Need advice! Thank you!

[whysolate]

Hi, I'm new to this forum. I'm doing some testing on High Temp Power MOSFETs. But I do not know much about electronics. I would like to know if there's any way to hold the device in place without soldering or little soldering. Something like the thing we use to hold ICs? Will it be able to withstand current up to 20A? There's a problem here cause the more desoldering and soldering to replace new test devices, there's oxidation and makes the connection bad. I attach a picture so that everyone can see clearly. I'm not sure if I explained clearly or not.

 

[audioguru]

It looks like you have tiny little surface-mount Mosfets. Their metal tab is usually soldered to a big copper area on the pcb for cooling. Unless they are turned on hard and turned on very quickly then they will be destroyed by heat when they conduct high current.

 

[whysolate]

Yes. You are right. During test, it will be pressed against a copper block to test its cooling capability. Any ideas how to avoid soldering? Cos I need to change new devices after a few runs.

 

[Optikon]

Those look like through hole parts with the leads bent. Like an I-pak or similar. The leads look too long to be a D-pak. So I am going to suggest you make a fixture that uses pin sockets to mount the devices into (no soldering) and use a clamp to rigidly sandwich the copper tab for a tight thermal connection to the heatsink.

Metal on metal will **NOT** make a good thermal connection. Take a look under a microscope at the metal surface to see why. for metal on metal use a thermal pad material on one of the mating parts to fill the voids with low thermal material. Stay away from thermal grease, nobody likes that stuff.

 

[stevez]

Clamping may be and effective way to handle this but it doesn't seem likely that you'd find something like a ZIF socket that will handle 20 amps. Still, it would seem that a fixture could be fashioned. I'd envision a two part fixture that is compressed - holding contacts against the device leads and also holding the body firmly against the heat sink. A toggle clamp might be used to apply the force.

I don't think the clamp will be exactly the same as soldering but it might be sufficient for your needs.

 

[Optikon]

Clamping may be and effective way to handle this but it doesn't seem likely that you'd find something like a ZIF socket that will handle 20 amps. Still, it would seem that a fixture could be fashioned. I'd envision a two part fixture that is compressed - holding contacts against the device leads and also holding the body firmly against the heat sink. A toggle clamp might be used to apply the force.

I don't think the clamp will be exactly the same as soldering but it might be sufficient for your needs.

No ziff socket... machined individual pin sockets. Done it before - ok for 20Amps if you get the big ones. Plenty to choose from in the market.

The only concern with the clamp is getting the thermal connection low enough. But I bet with enough force and a thermal pad one can achieve < 0.1 deg C/W which is likely going to be as good or better than the heatsink attachment in the final circuit.

 

[whysolate]

Do you guys mean a SIL pin socket?

Is this one suitable?

http://www.rsnewzealand.com/cgi-bin/bv/rswww/searchBrowseAction.do?N=0&Ntk=I18NRSStockNumber&Ntt=267-7416&Nty=1&D=267-7416&callingPage=/jsp/line/line.jsp&BV_SessionID=@@@@0965432106.1194424860@@@@&BV_EngineID=ccciaddmghfighecefeceeldgkidhgn.0&cacheID=nzie&Nr=avl:nz

 

[Optikon]

Do you guys mean a SIL pin socket?

Is this one suitable?

http://www.rsnewzealand.com/cgi-bin/bv/rswww/searchBrowseAction.do?N=0&Ntk=I18NRSStockNumber&Ntt=267-7416&Nty=1&D=267-7416&callingPage=/jsp/line/line.jsp&BV_SessionID=@@@@0965432106.1194424860@@@@&BV_EngineID=ccciaddmghfighecefeceeldgkidhgn.0&cacheID=nzie&Nr=avl:nz

I don't know.. your link doesnt work. If you go the socket route (Steves press connection idea can work too but may be more involved) just make sure you get sockets that are rated fro what you need.

IMO though, less than 20A sockets can be used if the vdrop is acceptable and the temp rise is acceptable.

What are you planning to do?

 

[whysolate]

Yeah..the link expired.

I can't seem to find any socket that has more than 10A current rating. Can you suggest one that is suitable for me? The max current is 20A, max Vdrop 3V, temp range from 90°C to 250°C.

This is a set up to measure the case temperature through thermocouples and junction temperature through Vds measurement.

 

[audioguru]

Why are you trying to destroy those tiny Mosfets?

 

[Nigel Goodwin]

I can't seem to find any socket that has more than 10A current rating. Can you suggest one that is suitable for me? The max current is 20A, max Vdrop 3V, temp range from 90°C to 250°C.

I doubt you will?, I wouldn't consider the transistor leads suitable for 20A, so a limited contact via a socket isn't going to be very good at all.

Although I'm still confused as to the reason?, surely you would only need to test a very small number of devices in this way?.

 

[Optikon]

I doubt you will?, I wouldn't consider the transistor leads suitable for 20A, so a limited contact via a socket isn't going to be very good at all.

Although I'm still confused as to the reason?, surely you would only need to test a very small number of devices in this way?.

what do you think will happen to the leads at 20A ? It is simply a question of temp rise (power dissipation) and v-drop. If both are acceptable, then the metal leads will survive.

The OP may want to look into how the manufacturer tests these parts.. and they do test them at 20A. But these are mostly pulse kind of tests.. if the OP could do the same, he can avoid high current rated sockets and expensive custom fixtures.

 

[whysolate]

Yes. This test is to accelerate the life time of the device. So high current will be used for that purpose. I forgot to mention that the device will be loaded with current only for 30 to 60 secs and will be cooled down for another 30 to 60 secs. This will be repeated continuosly for 1000 times. So I guess I do not need to use high current rated sockets? What about using a mount with screw terminals? Is that going to be better?

 

[ericgibbs]

Yes. This test is to accelerate the life time of the device. So high current will be used for that purpose. I forgot to mention that the device will be loaded with current only for 30 to 60 secs and will be cooled down for another 30 to 60 secs. This will be repeated continuosly for 1000 times. So I guess I do not need to use high current rated sockets? What about using a mount with screw terminals? Is that going to be better?

hi,
Why dont you strip out the screw terminals from a 15Amp barrier strip, solder them to the copper tracks and use the terminal screws to hold the device legs?

The 0.1inch pitch may be a limitation.

 

[stevez]

While the screws might provide good, reliable electrical contact I am not sure if the test fixture will be as stable or non-varying as you might need for your purposes. Given the test you are describing I would think that reliable and repeatable heat transfer is of some importance.

At higher power handling levels I would guess (I don't really know) that heat transfer from the leads can affect junction temp. On some terminal strips it might be difficult to control the heat transfer from the lead to the screw. A screw terminal with a piece of hardware to control compression might help. I'd be inclined toward a fixture with contacts that have large surfaces to mate up with a significant length of the lead surface (wide, flat side). That way lots of pressure could be applied without distorting the leads - and modest variability in pressure would not affect heat transfer much.

Think in similar terms for the case too. Some relatively constant and repeatable means of holding the device to the heat sink would be helpful. Toggle clamps might be useful here. A pneumatic cylinder might be useful. You might need to have a procedure to clean/prep the device and mating surface each time to reduce variability.

 

[mneary]

Your test fixture has machine-pin terminals for the incoming voltage/current. At a glance, they look like they would also work as a MOSFET socket if you placed them correctly. If they aren't the correct size, you could see if the vendor makes some that are suitable.