EEVblog Electronics Community Forum

Electronics => Repair => Topic started by: iamdarkyoshi on September 17, 2017, 01:29:37 am

Title: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 17, 2017, 01:29:37 am
So I bought a cheap 150w electronic load from ebay. The listing said 150w (yeah not with the stock heatsink), though with the massive CPU cooler I stuck on it, I had no issues dissipating 200W from a 19V laptop adapter. Or so I thought. After a couple hours of this, the mosfet just shorted.

Its a single IRFP250N, which is rated for about 200W disspation at 25C in the datasheet IIRC. However, mine was running at about 65C package temp.

I wouldn't be surprised if they put a knockoff mosfet in here, but I bought some replacements from a reputable source, but before I put one back in, I'd like to know what went wrong.

Any ideas? I used high quality thermal paste and everything when I put the CPU cooler on it, so its not that I had inadequate cooling.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: nidlaX on September 17, 2017, 01:41:11 am
This type of "150W" electronic load has been discussed before on the forum. There is variability in circuit design, component selection, and quality control. You should post a few pictures of your particular variant.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 17, 2017, 01:45:46 am
This type of "150W" electronic load has been discussed before on the forum. There is variability in circuit design, component selection, and quality control. You should post a few pictures of your particular variant.

This is what my unit looks like after a cooler upgrade.
(https://i.imgur.com/TBMgmmkr.jpg)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: nidlaX on September 17, 2017, 01:46:57 am
This type of "150W" electronic load has been discussed before on the forum. There is variability in circuit design, component selection, and quality control. You should post a few pictures of your particular variant.

This is what my unit looks like after a cooler upgrade.
(https://i.imgur.com/TBMgmmkr.jpg)
Pictures of the PCB and circuit so we can actually see which variant you're using...
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 17, 2017, 01:52:35 am
This type of "150W" electronic load has been discussed before on the forum. There is variability in circuit design, component selection, and quality control. You should post a few pictures of your particular variant.

This is what my unit looks like after a cooler upgrade.
-snip-
Pictures of the PCB and circuit so we can actually see which variant you're using...

I've made some small modifiations to mine.
(https://i.imgur.com/bwR0sj0r.jpg)
(https://i.imgur.com/dCVkw9rr.jpg)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: najrao on September 17, 2017, 02:20:57 am
You must be confused with the method of rating power mosfets: "200W@25C case temperature", the rating assigned, is just an extrapolated value, and applies if the user somehow managed to hold the metal base temperature at 25C. At 65C cooler temperature, the centre of the mosfet's base will be even hotter, may be 85C. It needs a correctly attached thermocouple  to check this. And at this 85C base, the chip inside will reach 150C for a power dissipation of just 100W!, not 200.That it survived for a time at 200W demonstrates a margin of performance not meant to be exploited by the user.
Though you might get a 400W@25C mosfet in the same case, it is best not to push temperatures so far. I would use THREE IRF250's in parallel, and hold their bases at 65C by the heat sink. Obviously,  the mosfet's should attach to the heat sink directly i.e., without any shims or spacers.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 17, 2017, 03:27:50 am
The cooler looks huge but
It is definitely necessary to know the reliable published thermal resistance performance rating of the cooler and not just the huge looking size.
By the look of it, the plates look very thin, and the region where the tube connects to the plates [compression flare] looks ..... ehm... not too promising.
Then you wonder whether it is a genuine Heat Pipe technology, and the preferred orientation of the installation for the migration and return of the flow of the refrigerant inside the tube.
So, I would say you firstly need to calculate the required thermal resistance performance of the heat-sink before you buy one.

Edit: Rated in the market as useless stuff, do not buy it. Further info on fake can be found here

https://www.eevblog.com/forum/repair/part-identification-150w-constant-current-electronic-load-60v-10a/ (https://www.eevblog.com/forum/repair/part-identification-150w-constant-current-electronic-load-60v-10a/)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: IanMacdonald on September 17, 2017, 06:16:17 am
Thermal resistance and heat dissipation calculations might be a good subject for a video. It's not that difficult a subject, but I've met plenty graduates who didn't understand it. Basically just Ohm's law applied to heat.

If the device's rating is quoted at 25C and it's running maxed-out with  the case at 65C then it's running well outside of its rating, because if the chip has a max temp of 175C, then it will actually be at 215C.

https://www.infineon.com/dgdl/irfp250n.pdf (https://www.infineon.com/dgdl/irfp250n.pdf)

Though, the 214W claim seems to be a bit of an exaggeration since the thermal resistance is quoted at 0.7C/W, with an additional 0.24C/W for mounting. Thus the actual max power is 159W, even with the mounting at 25C.

