Mosfit Blowout Prevention & Modification

[specter]

Ok as many of you know from my posts I have been experiencing mosfet blowouts/shorts causing units from different manufacturers to have to be replaced.

The cause in each case was the mosfet retention bar being overtorqued and causing hairline cracks in the mosfits. The torque on the steel retention bars have averaged 8-10 foot pounds when in fact the torque should have been in mere inch pounds. This was confirmed by an ISO Certified Tech using Certified/Calibrated Quality Assurance Tensioning Tools.

As seen in the pictures below (from several units) each mosfet had been cracked from overtorque on the mosfit retention bar screws. The unit when powered up would send a charge through the mosfet to the circuit board but because of the crack in the mosfet it would short against the steel retention bar and then cause the mosfet to blowout. And the mosfet blowout in turn causing the failure of the unit. In each case the units breaker tripped preventing a possible fire.

Both my shops tech and I came up with the following suggestion to the manufacturer as well as those of you currently expericencing this problem.

1. Make the mosfet retention bars out of plastic instead of steel. The new bars would be cheaper to produce. The new bars would also be nonconductive preventing arcing through the mosfets. The flexiblity of the plastic bars would still allow the bars to be properly tensioned down onto the mosfets.

2. If not going with #1 above consider the use of a conductive spacer inserted between the mosfets and retention bars during assembly of the unit. The nonconductive spacer should prevent any blowout/arcing against the steel retention bars.The spacer could be made out of the same nonconductive (pink) thermoplastic already in use behind the units mosfets and their aluminum retention board. In my case I chose clear thermoplastic.

3. In either suggested method or in your current method you should use a proper tensioning tool to ensure that the mosfets are not overtorqued. Proper assembly and QA will prevent mosfit hairline fractures and hopefully prevent further mosfet blowout and unit failure.

Sure hope this help anyone experincing the same problems I have had.

Tom

[specter]

Sorry it limited me to posting five pics with that first post..

Note that in my modification using the clear thermoplastic I extended the non conductive thermoplastic 1/16 to 1/8" beyond the edges of the metal retention bar. Purpose of this is to prevent arcing over the edge of the protective spacer.

addon: In this case I illustrate one spacer. This can be done throughtout the unit to prevent mosfet blowout. Or you can choose to make one long piece of material that would fit neatly behind the row of mosfets on each board.

[jetenginedoctor]

Mosfet

[specter]

oppsie

Thanks for pointing that out. I corrected all but the titles since it wouldn't allow title correction

Tom

[jetenginedoctor]

:)

Hell about this forum is they don't let you go back and fix spelling errors, etc. Good info and pictures, though. Hopefully it'll save somebody the headache you appearantly suffered. Sure wish I could spell more gooder the first time I post.

BK

[Tensaiteki]

They are MOSFETs not 'mosfits', there is no 'i.' MOSFET is an acronym for metal–oxide–semiconductor field-effect transistor, and should therefore be written in all capital letters. You managed to spell (and capitalize) it wrong 19 times in your two posts, not counting the thread title.

Also in your suggestion #2 you said "consider the use of a conductive spacer". Presumably you meant nonconductive spacer.

edit: jetenginedoctor beat me to pointing out the spelling errors.

[jetenginedoctor]

How did you get it to let you write MOSFET in all caps? Stupid forum software de-capped everything after the M in my post.

[Tensaiteki]

How did you get it to let you write MOSFET in all caps? Stupid forum software de-capped everything after the M in my post.

I simply typed it in all caps.

I think the forum software will decap if either the entire post is written in all caps or it will decap if and entire line/paragraph is written in all caps.

All the more reason to not make one-word posts.

[Fred Paul Jones]

Why not just use the handy scew hole in the mosfet itself? Drill and tap holes for each in the heat sink.

