still concerned about you comment on the heat sink. Are all of the mosfets firmly attached to it? If not, the loose ones may be over heating.

John
Ok after checking more i found a that there is a pink tape placed between the MOSFET's back and the heat sink , there is an additional tape that some one added by mistake , so i guess this additional tape may be causing less heat dissipation
 
Last edited:
Pics :

jp0ZPRW.jpg
hA6zmFN.jpg
 
That welder has a 25% duty cycle rating. That can mean 2.5 minutes out of 10 minutes or 1.25 minutes out of 5 minutes at rated capacity (supposedly). Are you sure you are not exceeding that duty cycle? What current are you running?
I don't know about that but the welding stops after less than 1 minute if usage ; in the other hand the system stays on .
 
That welder looks as if it has been repaired at some time. The bottom picture seems to be missing a mosfet. Did you remove it? Also, those mosfets in the lower picture are not properly attached to the heat sink. That could cause over heating by itself, and with a missing mosfet, you are certainly going to over heat.

John

Edit: Just checked the mosfet datasheet to be sure. The drains on those mosfets are connected to the tab ("heat sink"). That is why one set of the mosfets needs to have the tab insulated form the heatsink, unless two heat sinks are insulated form each other. Is it possible that the design has the heat sinks insulated from each other and someone later added the "pink tape" (looks like silicone heat conductive polymer). In any event, there should only be one layer of that tape.
 
Last edited:
No thats actually after disassembling it , the MOSFETS were attached very tight to the heat sink .
I will re-apply the thermal paste and see .
 
Edit: Just checked the mosfet datasheet to be sure. The drains on those mosfets are connected to the tab ("heat sink"). That is why one set of the mosfets needs to have the tab insulated form the heatsink, unless two heat sinks are insulated form each other. Is it possible that the design has the heat sinks insulated from each other and someone later added the "pink tape" (looks like silicone heat conductive polymer). In any event, there should only be one layer of that tape.
Yes that what i think too , maybe that layer was not supposed to be there .
 
With an ohmmeter, check to see if the two heat sinks are connected electrically. I suspect they are, and it would be a BIG mistake to remove the electrical insulation from that second set, if they are connected.

Since that is they way it was when you opened it, the set of mosfets with tape is not incorrectly connected to the heatsink and is missing one unit. That explains your overheating.

John
 
Ok jpanhalt , actually there is no missing MOSFET i removed that one to check if its working or not , and also everything was well connected before , now should i change the MOSFETs attached to that additional polymer tape ?? that tape may have caused poor heat dissipation due to the increased distance between the MOSFET's back and the heat sink ????
 
It is absolutely essential to determine whether the two heat sinks are electrically insulated from each other before you remove the insulating tape. I suspect they are not. If they are not insulated, and you remove the insulating tape, you will get a dead short (almost) from the power supply to ground. That will destroy your unit and possibly do other damage.

Parallel mosfets can fail one at a time. Just checking one, does not assure the others are OK.

As for heat, can you get a non-contact (IR) thermometer? You can check with that whether one bank of mosfets is getting hotter than the other.

John
 
As jp says the insulating sheet between the fets and the 'sink is vital, and the condition of it has to be immaculate.
Not all of these require heatsink compound, some dont.
An effective way to check those fets would be to use a component tester on a 'scope, I'm guessing you dont have access to one, stillyou could check the fets with a meter including measuring the body diodes voltage drop.
 
I can't add much here, but will re-iterate.

The surfaces must be flat and clean.
The thermal pad may or MAY NOT require grease. Slippery ones usually do. Rubber like ones with a patterned surface don't.
Use a VERY THIN layer of heat sink grease.

A cause of failure is that the MOSFETS aren't matched or for some reason are not at the same temperature. What happens is the hot one conducts more appreciably than the others in a parallel situation so it gets a disproportionate share and finally fails. A match might be some current, say +-10%, at a particular value of Vgs near the operating point.
 
A cause of failure is that the MOSFETS aren't matched or for some reason are not at the same temperature. What happens is the hot one conducts more appreciably than the others in a parallel situation so it gets a disproportionate share and finally fails. A match might be some current, say +-10%, at a particular value of Vgs near the operating point.

I thought one of the nice things about mosfets in parallel was that Rds increased with temperature. Thus the hot one carries less current. Here is the chart for an IRF1010E:
upload_2013-10-4_7-1-9.png

It is , of course, important to balance the gate turn on with resistors.

John
 
Ok , thanks all for the comments .
the heatsink attaching the hyperfast rectifiers is insulated from the other heatsink containing the MOSFETs and the polymer tape .
Now i applied new thermal past , changed two MOSFETs , and reattached everything tightly .

For now the soldering goes very well , seems like heat dissipation is normal .
 
Last edited:
Congratulations.

And, thanks for the follow-up. So often, posters forget to do that.

John
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top