If you were to allow a max heatsink temp of 65C, and assuming a direct mount with no washer, then the rating is 117W.

So, this is partly a case of not understanding ratings, but also partly a case of manufacturers making bullshit claims, there being no way that a 214W rating could be achieved in practice, short of refrigeration.   
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: najrao on September 17, 2017, 09:03:28 am
NidlaX's comment on the manufacturer's claim being b-s,- is misplaced to the extent that the 214W rating assigned applies at 25C BASE temperature (actually specified and measured at an exact location directly under the mosfet die inside the package); the manufacturer has no control over the extra thermal resistance between the base and the heatsink, as much depends on the quality of the surfaces and their coupling through air/thermal grease and  the clamping pressure. Thus, the 214W results from 175-25=150÷0.7=214. It only needs elementary application of mind to understand this well standardized rating method, and to apply all the 'deratings' needed in a practical application.
The finned heatpipe cooler is, again, a highly optimized design; there is nothing wrong with the fins being "too thin" or "too closely spaced", as the heatpipe carries the heat with little temperature gradient right to the centre of each fin, and as a powerful draft is set up by the fan.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 17, 2017, 09:17:07 am
The finned heatpipe cooler is, again, a highly optimized design; there is nothing wrong with the fins being "too thin" or "too closely spaced", as the heatpipe carries the heat with little temperature gradient right to the centre of each fin, and as a powerful draft is set up by the fan.

Your statement implies All finned heatpipe coolers are constructed same regardless from china, usa, or from reputable manufacturer or from street side manufacturer?
Be real, ask for the certificate.
Tell us the exact liquid and the material inside the tube and of this particular cooler in question, since you know all finned heatpipe cooler.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 17, 2017, 09:56:28 am
You're also using what looks like a switching specified mosfet in its linear region (I assume the gate drive is linear rather than PWM- I don't see any inductors).

Most switching mosfets get unhappy if used in their linear region and can fail due to localised hot-spots on the die. It's normal to significantly de-rate them for linear operation, regardless of heatsinking.

200W is an awful lot of dissipation for one package anyway. I would look at parallel operation across a couple of FETs. Such a massive heatsink on a single package just looks 'wrong'.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: mikeselectricstuff on September 17, 2017, 10:00:47 am
Thermal resistance and heat dissipation calculations might be a good subject for a video.
https://www.youtube.com/watch?v=8ruFVmxf0zs (https://www.youtube.com/watch?v=8ruFVmxf0zs)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 17, 2017, 10:11:08 am

 there is nothing wrong with the fins being "too thin" or "too closely spaced",

Elementary question;?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: najrao on September 19, 2017, 05:01:50 am
Armadillo, I do not claim to be the sole depository of all knowledge on this subject. Nor that there are no differences between 'reputable- read American' manufacturers and 'street side vendors'.
Other things being equal, thick fins conduct heat to their extremities with less temp drop, so should cool better. By how much? In engineering practice, it is not uncommon to compromise when the returns start dropping off. It is called optimisation, just in case.
The most common liquid - and the most appropriate at that - in low temp heat pipes is demin water. There is no way of telling what it is in a specific model.
Thank you for your comment.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Assafl on September 19, 2017, 06:02:30 am

 there is nothing wrong with the fins being "too thin" or "too closely spaced",

Elementary question;?

Interesting quote. I wonder why A/C manufacturers bent on getting an A energy rating use a large amount of pipe, and lots and lots of thin fins.

My ongoing assumption is that indeed it is the amount of pipe + square section of fins that affect cooling performance. However, you'd also add to that the fins access to convective cooling "air" (which would want a larger surface area).

Maybe in Aerospace, the fins only radiate energy (IR) and that requires less fins?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 07:28:10 am
I wonder why A/C manufacturers bent on getting an A energy rating use a large amount of pipe, and lots and lots of thin fins.

Number of fins per inch?

How many FPI does the cooler in question has?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Simon on September 19, 2017, 11:54:06 am
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 12:21:23 pm
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.

I did say HeatPipe Technology, and you are referring to gaseous phase, but heat exchanging should be the same as m, c, delta T.
The pertinent matter would be the actual published Thermal Heat Resistance of the said cooler other than contorting into abyss.
Cool!
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 12:50:52 pm
Armadillo, I do not claim to be the sole depository of all knowledge on this subject. Nor that there are no differences between 'reputable- read American' manufacturers and 'street side vendors'.
Other things being equal, thick fins conduct heat to their extremities with less temp drop, so should cool better. By how much? In engineering practice, it is not uncommon to compromise when the returns start dropping off. It is called optimisation, just in case.
The most common liquid - and the most appropriate at that - in low temp heat pipes is demin water. There is no way of telling what it is in a specific model.
Thank you for your comment.