[specter]

well so much for touch typing and a failure to spell check

[specter]

Hi Fred I am pretty sure that if the assembly line overtorqued the retention bars they would have done same on each mosfet. I think the idea behind the retention bar is to spred out the torque.

[kepe]

Both my shops tech and I came up with the following suggestion to the manufacturer as well as those of you currently expericencing this problem.

1. Make the mosfet retention bars out of plastic instead of steel. The new bars would be cheaper to produce. The new bars would also be nonconductive preventing arcing through the mosfets. The flexiblity of the plastic bars would still allow the bars to be properly tensioned down onto the mosfets.

2. If not going with #1 above consider the use of a conductive spacer inserted between the mosfets and retention bars during assembly of the unit. The nonconductive spacer should prevent any blowout/arcing against the steel retention bars.The spacer could be made out of the same nonconductive (pink) thermoplastic already in use behind the units mosfets and their aluminum retention board. In my case I chose clear thermoplastic.

3. In either suggested method or in your current method you should use a proper tensioning tool to ensure that the mosfets are not overtorqued. Proper assembly and QA will prevent mosfit hairline fractures and hopefully prevent further mosfet blowout and unit failure.

4. Purchase a unit from company which doesn't have such quality control issues. :drinkup:

[specter]

Well I was trying to save some money. And , all the review on the units forum looked great.

[kepe]

I totally understand where you are coming from, but was/is it worth the savings? Personally, I can't afford to buy a particular piece of machinery more than once.

[specter]

well your right!! All this going back and forth between manufacturers and dealing with defective products is a bit too much.

I have always bought brand name high end equipment before.

And like you say in the end you get what you pay for. Well I certainly got it in the end. ow

[Sandy]

A large portion of the R&D gets done by the end user which equals a big savings for the developing company. And if a lot of the little things just get tolerated, bigger savings yet.

[duaneb55]

So what you had there was a bunch of misfit MOSFET mounts.

Are you sure about that 8-10lb/ft of torque on those - what are they - #6 or #8 screws? Seems a little high for that size machine screw to take in the first place.:confused:

I sure hope the modification cures it for you and that Simon comes through with a healthy batch of consumables for your time and trouble. I was going to suggest a smokin' deal on your next purchase but I guess you already got that.:rolleyes:

[kepe]

spector, I am sorry this happened to you. Hope you get everything worked out sooner rather than later.

[jetenginedoctor]

How much are you out in terms of time and money to fix this stuff? Were you able to do this yourself, or did you end up having to pay somebody else? You mentioned a certain ISO certified tech, so it sounds like maybe it was somebody else (seeing as how you have a farm, I'm guessing it isn't you?) Maybe it IS you? I don't know why somebody would be a tech and have a farm, but I suppose it's possible? :) I know that'd be one busy dude!

BK

[specter]

Well its been since the beginning of Aug I am out more time than money since the item is to be used on my farm. I co-own a cnc shop but I do not stay there all the time. Most of the farm is fields for hay that I sell. What can I say the field grows you mow it then bale. The rest of the time on the farm is side jobs for friends.

As far as fixing these units I didn't!! Our shop tech help diagnose the problem but then I notified the company of the defective unit and they replaced it. Of course having the shop tech away from paying work does cost and after all this is said and done the only thing I seem to get from the manufacturer is more defective units. They do send extra consumsables with each defective unit. Yes I keep those each time I replace the unit. I am building up quite a collection of consumables!! At this rate I'll have a lifetime of consumables!! Just hope I eventually get a working unit to run them on.

[Mondo]

It appears to me the manufacturer is using the metal bars as supplemental heat sinks. Although the MOSFETs are installed with thermocoupling compound (you can see the white goo oozing from under the components in the photos) the engineers are still looking to sink off more heat. The metal bars then conduct that additional heat away from the components and to the aluminum heat sink through the screws. Using a plastic retention bar, or putting a plastic shim between the MOSFET and the retention bar is counter to that design intent. Using the mounting holes provided by the MOSFET manufacturer is probably inadequate to that end.

