fake MOSFETs on Ebay?

Thread Starter

strantor

Joined Oct 3, 2010
5,025
I found these mosfets on Ebay. They are in the TO-3P package and stamped "IXTH88N30P", which according to the IXYS datasheet they should be in a TO-247 package. They are also stamped "philippines" and shipping out of China, at half price. seems shipping from philippines to china, to america should drive up the price at least a little bit. Should I be suspicious?
 

Wendy

Joined Mar 24, 2008
21,840
Actually lots. I wouldn't mind some TO3 case styles myself.

With TO220 case styles the 80A is a bit deceptive. The MOSFET might be able to do it, but the leads for the package can't. A TO3 can handle much higher currents IMO.

************************

OUCH! $80 for a MOSFET!!! I think I would pass myself.
 

shortbus

Joined Sep 30, 2009
6,893
@strantor - they are in the TO-247 package. according to the data sheet you link to, they come in four different packages and the prefix 'IXTH' is the one for TO-247.

the '88N30P' is the device type, an the prefix is for the different packages. The D/S also states that they are limited to 75amps rms by the external leads.
 

SgtWookie

Joined Jul 17, 2007
22,201
The TO-3P packages aren't the good old hermetically sealed TO-3 steel package.

It does have more current handling capacity than the TO-220 package, and better thermal characteristics.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,025
@strantor - they are in the TO-247 package. according to the data sheet you link to, they come in four different packages and the prefix 'IXTH' is the one for TO-247.
sure about that?? from the data sheet:


from ebay:


looks to me the FET in the Ebay pic should be stamped IXTQ and not IXTH


Also, regarding the lead current limit; I am aware that all the FETs I find in this package are going to have a current limit of 70 or 75A on the leads. At first I decided that it was pointless to look for a FET with a Amp rating higher than 70, but now I've realized that if the FET starts out at 70, the hotter it get, it just goes down from there. So, I'm now again looking at FETs in this type package with Amp rating much higher than 70, so that even at 125c I can still get 70A out of them. I know we talked about the better MOSFET packages such as ISOTOP and SOT-227, but IMO I am probably going to blow stuff up in the beginning and I don't want to start off blowing up expensive stuff. Once I come up with a design that works and I've ruined enough stuff to actually know what I'm doing, then I'll buy the good stuff and make a good, long-term reliable design.
 

Attachments

tom66

Joined May 9, 2009
2,591
I would like to see a three phase design using these. 6 per phase (3 high, 3 low) powering a delta(?)-wired motor. 88A per MOSFET, but by thermal calculations assuming a good 0.5°C/W heatplate, I've come up with 40A per device, maximum. 120A @ 288V (for 20 Pb 12V 17Ah batteries) per phase would do 20kW nicely, about 30hp... that's more than enough for city driving.

And 20 lead-acid batteries @ 17Ah each would give around 4.8kWh. Might want to double that up for extra range. The Volt has 16kWh LiPo.

Just wonder what the range would be on it, according to several sources average kWh per mile is 0.1-0.2 in a Prius, so that's 24-48 miles...

Maybe look into IGBT's as a good IGBT will only drop 0.5-1V in operation which is possibly lower than the resistive loss of the MOSFET.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,025
I would like to see a three phase design using these. 6 per phase (3 high, 3 low) powering a delta(?)-wired motor. 88A per MOSFET, but by thermal calculations assuming a good 0.5°C/W heatplate, I've come up with 40A per device, maximum. 120A @ 288V (for 20 Pb 12V 17Ah batteries) per phase would do 20kW nicely, about 30hp... that's more than enough for city driving.

And 20 lead-acid batteries @ 17Ah each would give around 4.8kWh. Might want to double that up for extra range. The Volt has 16kWh LiPo.

Just wonder what the range would be on it, according to several sources average kWh per mile is 0.1-0.2 in a Prius, so that's 24-48 miles...

Maybe look into IGBT's as a good IGBT will only drop 0.5-1V in operation which is possibly lower than the resistive loss of the MOSFET.
Tom I already have 3 phase motors (have a nice 2hp 3ph pump motor and several fractional HP 3 phase mtrs.); I have been thinking of making a 3phase controller, but I need to get my hands dirty with something simpler first. Once I prove to myself that I can get MOSFETs to control something, I will quickly outgrow this DC thing. Like I said in another post, for every question answered I come up with 10 new questions. This rabbit hole is too deep to go any further without experience.

EDIT: the reason I'm still chasing MOSFETs instead of IGBTs is because of the negative temp coefficient. I have never seen how these things heat up or proper ways to heat sink them. Let me blow up a few and get the idea, then maybe I'll be ready for IGBTs
 

tom66

Joined May 9, 2009
2,591
Tom I already have 3 phase motors (have a nice 2hp 3ph pump motor and several fractional HP 3 phase mtrs.); I have been thinking of making a 3phase controller, but I need to get my hands dirty with something simpler first. Once I prove to myself that I can get MOSFETs to control something, I will quickly outgrow this DC thing. Like I said in another post, for every question answered I come up with 10 new questions. This rabbit hole is too deep to go any further without experience.

EDIT: the reason I'm still chasing MOSFETs instead of IGBTs is because of the negative temp coefficient. I have never seen how these things heat up or proper ways to heat sink them. Let me blow up a few and get the idea, then maybe I'll be ready for IGBTs
I suggest a massive heat plate possibly ripped from an industrial control system with the devices bolted onto it. Use a small mica insulator, or a good glob of non-conductive thermal paste, because the tab is connected to drain or source. I'd also suggest forced air cooling or perhaps even water coolant pumped through it to a standard automotive radiator. Phew!