How can you assert that the specific model cooler had been "Optimised" when there are so many unknown to you particularly the heat pipes running on "demin water" is kind of a joke, don't you think so? If you truly do not claim to be the sole depository of all the knowledge, would you not faults others comment like you had own the knowledge of the subject?


NidlaX's comment on the manufacturer's claim being b-s,- is misplaced to the extent that the 214W rating assigned applies at 25C BASE temperature (actually specified and measured at an exact location directly under the mosfet die inside the package); the manufacturer has no control over the extra thermal resistance between the base and the heatsink, as much depends on the quality of the surfaces and their coupling through air/thermal grease and  the clamping pressure. Thus, the 214W results from 175-25=150÷0.7=214. It only needs elementary application of mind to understand this well standardized rating method, and to apply all the 'deratings' needed in a practical application.
The finned heatpipe cooler is, again, a highly optimized design; there is nothing wrong with the fins being "too thin" or "too closely spaced", as the heatpipe carries the heat with little temperature gradient right to the centre of each fin, and as a powerful draft is set up by the fan.

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Karlo_Moharic on September 19, 2017, 04:13:14 pm
Quick question , why the hell don't just buy a decent DC load , instead of cheap crap off of ebay?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 04:25:01 pm
Can the heatsink sit flush to the mosfet.

And also the current sense wire, is it the original wire material? What material wire is used if not.

Just curious.

Thanks.

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Simon on September 19, 2017, 05:07:40 pm
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.

I did say HeatPipe Technology, and you are referring to gaseous phase, but heat exchanging should be the same as m, c, delta T.
The pertinent matter would be the actual published Thermal Heat Resistance of the said cooler other than contorting into abyss.
Cool!

Believe it or not your not the only one talking on this forum! a comparison was made to AC heat exchangers.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 05:30:29 pm
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.

I did say HeatPipe Technology, and you are referring to gaseous phase, but heat exchanging should be the same as m, c, delta T.
The pertinent matter would be the actual published Thermal Heat Resistance of the said cooler other than contorting into abyss.
Cool!

Believe it or not your not the only one talking on this forum! a comparison was made to AC heat exchangers.

Now, you weigh on something irrelevant and you ought to substantiate the difference.

Well, can you show or demonstrate that the mathematical formulae are much different between the cooler and the AC heat exchangers regardless whether it is mentioned or not mentioned. By the way, in the HVAC industry, they are also using the heatpipe system as part of energy conservation particularly in Fresh Air treatment in industry like wafer fabrication foundries, cleanroom, data center or what have you not.

So my question is WHAT IS YOUR POINT?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Simon on September 19, 2017, 05:33:00 pm
I suggest you read the whole topic.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 05:49:28 pm
I suggest you read the whole topic.

I suggest you do too.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 19, 2017, 06:26:39 pm
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.
(http://www.heatpipe.nl/images/heatpipe.jpg)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 19, 2017, 06:37:48 pm
By the look of it, the plates look very thin,
All CPU coolers with heat pipes have thin plates, show at least one with thick plates to prove opposite.
Quote
And the region where the tube connects to the plates [compression flare] looks ..... ehm... not too promising.
All coolers with aluminium plates are made like this.
Quote
Then you wonder whether it is a genuine Heat Pipe technology, and the preferred orientation of the installation for the migration and return of the flow of the refrigerant inside the tube.
So, I would say you firstly need to calculate the required thermal resistance performance of the heat-sink before you buy one.
There are no non genuine heat pipes, it simply won't work at all. Such cooler won't be able to dissipate any significant heat.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 06:42:52 pm
Hahahah, looks like the manifold disappeared.

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 19, 2017, 06:46:55 pm
LOL, you compare photos taken from different angle. It would be really surprising if aluminium plates became transparent.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 06:55:41 pm
By the look of it, the plates look very thin,
All CPU coolers with heat pipes have thin plates, show at least one with thick plates to prove opposite.
Quote
And the region where the tube connects to the plates [compression flare] looks ..... ehm... not too promising.
All coolers with aluminium plates are made like this.
Quote
Then you wonder whether it is a genuine Heat Pipe technology, and the preferred orientation of the installation for the migration and return of the flow of the refrigerant inside the tube.
So, I would say you firstly need to calculate the required thermal resistance performance of the heat-sink before you buy one.
There are no non genuine heat pipes, it simply won't work at all. Such cooler won't be able to dissipate any significant heat.