There are two solutions:
QA needs to enhance efforts on the factory floor to assure the screws are installed with the proper tools using maximum torque settings.
Engineering need to do a better job in the design of the product to assure they do not have this heat problem. There are some questions engineering needs to review and answer:

Why did they put that pink strip between the components and the aluminum heat sink to which the components are mounted? If the MOSFETs were mounted directly to the heat sink with only the white thermoucoupling compound (and the appropriate quantity, too!) between them and them sink, and each was individually screwed down as it is designed to be, they might not be having this problem. But there must be some reason for the pink shim.

You noted you had been haing similar problems with various manufacturers. What manufacturers? Are they all mounting the MOSFETs this way?

-Mondo

[specter]

Mondo You have a very valid point!!

Since each mosfet is securely fasten to the board by solder it does make sense to also have those metal (retention) bars also act as a heat sink. In which case putting the plastc between the moset and the bar would further retain heat. As for the pink plastic it too maybe hinder the cooling of the aluminum heat sink plate.

But like you say that has to be addressed by their engineerers.

As far as the manufacturers!! It appears on all the China made products.

[distrbd]

Specter,thank you for the info.
I'm expecting to receive a Chinese made tig welder soon and now worried that I'll have the same problem you described,how exactly do you recommend I tighten these bolts after installing the non conductive spacers?
I don't own a torque wrench but was wondering if there is any easier way to do this?
Thanks.

[specter]

My shop tech says the easiet way is to loosen up the screw. Next retighten the screw till it starts to feel tension, then turn the screw an additional 1/3 to 1/2 turn. That should get you close.

[distrbd]

Alright thank you.:drinkup:

[Fred Paul Jones]

Looks like the bars are a cheap way to isolate the fets that did not work.

[joebiplane]

Specter,

I, too, am waiting for a plasma cutter / tig welder and wonder if the MOSFET problem has been addressed in the new, updated machines just coming to Simon this week ?? Sure hope so.!
joe

[jimcolt]

Most US manufacturers of plasma cutting systems stopped using MOSFET devices for power switching years ago. This technology is prone to failure based on relatively high voltages used in plasma cutting systems as well as from input power line spikes.

Most industrial grade plasma systems today use IGBT switching devices (Insulated Gate Bipolar Transistors) which have higher speed switching capabilities as well as higher current and voltage tolerances...meaning better reliability.

If you open up a Hypertherm plasma....you will find IGBT technology!

Jim

[specter]

Well I agree with you on that on Jim. Our shop's Hypertherm has never given a problem of any kind. But then you get what you pay for High Quality.

[distrbd]

Specter, I don't know exactly what a "thermoplastic" is or where I can find it,I don't want to use any clear plastic I find around the house in case it would melt ,could you or anyone help me by directing me to a place where I could purchase some of this material?
It would be greatly appreciated.
Ken.

[buhler]

That pink pad is a thermal pad and an electrical isolation pad.

It electrically insulates the fets from the heatsink. The grease is a band-aid to improve thermal performance. There is an alternative product which is a hard plastic sheet coated with wax on both sides. The wax melts and gives a good thermal conductivity. The hard plastic is immune to push-thur or cut-thru problems with the fets on a soft (conformal) rubber pad.

The metal bars should be designed so they have a gentle dimple in the center of each fet. As it is now, the flat bar bends when the screw is tightened and it puts all the pressure on the side (edge) of the fet.

Even better if the flat bars were changed to a stiffer "C" channel with dimples over the centers of the fets.

A bellville washer or two should be put under the screw head to put constant pressure on the bar. The screw can then be tightened untill the belville washer is just NOT bottemed out.

Also in one of the pics the screw hole seems way too close to the right hand fet to meet clearance and creepage rules. The design needs a better way to register the pc board to the heatsink.