If you use something like pure oil or distilled water, you can completely submerge the devices; but be careful of a short circuit due to impurities. And the water may end up boiling.

By the way, when MOSFETs blow up, they fail shorted ~90% of the time. This is the problem with a bridge 3-phase design; a single device failure is likely to cause a complete failure of at least the other three devices in the lower or upper half, and possibly some in the other half, too. However, that failure will directly short out the battery, so should quickly incinerate the fuse. Whereas a low-side drive circuit for a DC motor may well lead to uncontrolled acceleration. Having a high-current isolation switch / relay will be essential. Make sure the brakes in the car can stall the motor, too.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,025
I suggest a massive heat plate possibly ripped from an industrial control system with the devices bolted onto it. Use a small mica insulator, or a good glob of non-conductive thermal paste, because the tab is connected to drain or source. I'd also suggest forced air cooling or perhaps even water coolant pumped through it to a standard automotive radiator. Phew!

If you use something like pure oil or distilled water, you can completely submerge the devices; but be careful of a short circuit due to impurities. And the water may end up boiling.

By the way, when MOSFETs blow up, they fail shorted ~90% of the time. This is the problem with a bridge 3-phase design; a single device failure is likely to cause a complete failure of at least the other three devices in the lower or upper half, and possibly some in the other half, too. However, that failure will directly short out the battery, so should quickly incinerate the fuse. Whereas a low-side drive circuit for a DC motor may well lead to uncontrolled acceleration. Having a high-current isolation switch / relay will be essential. Make sure the brakes in the car can stall the motor, too.
Yeah My plan was just to bolt them to the largest chunk ouf aluminum I can find (with thermal grease) and go from there. I have run across formulae for heat sink sizing but didn't read it because I was on a different mission; one of the many useful bits I've neglected. I've put probably 60hrs of research into this in the past 4 days and I still feel like I haven't even scratched the tip of the iceberg. And let's not start throwing around the big "C" word just yet, remember this is for a go-cart.
 

tom66

Joined May 9, 2009
2,591
Yeah My plan was just to bolt them to the largest chunk ouf aluminum I can find (with thermal grease) and go from there. I have run across formulae for heat sink sizing but didn't read it because I was on a different mission; one of the many useful bits I've neglected. I've put probably 60hrs of research into this in the past 4 days and I still feel like I haven't even scratched the tip of the iceberg. And let's not start throwing around the big "C" word just yet, remember this is for a go-cart.
"C" word? Anyway, try using copper as it has high thermal conductivity. Expensive, though...
 

Thread Starter

strantor

Joined Oct 3, 2010
5,025
"C" word? Anyway, try using copper as it has high thermal conductivity. Expensive, though...
= car. I have a bunch of scrap aluminum, somewhat banged up. May need to be resurfaced to make good thermal connection.

Anyways, I'm still not sure if those MOSFETs are legit. I sent the guy an email and said I was 99% sure they were factory rejects and that I'd give him 175$ for 100 of them (he's asking like 300-something for 100).
 

tom66

Joined May 9, 2009
2,591
= car. I have a bunch of scrap aluminum, somewhat banged up. May need to be resurfaced to make good thermal connection.

Anyways, I'm still not sure if those MOSFETs are legit. I sent the guy an email and said I was 99% sure they were factory rejects and that I'd give him 175$ for 100 of them (he's asking like 300-something for 100).
High voltage power MOSFETs are pretty pricey. Was looking at 300V, 56mOhm here: http://www.fairchildsemi.com/pf/FD/FDA59N30.html

That's $3.08 in 1k units, each.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,025
High voltage power MOSFETs are pretty pricey. Was looking at 300V, 56mOhm here: http://www.fairchildsemi.com/pf/FD/FDA59N30.html

That's $3.08 in 1k units, each.
Am I missing something? I can't see the datasheet. The page shows 59A. I (keep in mind I really don't know what I'm doing since I've never done it before) am looking for something with at least 110A (which I will limit to 70A) capacity so that @125C I can still get the 70A out of it that the legs are rated for. After that requirement, then I would like at least 180 or 200V (leaving room to grow; later if my design works I will be using higher voltage motor) and a low Rds(on). MOSFETs like that can be found, but not for cheap, so I'm in a battle as to if I want to spend the money for the good stuff and blow it up. If I have to skimp on something, I guess it would be voltage; for my lower voltage I might go down to 50V or something; can't see myself going above 50V on a starter motor. I am tracking down several things in parallel. For every mosfet I look at, I also need to figure out what driver to use and I still haven't figured out how to do that.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,025
Man, I feel you guys might be trying to polish a turd, giving me all these good links. I just don't understand enough to know a golden nugget if it fell in my lap. The first link looks cool, like all the work is alreay done for me, but I probably wouldn't be able to recreate it if I don't understand what's going on in there. The 2 IGBTs in the middle; I had found a 1200V 1200A module like that but when I went to look for a driver for it, I couldn't find the right IC. The last link to the ISOTOP one looks good but I'd be afraid to play with them for the price.

I think I need to take one more step back and buy some cheap MOSFETs, low voltage, and make a controller for something a little more tame. I'm putting the cart before the horse here; biting off more than I can chew; eyes too big for my belly, and all that other stuff.

Sorry if I've wasted you guy's time, but I will be back to waste more of it, once I order some stuff and catch it on fire.
 
Top