The context was for the 200W DC load mosfet and NOT CPU, the fins look thin in terms of thin and fews for the load in question. It meant, If the fins is so widely spaced, it should be thicker for the loads.

@All coolers with aluminium plates are made like this. - hahahahahahah a piece of joke. So we should trust and buy without a doubt based on your assertion?

@There are no non genuine heat pipes, it simply won't work at all. Such cooler won't be able to dissipate any significant heat. - That may be why the mosfet blew in the first place. Hahahah LOL.  :-DD





Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 06:57:19 pm
LOL, you compare photos taken from different angle. It would be really surprising if aluminium plates became transparent.

LOL ok you had your point.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 07:13:48 pm
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.
(http://www.heatpipe.nl/images/heatpipe.jpg)



Then you wonder whether it is a genuine Heat Pipe technology, and the preferred orientation of the installation for the migration and return of the flow of the refrigerant inside the tube.

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 19, 2017, 07:15:15 pm
The context was for the 200W DC load mosfet and NOT CPU, the fins look thin in terms of thin and fews for the load in question. It meant, If the fins is so widely spaced, it should be thicker for the loads.
It would not help, thickness won't improve heat dissipation significantly, more dissipation area is needed for that. Fins are spaced normally. Closely spaced fins obstruct airflow and collect a lot of dust very fast. I did not say this heatsink is adequate for 200W heat dissipation, most likely it isn't.
Quote
Such cooler won't be able to dissipate any significant heat. - That may be why the mosfet blew in the first place.
MOSFET was stressed a lot beyond absolute max spec regardless if cooler was good or bad. And by significant heat I mean that if heat pipes were fake, such cooler won't be able to dissipate even 20W.
Quote
@All coolers with aluminium plates are made like this. - hahahahahahah a piece of joke. So we should trust and buy without a doubt based on your assertion?
Try soldering aluminium. Copper plates can be either compressed or soldered.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 07:40:31 pm
The context was for the 200W DC load mosfet and NOT CPU, the fins look thin in terms of thin and fews for the load in question. It meant, If the fins is so widely spaced, it should be thicker for the loads.
It would not help, thickness won't improve heat dissipation significantly, more dissipation area is needed for that. Fins are spaced normally. Closely spaced fins obstruct airflow and collect a lot of dust very fast. I did not say this heatsink is adequate for 200W heat dissipation, most likely it isn't.
Quote
Such cooler won't be able to dissipate any significant heat. - That may be why the mosfet blew in the first place.
MOSFET was stressed a lot beyond absolute max spec regardless if cooler was good or bad. And by significant heat I mean that if heat pipes were fake, such cooler won't be able to dissipate even 20W.
Quote
@All coolers with aluminium plates are made like this. - hahahahahahah a piece of joke. So we should trust and buy without a doubt based on your assertion?
Try soldering aluminium. Copper plates can be either compressed or soldered.

Wraper to cut the story short - I was asking for the certified published thermal resistance value for this cooler. Do you  think is necessary to calculate it first before buying or using the cooler? Bear in mind, if you have not read it all, that the cooler is not the original cooler for the DC load.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 08:04:16 pm

It would not help, thickness won't improve heat dissipation significantly, more dissipation area is needed for that. Fins are spaced normally. Closely spaced fins obstruct airflow and collect a lot of dust very fast.

Curious, Have you open up a proper DC load instrument before and appreciate the heatsink inside.?


MOSFET was stressed a lot beyond absolute max spec regardless if cooler was good or bad. And by significant heat I mean that if heat pipes were fake, such cooler won't be able to dissipate even 20W.

You calculated? can it be 20.1W .... hahahahah LOL.



Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Monkeh on September 19, 2017, 08:13:53 pm
Wraper to cut the story short - I was asking for the certified published thermal resistance value for this cooler.

And just where do you think you'll find that for an aftermarket PC heatsink?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 09:17:49 pm
Wraper to cut the story short - I was asking for the certified published thermal resistance value for this cooler.

And just where do you think you'll find that for an aftermarket PC heatsink?

You should answer the question if its important to know the value prior to the application in the first place and not a matter of excusing where to find it. That should beget the subject matter in question, don't you think so?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 19, 2017, 09:22:25 pm
I see the OP hasn't bothered to come back - probably can't find anything of interest in the last page or so!  :palm:
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 19, 2017, 09:26:33 pm
Hahahah, looks like the manifold disappeared.