[distrbd]

Two ideas come to mind:
1)=the lens on a welding helmet.
2)=spray the metal bar with a heat resistant paint which is normally used on engines,Contains ceramic resins for maximum heat dissipation .
Resists temperatures up to 500° F.
Any thoughts?

[longevity-inc.com]

HI All,

The spray is a very interesting suggestion. We are looking at many alternatives and are testing some of Specter's suggestions to see what will be the best solution.

Please keep in mind that we are here to improve quality and are also testing mosfets to see what issues if any there are with the chips used at that time in the production line.

However, please understand that our return ratio is still very low for our company. We will be improving a lot of things here and we thank Tom and other for suggestions and input.

Simon

[specter]

Yep everyone has some good suggestions. The idea of using hheat dispating ceramic paint could be used on existing retention bar stocks till they are used up.

The idea of the plastic retention bars that shop came up with would have a several good advantages

cheaper to produce
the plastic bars would have some flex
the plastic bars are nonconductive

Of course in the meantime if the factory would decrease tension on the screws holding down the steel retention bars onto thge mosfets that should elinate the hairline cracks caused by overtorquing the screws.

[GiantTechGuy]

Are you sure those hairline cracks aren't a product of the mosfet explosion and not the retention bars overtorqued?

My email address.... lpb60@roadrunner.com

[specter]

In each case where the mosfet was blown we used a k-d quality control screw torque meter and found the screw averaged 6-8 foot pounds of torque which is far in excess of what it should be for a screw that size. Torgue should have been in inch pounds. In each case the mosfets under additional retenttion bars we found hairline cracks on the non burnt out/explosion mosfets

[GiantTechGuy]

Were those non-exploded mosfets with hairline in the same bank of mosfets as were the exploded ones?

[specter]

Yes they were on both side of the unit's upper level on both left & right sides.

add on I got the new replacement unit here with me in Ft Myes FL. I have have taken the units cover off and all wires were properly attached. It was missing theh grounding lug onb the back of the unit so I called Simon to address this. He said it was one of the newer units not requiring a separate grounding lug. I also asked about the modifications I brought up. He addressed the issue at the factory and their engineerers are looking into it. I addreess the nonconduuctive spacer issue and he saidadding them to the machine would not void the existing warranty. I checked all the mosfet retention bars and they are still overtorqued so much so that I stripped outy of of the phillips screw heads while loosing up the screws. I will add the nonconductive spacers before I try firing up the unit.

[GiantTechGuy]

Another thing you might want to check is the amp draw of the unit. In 220 volt mode while cutting with machine at full output see what it draws with an amprobe around one of the incoming hot legs of the power supply cord.

[GiantTechGuy]

Yes they were on both side of the unit's upper level on both left & right sides.

.


So the non-exploded mosfets with hairline cracks were next to exploded mosfets on each side of the machine?

[GiantTechGuy]

It was missing theh grounding lug onb the back of the unit so I called Simon to address this. He said it was one of the newer units not requiring a separate grounding lug. .

That grd lug was never needed if the machine had a 3 wire power cord which I am sure it did. I am told that in China they run an external ground to that lug from where I do not know. They told me that their power cords are 2 wire, cords only house the power wires no equipment grounding wires, in fact I have seen this myself on a sample order of machines I recieved last winter from a factory that normally only supplies the domestic market. I had to change all the power cords to 3 wire :laugh: .

[Capnbondo]

The bottom line here is that those MOSFET's were not designed to be mechanically retained by a metal bar strapped across the side of them.

They have a hole for an attachment screw on the end of them, and you can buy heatsinks designed specifically for that style of transistor, that attach via the existing mounting hole.

I would launch that whole piss-poor system into the garbage can and use proper heatsinks attached via the proper mounting hole. Use the product as it was intended and engineered to be used. Problem solved.

[specter]

Yes the hairline cracked mosfits were along side the burntout mosfets.