This mounting somehow don't look right. Has the cooler been modified?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 19, 2017, 09:56:49 pm

It would not help, thickness won't improve heat dissipation significantly, more dissipation area is needed for that. Fins are spaced normally. Closely spaced fins obstruct airflow and collect a lot of dust very fast.

Curious, Have you open up a proper DC load instrument before and appreciate the heatsink inside.?
Big aluminium heatsink(s) with a lot of mosfets mounted on it(them). No heatpipes because generally this is not needed as heat is evenly distribuited across the heatatsink by many mosfets. Fins are significantly thicker as distance heat needs to travel within fin is much longer. Do you even understand why thin plates are used in heatsink with heatpipes. Heat is delivered directly to the place where it is dissipated. If thicker fins will will be used, it will result in lower count of them in the same volume thus reducing performance. Can you explain by your wonderful theory why best CPU heatsinks like $90 Noctua NH D14 still use the same thin fins compressed onto the heatpipes? Also don't forget that CPU heatsinks are optimized for slow low noise fans, thus a lot of surface area is more important than thermal conductivity of the fins. Too low distance between fins besides dust accumulation cannot be used because fan at slow speeds don't produce high static pressure required to push air through.

(http://noctua.at/media/catalog/product/cache/2/image/9df78eab33525d08d6e5fb8d27136e95/n/o/noctua_nh_d14_1.jpg)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 19, 2017, 10:24:15 pm
BTW surprise, surprise. This is a heatsink which was cooling pentium 4 in old Dell PC, turned out I had one lying around. And fins are actually soldered to heatpipes. Also I stand corrected, this is the first time I see manufacturer bothered with trouble soldering aluminium to the heatpipes.

(https://www.eevblog.com/forum/repair/blew-a-mosfet-dissipating-200w-fake-mosfet-or-bad-circuit/?action=dlattach;attach=352632;image)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 20, 2017, 03:23:02 am
Here is the List of CPU thermal designed power from https://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures (https://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures)

If the cooler was designed for it, would it be safe to assume those figures relatively speaking?

The next interesting things to know would be;

Do the heatpipe CPU coolers fail? How long will it last.?

How to test it?     http://www.dansdata.com/coolercomp.htm (http://www.dansdata.com/coolercomp.htm)

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 20, 2017, 06:53:24 am
Big aluminium heatsink(s) with a lot of mosfets mounted on it(them). No heatpipes because generally this is not needed as heat is evenly distribuited across the heatatsink by many mosfets. Fins are significantly thicker as distance heat needs to travel within fin is much longer. Do you even understand why thin plates are used in heatsink with heatpipes. Heat is delivered directly to the place where it is dissipated. If thicker fins will will be used, it will result in lower count of them in the same volume thus reducing performance. Can you explain by your wonderful theory why best CPU heatsinks like $90 Noctua NH D14 still use the same thin fins compressed onto the heatpipes? Also don't forget that CPU heatsinks are optimized for slow low noise fans, thus a lot of surface area is more important than thermal conductivity of the fins. Too low distance between fins besides dust accumulation cannot be used because fan at slow speeds don't produce high static pressure required to push air through.

(http://noctua.at/media/catalog/product/cache/2/image/9df78eab33525d08d6e5fb8d27136e95/n/o/noctua_nh_d14_1.jpg)

Elementary maths question?.

I think - The rate of heat input governs the thickness of the heat sink core and fins. Attached calculator for your learning.

The thicker the fins, the better the thermal heat resistance.

http://www.myheatsinks.com/calculate/thermal-resistance-plate-fin/ (http://www.myheatsinks.com/calculate/thermal-resistance-plate-fin/)

To understand the problems involved in higher load cpu cooler, read this; http://www.powerelectronics.com/content/heat-pipe-reliability-high-power-applications (http://www.powerelectronics.com/content/heat-pipe-reliability-high-power-applications)

For DC Load Bank use custom heatsink, period.

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: NANDBlog on September 20, 2017, 09:07:57 am
Here is the List of CPU thermal designed power from https://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures (https://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures)

If the cooler was designed for it, would it be safe to assume those figures relatively speaking?
These are the ghetto pentium 4 CPUs for poor people. Pentium 4 went up to some 130W, and they were easily dissipating above their TDP limit.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 20, 2017, 09:25:30 am
Here is the List of CPU thermal designed power from https://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures (https://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures)

If the cooler was designed for it, would it be safe to assume those figures relatively speaking?
These are the ghetto pentium 4 CPUs for poor people. Pentium 4 went up to some 130W, and they were easily dissipating above their TDP limit.

Ghetto Cooler for Ghetto CPU, the subject in question.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 20, 2017, 11:12:32 am
Elementary maths question?.

I think - The rate of heat input governs the thickness of the heat sink core and fins. Attached calculator for your learning.

The thicker the fins, the better the thermal heat resistance.

http://www.myheatsinks.com/calculate/thermal-resistance-plate-fin/ (http://www.myheatsinks.com/calculate/thermal-resistance-plate-fin/)
Educate yourself. Fin thickness of this heatsink is 0.4mm, fin pitch 2 mm, distance heatpipe-heatsink center 25mm. I used this calculator, using this heatsink dimensions 105x88x55 mm (WxHxL). increasing fin thickness in this calculator from 0.4mm to 0.8mm increases pressure drop by around 3 times but reduces thermal resistance by 20% at the same airflow. The issue is, because of the pressure drop you won't be able to push the same airflow through it. All of this calculated for usual heatsink with 20mm base thickness, and not considering heatpipes which would make thermal resistance difference even less.
EDIT: if calculation is modified to use 25mm fin length to take heatpipes into account, when going from 0.4mm fin thickness to 0.8mm, thermal resistance decreases by only around 3-5% depending on airflow, yet again with drastic 3x pressure drop increase.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 20, 2017, 01:04:58 pm

Educate yourself. Fin thickness of this heatsink is 0.4mm, fin pitch 2 mm, distance heatpipe-heatsink center 25mm. I used this calculator, using this heatsink dimensions 105x88x55 mm (WxHxL). increasing fin thickness in this calculator from 0.4mm to 0.8mm increases pressure drop by around 3 times but reduces thermal resistance by 20% at the same airflow. The issue is, because of the pressure drop you won't be able to push the same airflow through it. All of this calculated for usual heatsink with 20mm base thickness, and not considering heatpipes which would make thermal resistance difference even less.
EDIT: if calculation is modified to use 25mm fin length to take heatpipes into account, when going from 0.4mm fin thickness to 0.8mm, thermal resistance decreases by only around 3-5% depending on airflow, yet again with drastic 3x pressure drop increase.

Wraper says and wraper thinks but not a pinch of technical reference to support your warp logics... hahahahaha LOL.
At least those references provided by me gave you an insight on how the thicknesses of the fins are selected, not just always thin, thin and thin. And when the next time you open up that commercial DC load banks, you can appreciate more on how designers do it the right way. In future, if your logic is too wayward, and out of context, I may not consider to reply.

To sum it, the cooler used may not be sufficient to run the mosfet at 200 watts. It may be over-drived but further detailed tests and calculation may need to be done to support this. Hence custom designed heatsink or the manufacturer recommended heatsink will be more appropriate to be used for DC load banks, safely. This does not exclude other possible causes for the blown mosfet.    ;D
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on September 20, 2017, 01:35:34 pm
Wraper says and wraper thinks but not a pinch of technical reference to support your warp logics... hahahahaha LOL.
LOL, you posted a link to calculator, I made calculations. You call it warping logic  :palm: There is a compromise between surface area / fin count and fin thickness to achieve least thermal resistance in particular volume. If you want to increase fin thickness, you must decrease fin count  :palm:. Less fin count, less dissipating area. As with heat pipes distance heat need to travel through the fin is low, low fin thickness is optimal. All heatsink manufacturers must be stupid, Armadillo is smarter than all of them. Also in electronic loads there in no such size, noise, small heat source constrain, therefore cheaper heatsinks with larger size, thicker fins and larger gaps are used. Also such heatsinks are usually extruded, and you cannot extrude 0.4mm fins even if you want.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 20, 2017, 10:33:18 pm
Can the heatsink sit flush to the mosfet.

And also the current sense wire, is it the original wire material? What material wire is used if not.

Just curious.

Thanks.

Its the original shunt. Anyway, I've been away and this topic kinda exploded. As suggested earlier, I'm going to parallel mosfets for better heat dissipation. I can fit four on this particular cooler.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 21, 2017, 08:58:20 am

Its the original shunt. Anyway, I've been away and this topic kinda exploded. As suggested earlier, I'm going to parallel mosfets for better heat dissipation. I can fit four on this particular cooler.