The power wire is a proper three wire setup. Two 120vdc lines and one grounding/neutral line.

I am not sure what the design engineers had in mind with this design but it appears to be the same design on all Chineese made plasma cutters.

[GiantTechGuy]

Yes the hairline cracked mosfits were along side the burntout mosfets.



OK, that is what I expected. Now let me tell you my thoughts. I have inspected literally hundreds of these machines, I do so as a regular practice before shipping out to my customers, and I have NEVER seen any hairline cracks except after a mosfet explosion. I am not saying the over torque will not cause a crack but only that I have not seen it. I think the crack you are seeing is from the exploding mosfet causing a shock to the retention bar or maybe just a shock wave to the surrounding area? When they explode it's a hella BANG!
My guess is that your problem may be from over current caused by the factory not dialing in the machine. I say this because last summer I experienced a nightmare of mosfet explosions from machines that came from the same factory as the ones yours are coming from. After several machines exploding during testing here at my shop and a few returning from customers I decided I had to find the reason. At 1st I thought it was from a new brand of mosfets the factory had installed without 1st approving it with me, that turned out to be wrong! What I found was that after placing an amprobe on the incoming ac line the current draw on some machines was in the high 30s to mid 40 amps range where it should have been no more than 28. Well that high current means higher heat output from certain devices within the machine, mosfets being one of them. Look here to see the effect of too much heat within a mosfet.... http://www.ipes.ethz.ch/ipes/2002thermal/runaway/runaway.html
So needless to say I test all machines for proper current draw. Also needless to say I do not purchase from that factory, Chiry,any longer and not just because of what I mentioned above but also because the quality of machine builds would vary so much from shipment to shipment and because of several other factors I wont get into here.
So to summarize my thoughts I think you're on the wrong track concerning the cause of your mosfet failures.
Hope this helps a bit. I am by no means an electronics expert but the above might mean something coming from someone who has more than average experience with these machines.

[GiantTechGuy]

Yes the hairline cracked mosfits were along side the burntout mosfets.

The power wire is a proper three wire setup. Two 120vdc lines and one grounding/neutral line.

I am not sure what the design engineers had in mind with this design but it appears to be the same design on all Chineese made plasma cutters.

I think you ment to write ac and that third wire is an equipment ground, never a neutral.

[duaneb55]

I don't mean to hijack the thread but how is all this free troubleshooting and R&D for Simon a topic for a 'Safety Forum'?:confused:

Just curious.

[GiantTechGuy]

I don't mean to hijack the thread but how is all this free troubleshooting and R&D for Simon a topic for a 'Safety Forum'?:confused:

Just curious.

I think it was placed here by mistake, should maybe have been placed in the electrical forum?
A better Chinese machine is good for all concerned! But I wont be giving away any of our up coming improvements :waving:

My email... lpb60@roadrunner.com

[GiantTechGuy]

Maybe I'd leave it here in the safey department, never power up your machines with the cover removed and your face near a mosfet :jester:



My email... lpb60@roadrunner.com

[duaneb55]

Maybe I'd leave it here in the safey department, never power up your machines with the cover removed and your face near a mosfet :jester:



My email... lpb60@roadrunner.com

There ya go! That works.;)

[specter]

Thanks for that info GiantTech,

Well at this point I am just crasping at straws trying to figure out what is causing the problem. I got the third unit but I haven't tried it out yet. I have taken the cover off and the unit and performed a visual inspection. Then I performed a check of the mosfet retention bar screws! Again the screws are tightened down far to tight. I stripped two of the phillps head with my screw driver just loosing the screws. I am pretty sure the tension on these screws does not have to be so high.

I also haven't tried my modification with the thermoplastic spacer between the mosfet and its retention bar. After reading about the high heat these mosfets generate it makes me wonder if inserting a plastic spacer would be a good idea. Sure the spacer would isolate the mosfet from the metal retention in the event of a short! But!! If the mosfets get that hot they may in fact melt the plastic and cause an additional problem.