I see that you like to experiment. So how about progress to 400Watts 1KW Peak?

https://www.youtube.com/watch?v=WUPrj03UbTM (https://www.youtube.com/watch?v=WUPrj03UbTM)

https://www.youtube.com/watch?v=Vr6YsT403DM (https://www.youtube.com/watch?v=Vr6YsT403DM)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 21, 2017, 03:01:32 pm

Its the original shunt. Anyway, I've been away and this topic kinda exploded. As suggested earlier, I'm going to parallel mosfets for better heat dissipation. I can fit four on this particular cooler.

I see that you like to experiment. So how about progress to 400Watts 1KW Peak?

-snip-

Probably not a bad idea, I've got enough computer heatsinks to easily dissipate like 5KW.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 23, 2017, 12:32:16 am
Well guys, I tried a set of two parallelled mosfets, and I got nowhere. The unit started to draw crazy currents at like 16V and higher. I switched it back to a single mosfet and it works fine. Any ideas? I'd really like to go over 100W on this thing.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 23, 2017, 09:19:03 am
That sounds like parasitic oscillation. I'm not sure how long and tidy your wiring is but try a series 100R resistor right at the gate pin of each mosfet.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 23, 2017, 11:26:14 am
Such a massive heatsink on a single package just looks 'wrong'.

probably can't find anything of interest in the last page or so!  :palm:

That sounds like parasitic oscillation. I'm not sure how long and tidy your wiring is but try a series 100R resistor right at the gate pin of each mosfet.

Getting interesting.     :popcorn:
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 23, 2017, 11:57:08 am
Can't say it's that interesting.  :-\   It would be nice if the OP can achieve his 200W goal though.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Simon on September 23, 2017, 04:54:21 pm
Please can you calm down Armadillo!

AC heat exchangers are likely different from CPU coolers as gas is pumped through them. I don't see any circulation in CPU coolers, just conduction.
(http://www.heatpipe.nl/images/heatpipe.jpg)

And what is the speed of flow? would this make a difference to the type of construction like how many FPI?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on September 24, 2017, 01:04:26 am
That sounds like parasitic oscillation. I'm not sure how long and tidy your wiring is but try a series 100R resistor right at the gate pin of each mosfet.

I had the mosfets sitting right where the old one was, and some clever lead bending to get everything but the gate connected without extending them. The gate had probably about 1cm of extra wire.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 24, 2017, 09:07:07 am
The extra gate capacitance of 2 mosfets causing the opamp(?) that's driving them to go unstable? The gate resistors would help there too btw.

It's really hard to say without some sort of schematic.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 24, 2017, 09:53:52 am
By the look of it, the plates look very thin,
All CPU coolers with heat pipes have thin plates, show at least one with thick plates to prove opposite.
Quote
And the region where the tube connects to the plates [compression flare] looks ..... ehm... not too promising.
All coolers with aluminium plates are made like this.
Quote
Then you wonder whether it is a genuine Heat Pipe technology, and the preferred orientation of the installation for the migration and return of the flow of the refrigerant inside the tube.
So, I would say you firstly need to calculate the required thermal resistance performance of the heat-sink before you buy one.
There are no non genuine heat pipes, it simply won't work at all. Such cooler won't be able to dissipate any significant heat.

This is a true high performance cooler with thermal heat resistance published. 
Caught your bullshits right there. Innocent members should know the truths.

(https://www.eevblog.com/forum/repair/blew-a-mosfet-dissipating-200w-fake-mosfet-or-bad-circuit/?action=dlattach;attach=354040;image)

(https://www.eevblog.com/forum/repair/blew-a-mosfet-dissipating-200w-fake-mosfet-or-bad-circuit/?action=dlattach;attach=354042;image)

Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 24, 2017, 11:38:46 am
Curious that they describe a heatsink which is specified for 10-50CFM airflow as "Passive".

Presumable mainstream servers have big active fans.

My normal understanding of Passive is free-air.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 24, 2017, 11:51:16 am
Curious that they describe a heatsink which is specified for 10-50CFM airflow as "Passive".

Presumable mainstream servers have big active fans.

My normal understanding of Passive is free-air.

Probably at this low and quite airflow is classified as passive.
As a comparison, the 4'X2' ceiling diffuser in a library would be about 500 cfm.
While the Noctua NHD14 is doing around 55cfm to 65cfm.

Edit: it would be interesting to know how Noctua performs in terms of Thermal heat Resistance.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Simon on September 24, 2017, 11:55:06 am
Active versus passive us about method not quantity.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 24, 2017, 12:05:18 pm
Active versus passive us about method not quantity.