You idea of an amperage surge does make sense as well!! But down here I do not have access to my shop tech up in NY and to all his testing equipment as well as his knowledge.

I will try hooking up the unit as normal and see what happens. Being the third unit if this happens again at a different location than the previous two units I can only assume it is defective mosfets. As the seller said this is an entirely new unit for them so it hasn't been proven yet I can only guess the design and the units components should have been tested better.

Tom

add on: Oh Yeah I do not plan to have my face anywhere near the unit when I turn it on. The previous two fireworks displays from exploding mosfets was impressive! Just glald there wass nothing flamable near it.

[GiantTechGuy]

If I were you I'd check all the mosfets now to see if you can spot any hairline cracks.
I am not so sure that the spacer idea is worth much anyway. Is that plastic spacer rated to insulate against the voltage present at that mosfet, if not that voltage may just travel right through that spacer anyway.

[GiantTechGuy]

Thanks for that info GiantTech,





add on: Oh Yeah I do not plan to have my face anywhere near the unit when I turn it on. The previous two fireworks displays from exploding mosfets was impressive! Just glald there wass nothing flamable near it.

DO NOT fire up the unit without the cover in place, just my advice;)

My email... lpb60@roadrunner.com

[GiantTechGuy]

Oh and also the non-exploded mosfets with hairline cracks could very well be exploded just not to the degree of the visually apparent ones.I think if you tested those you'll find they too will not test as usable any longer.

[specter]

I plan to examine all the mosfets for any crack before I fire it up. I just have to find a large enuf magifiying glass in my tool box.

[distrbd]

I found was that after placing an amprobe on the incoming ac line the current draw on some machines was in the high 30s to mid 40 amps range where it should have been no more than 28. Well that high current means higher heat output from
So to summarize my thoughts I think you're on the wrong track concerning the cause of your mosfet failures.
.
So from what I understand if my inverter Mosfets happens to draw more than 28 amps(or more than what are designed to draw) they will explode regardless of how much those holding bars are overtightened.
I'm keeping my fingers crossed that they(Mosfets) will perform properly so I don't have to fiddle with them and/or plastic spacers,heat resistant paint, etc.
Thank you and Spector and others for your posts in a very interesting thread.

[GiantTechGuy]

So from what I understand if my inverter Mosfets happens to draw more than 28 amps(or more than what are designed to draw) they will explode regardless of how much those holding bars are overtightened.
I'm keeping my fingers crossed that they(Mosfets) will perform properly so I don't have to fiddle with them and/or plastic spacers,heat resistant paint, etc.
Thank you and Spector and others for your posts in a very interesting thread.


It's like Jim Colt said, there are also other factors that might contribute to their failure.

[specter]

Third failed 80 amp plasma cutter.
Fourth failed Chinese unit.
Problems occured with units set at 80amps before this time it happened at 40amps. On the fourth unit it failed at max setting of 50amp. So in three out of four units they failed when used on their max setting. In everycase it was a failure of the mosfets when they exploded and always on the upper board of the unit.

Of the four units three failed on initial powerup and start of arc. One of the four units did make it to four use and did make some impressive cuts in 1.25" hardened alloy dragster axle before that unit also failed.

I really have to believe that its a design fault! Either improper torquing of the mosfet retention bars or faulty mosfets.

Several folks have said that mosfets were a old technolgy and do not hold up well on higher amp plasma cutters. Some of you say that IGBT Technolgy is the way to go in the higher amp cutters.

Today I got to go to Airgas to pickup a bottle of argon for my lincoln mig. While there they had several manfacturer's units open for display and actual hands on testing. Hypertherm, ThermalDynamics, Miller and Lincoln were present.

The power of the Microprocessor IGBT Technolgy versus the older Mosfet Technology was the clear winner overall for fast consistant powerful clean looking cuts.

[GiantTechGuy]

Did you perform a current reading like I suggested?

For your problems please don't place so much blame on mosfet technology. I have several cut80 machines that have been out for over a year and a half without a problem. I think it's more than likely a problem with the circuit desgin by the factory your machines came from.

[duaneb55]

GTGuy,
So are you suggesting a 50% failure rate is good or acceptable?

I'm still trying to figure out why this thread is in the Safety Forum though.:confused:

[GiantTechGuy]

GTGuy,
So are you suggesting a 50% failure rate is good or acceptable?

I'm still trying to figure out why this thread is in the Safety Forum though.:confused:


I don't understand where you are getting that impression from?
Of course I don't think it's ok.
Again, for safety don't stick your head near one of these exploding mosfets :drinkup:

[GiantTechGuy]

GTGuy,
So are you suggesting a 50% failure rate is good or acceptable?

I'm still trying to figure out why this thread is in the Safety Forum though.:confused:

Oh I see where you got it from... let me rephrase that....


Did you perform a current reading like I suggested?

For your problems please don't place so much blame on mosfet technology. I have several cut80 machines that have been out for over one and a half years without a problem. I think it's more than likely a problem with the circuit desgin by the factory your machines came from.

:cool2:

[specter]

I am pretty sure that giantechguy isn't suggesting a 50% failure rate is acceptable. Seems he is saying failures do happen and in these units it appears to be design related.

I did not get to perform the test that gianttechguy suggested. I was on the road in my rv heading back to my house in FL (yep I'm a disabled vet and a snow bird). Down here I do not have the shop, experienced help and testing equipment I had access to up north.

I spoke to other folks out there that have had outstanding performance of their units. The one unit I had received that made it to four uses was truely impressive for the price. & cutting abilities But like many said at what point to you cut your losses and look elsewhere?

The thread was posted here to point out a potential fire hazzard in addition to the failure of the units. As I said when the units failed they put on a fireworks display even with the units casing installed. Had those units been near flamable materials or explosive gas(es) I might not be here! So I posted this thread for others with similiar units to watch out for this potential hazard.

[specter]

Ok here are the pics of the latest unit.

Always the top circuit board of the unit seems to be the cause. Anyone have some suggestions.

[kepe]

Yep, send it back and buy a proven unit. You have spent WAY too much time dealing with this and it's just not worth it.

[tx_swordguy]

A couple of questions. 1st let me say I just bought a CT520d from Giant Tech and am absolutely 100% satisfied with the machine. The machine has performed flawlessly in all aspects and I have run plasma (just sheet metal, but ran it up to 50amps).stick, and tig. It is one fantastic machine. With that being said, I am a little on the worried side about the mosfet problems being discussed here. Q1. We all know that Specter has had problems with all of his machines (4 stand alone plasmas I believe). How many OTHERS have had similar issues and was anything in common (length,time running,thickness of cut etc)? I suspect the people that have had good luck with their machines will not respond to this. Q2 If for some reason I do have a "fet" blowout down the road. Is it possible to replace the Mosfets with IGBT chips? If so what type would I need? I realize you warranty the machine Larry, just trying to cover any issues that may occur after the warranty runs out. I guess this maybe aime more at you Larry because you really seem to know your way around these machines.
Thanks Mark

[GiantTechGuy]

The machines I had from last summer blowing mosfets came from my old supplier, Chiry. None of them were tuned at the factory, all were drawing way too much current. I do not purchase from Chiry any longer. I purchase from a new up start factory and although I have had some issues they have been very willing to work together with me to improve.
The thing I have found with some of these Chinese factories is that they are all to happy to throw the latest and greatest new welding or cutting machines out into the market. Most if not all of these machines spend little to no time in any field testing prior to release into the market place, one example is the Super200P series and there are plenty of horror stories about that machine to read on the net :rolleyes:.
I no not know anything about IGBT other than what I have read, never seen one. My guess is that it would not be some easy retrofit.