From the commercial world, apparently judging from the graph of quantified airflow, their definition is not the same as pure English....
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 24, 2017, 12:26:02 pm
Haha yes, they commercially ommitted the high volume, high static pressure fan that needs to be ducted into it. We understand commercial English, it's called Marketing speak, nothing 'pure' about that!  ;D
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on September 24, 2017, 12:36:25 pm
Haha yes, they commercially ommitted the high volume, high static pressure fan that needs to be ducted into it. We understand commercial English, it's called Marketing speak, nothing 'pure' about that!  ;D

Hahahaha,
Unless you can show case of that "BIG HIGH STATIC PRESSURE FAN" hook up, otherwise I tends to think more likely is just "BIG TALK".
I understand static pressure is proportional to velocity square V2, with that low airflow requirements, apple to apple comparison of the thermal resistance performance would be a better judge of the case really. Otherwise would we say, empty talk, the big version type.   :D

Edit:  Hahahahahahaha
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Gyro on September 24, 2017, 01:11:00 pm
Haha no, I wouldn't presume to claim to be an expert in fluid dynamics. I could easily pick up that it wouldn't work in 'pure Engish' free-air passive any better than the same outer outline of solid Aluminium though (probably worse because the heatpipes wouldn't be able to condense and would stall) but much beyond that and normal enclosure airflow design and optimization...

Just by looking at the picture though (and roughly estimating it's dimensions from the base footprint as no dimensions are given) I can only guess that it would take a fair amount of static pressure to force anything like 50CFM though that block of so many, tightly spaced, fins.

Just my thoughts.  ;)
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on October 01, 2017, 06:17:32 am
Just curious with OP updates?.    :popcorn:

Oh BTW, the dP [differential pressure] is shown on the right side of the graph.!
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: iamdarkyoshi on October 18, 2017, 06:20:14 pm
Just curious with OP updates?.    :popcorn:

Oh BTW, the dP [differential pressure] is shown on the right side of the graph.!

Ah I've been working on other projects but figured I'd check back here.

I'm going with a 140mm fan and two 1st gen mac pro heatsinks. If they can each cool a 95w CPU at fuckall airflow, they should cool a couple mosfets with a bigass fan :P
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on October 19, 2017, 01:08:54 pm
This is a true high performance cooler with thermal heat resistance published. 
Caught your bullshits right there. Innocent members should know the truths
Please elaborate how those pictures "caught me on bullshits" in any way?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on October 19, 2017, 03:32:57 pm
This is a true high performance cooler with thermal heat resistance published. 
Caught your bullshits right there. Innocent members should know the truths
Please elaborate how those pictures "caught me on bullshits" in any way?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: wraper on October 19, 2017, 03:50:11 pm
OK, heatsink on the picture has heatpipes and very thin plates. Where is debunking ?
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Monkeh on October 19, 2017, 03:55:33 pm
OK, heatsink on the picture has heatpipes and very thin plates. Where is debunking ?

Perhaps he's seriously confused and was originally referring to my comment about finding thermal resistance numbers for aftermarket PC heatsinks. Which.. still isn't debunked as of yet, because that's not a PC heatsink sold to consumers, it's a server heatsink for ducted systems sold to OEMs..
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Armadillo on October 19, 2017, 04:16:00 pm
OK, heatsink on the picture has heatpipes and very thin plates. Where is debunking ?

Perhaps he's seriously confused and was originally referring to my comment about finding thermal resistance numbers for aftermarket PC heatsinks. Which.. still isn't debunked as of yet, because that's not a PC heatsink sold to consumers, it's a server heatsink for ducted systems sold to OEMs..

I didn't even read your previous posts, hence what are you referring  to I don't know. ?????

Now, such ways of arguments are providing no added values to other members and is totally meaningless and a waste of time especially in a repair forum like this one, hence I will not entertain to further response regarding this subject.  Keep the peace.  ;Ds
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Monkeh on October 19, 2017, 04:20:38 pm
OK, heatsink on the picture has heatpipes and very thin plates. Where is debunking ?

Perhaps he's seriously confused and was originally referring to my comment about finding thermal resistance numbers for aftermarket PC heatsinks. Which.. still isn't debunked as of yet, because that's not a PC heatsink sold to consumers, it's a server heatsink for ducted systems sold to OEMs..

I didn't even read your previous posts, hence what are you referring  to I don't know. ?????

Didn't read it? You replied to it. Please take your alternate fact elsewhere.
Title: Re: Blew a MOSFET dissipating 200W.... Fake mosfet or bad circuit?
Post by: Simon on October 19, 2017, 04:44:20 pm
Can we calm it down please. This thread is becoming pointless and a candidate for locking.