[specter]

Hi GiantTechGuy, Kepe and tx_swordguy,
MagnetMechanic just posted a reply in the Mosfet vs IGBT Technology thread. From the pics posted along with his desription it sure sounds alot like the problems I am having. His pics show that he must have had a bigger fireworks display than what I had.
I wish that I had been able to check that amperage. on those mosfets. It must be very high if it can burn/melt those metal bars holding those mosfets. In his case it also appeared to be circuit board problems as well.

Makes you wonder if they all get their circuit boards and mosfets from the same supplier.

Good thing is that the manufacturer sent me a prepaid dhl shipping label to return the defective unit.

I did however have to sign one of their return policy affidavits but I guess this is also a industry wide practice now when asking for a refund.

I added a note to the addidavit

Dear Longevity:


I, (Name) ____________________________the purchaser in regards to the Longevity plasma cutter/welder order am returning the unit:
_____80amp plasma cutter __________. I am asking Longevity to ship back the unit under their return policy. The unit was not damaged at the time of delivery and is being returned in original condition with outer casing and all internal components properly attached. I am asking Longevity to ship the unit back for me. Although I understand this affidavit was prepared by Longevity, by signing this form, I certify these statements to be true.

Sorry for the troubles.

Thank You,

(Name):
(Signature):

Please provide copy of your driver license and fax back to 1-510-887-7091

Note: Unit is being returned as non functional. The main upper circuit board was damaged upon powering up the unit and is being returned per warranty return policy.
And a full refund issued.

Consider me paranoid but after all these problems I just wanted to protect myself since that is a signed affidavit. Oh I did not include a copy of my drivers license.

DHL will be picking up the unit tommorrow.
__________________
Own farm:Harris oxyacy torch and lincoln 225s stick, Longevity ForceCut 80amp Plama Cutter
Co-Own CNC shop with various toys such as:Miller :1251 plasma cutter, Syncrowave 350 LX TIG, & XMT 456 Multiprocess Welder.Hypertherm HPR260 Plasma Cutter

[awright]

Mondo is undoubtedly correct about the metal clamping bar acting as a supplementary (but not very effective) heat sink. One way to spread the clamping force uniformly over the contact area, provide thermal conductivity, AND provide electrical insulation is to use thermally conductive elastomeric pads between the bar and the plastic MOSFET package. This thin, gray, rubber-like material is sometimes used between power transistor cases and heat sinks in lieu of gooey heat sink compound in cases where electrical insulation of the transistor from the heat sink is required. The elastomeric pads do result in greater thermal resistance than direct mounting with heat sink compound. This could be partially compensated for by substituting small finned extruded aluminum heat sinks for the steel bars. Deep fins (where space permits) could recover the stiffness lost by substituting aluminum for steel. Stiffness is necessary in the clamping bars to achieve uniform pressure distribution across the case.

Clamping force can be easily regulated by use of calibrated "spring washers," such as wavy spring steel washers or thin Belleville spring washers (see http://en.wikipedia.org/wiki/Belleville_washer or http://www.internationalsprings.com/wave_spring_washers.htm). Such washers can provide predetermined clamping force by being deflected by the clamping screw until they are just flattened or until a specific thickness of shim stock (as in a feeler gauge) can be just inserted under the edge of the washer.

Has anybody ever actually observed cracked MOSFET cases in welders BEFORE a failure occurs, rather than after? Unlikely, I think, because few people would be motivated to disassemble a new welder. Thermal failure of plastic packaged semiconductors due to simple overheating is likely to lead to cracking of the plastic case even when the failure was not induced by a cracked case. My question is, after a dramatic thermal failure, how does the observer making the autopsy determine if the observed cracks caused the failure or resulted from the failure? It could just as well be a case of inadequate thermal design leading to failure in the absence of a clamping problem.

awright