Best 100v TO220 package FET made? It's cheap too. :-)

[liveforphysics]

Stumbled across this awesome FET tonight.

http://www.newark.com/texas-instruments/csd19536kcs/mosfet-n-ch-100v-150a-to-220-3/dp/38X4675

 

[Thud]

Looks great on paper....
Min qty=150......9 week lead time.

I am down 50 of em to play with on my last few infinions.......any one care to group buy with me?

 

[Doctorbass]

Nice find Luke!

As the rds ON become lower on these to-220 fets ( 2.3 miliohm compare to the 4110 with 3.7 miliohm or 3077 over their evolution i wonder how the thinny lead can manage that current and how is calculated the thermal dissipapation with the pcb traces and bus bar on the drain and source lead.

I mean the heat produced by the thiny drain and source lead compare to the case.. That ratio become lower...

These Icharger using some of the IRFB3205 have their lead very short ( 2-3mm long) between the case and the pcb to avoid them producing too mych heat and have high inducatance i guess..

Some used a single 6 fet 4110 and used at 50A... so this new fet with 2.3 miliohm it seem it woudl be difficult to imagine the same 6 fets controller used at 75-80A !

Doc

 

[sardini]

I think is time to repair my 18 fet crystalyte controller count me in for 20

 

[Honk]

A great mosfet at great price.
If 100V isn't a must then I prefer this heavy duty 60V mosfet at 1.65mR RDSon.
FDP020N06B @ Digikey

It has performed excellent in my bike controller where other similar mosfets has failed.

 

[whereswally606]

Honk that's a nice mosfet too but the digikey price ranges from 3.4 to 3.9$ for the 150 quant range. this one is cheaper still. nice find Luke

 

[mauimart]

Just came across the same FET in an email I received from TI "World's Lowest On-Resistance 80V and 100V TO-220 MOSFETs - Sample Now!" I put in for a couple of samples of both the 80V and 100V parts. Maybe I will put together a simple 6-FET power stage driven by my InstaSpin dev board.

 

[zombiess]

To-220 package has a real world limit of 75a max due to the leads. Still a nice MOSFET, for only 33% more than an irfb4110. I doubt its 33% better though.

I need to look at the datasheet, but if it has a lower Qg and lower diode v drop it could be a good improvement over the 4110s. I would like to see lower diode v drops vs lower RDS on. Way more heat generated in the free wheeling MOSFET than the conduction loss when its on.

 

[zombiess]

Just checked the specs, it is nice! Qg of 153nC vs 210 nC and the vf is 1.1v vs 1.3v. This blows the irfb4110 away. Now it needs thicker leads as that 150a package limit is datasheet koolaid. 150a for 1mS or 1 hour? The legs start to glow at 100a.

 

[neptronix]

Could someone just straight up replace 4110's in a controller ( keywin/xiechang/infineon type ) with these and expect everything to work peachy keen after?

Would the result be significantly less heat and higher efficiency for a 12FET which is on the verge of meltdown at 60-80A continuous?

 

[liveforphysics]

Could someone just straight up replace 4110's in a controller ( keywin/xiechang/infineon type ) with these and expect everything to work peachy keen after?

Would the result be significantly less heat and higher efficiency for a 12FET which is on the verge of meltdown at 60-80A continuous?

Yes to both of those questions to the limit of my understanding.

I've not personally had the chance to test one yet to verify the datasheet claims, but Texas Instruments has typically been a pretty straight-shooter with specs.

 

[zombiess]

If people on here are interested in these MOSFETs, I could buy say 200ish of them and make some miller plateau matched sets for replacements. I'd charge $3.50 per MOSFET for the matched sets.

I'm kinda curious to see how far an 18 FET with matched sets of these could be pushed. I'm thinking probably close to 150A battery 200A phase. Should have less heat issues. I've done close to this current with IRFB4110's at 75V batt and it works, but gets a bit warm. Biggest issue with the 18 FET controllers is they don't have enough surface area or mass to keep up with some MOSFETs. I know with IRFB4115's I have to limit them to 80-100A battery and 140A phase to keep them from getting really hot. With better thermal management an 18 FET IRFB4115's could do 130A batt, 200A phase.

 

[liveforphysics]

If people on here are interested in these MOSFETs, I could buy say 200ish of them and make some miller plateau matched sets for seconds replacements. I'd charge $3.50 per MOSFET for the matched sets.

I'm kinda curious to see how far an 18 FET with matched sets of these could be pushed. I'm thinking probably close to 150A battery 200A phase. Should have less heat issues. I've done close to this current with IRFB4110's at 75V batt and it works, but gets a bit warm.

Yes. I think it may be the highest power transfer you're going to find in TO220 package form at the moment.

An 18 FET laid out well enough to share current evenly between parallel devices should be capable of at least a reasonable 300A phase for perhaps a couple of seconds at least before over-heat. There could be an option to look at controller FET temps and just give the most phase current it safely can.

Think if the first two seconds of your 0-40mph acceleration was given a few seconds of 300A phase current to get the system moving.

You could do the 300A starts and not violate even the 250w regulations.

 

[zombiess]

Yes, this looks like a very promising MOSFET for those using < 80V max battery voltage. Too bad the min order is 150 units. I don't have a need for that many myself, but I'd like to build up an 18 FET Xie Chang board and compare it to an IRFB4110 built board that I have here. I have temp probes on the high side MOSFETs of all my test controllers so I can monitor what's going on and when I'm pushing the limits.

 

[liveforphysics]

Yes, this looks like a very promising MOSFET for those using < 80V max battery voltage. Too bad the min order is 150 units. I don't have a need for that many myself, but I'd like to build up an 18 FET Xie Chang board and compare it to an IRFB4110 built board that I have here. I have temp probes on the high side MOSFETs of all my test controllers so I can monitor what's going on and when I'm pushing the limits.

It sounds like you only would wish to buy perhaps 18-36 for yourself?


I would make the purchase and trust I would end up reimbursed through PayPal or whatever. I do not want them though, I want them shipped to someone who enjoys hosting group buys.

 

[johnrobholmes]

Great find Luke, thanks for sharing!

You should be able to use these on a xie chang with no worries. I don't think it worth the trouble to remove 4110s unless the controller is already overheating though. Just get a new one and install the new type.

 

[grindz145]

2.7mOhm? damn. Bye Bye 4110s, you had a good run. This basically allows the same current in an 18 Fet instead of a 24 Fet controller.

Are these GAN?

 

[liveforphysics]

2.7mOhm? damn. Bye Bye 4110s, you had a good run. This basically allows the same current in an 18 Fet instead of a 24 Fet controller.

Are these GAN?

Not GAN, just doped Si like a regular MOSFET, but a more optimized die design than anything else available at the moment.


You're a known-responsible guy Grindz, you want to receive a box with a few hundred of them and mail them off to various members who want them?

 

[grindz145]

Umm, yes. I would be happy to do this.

 

[bigmoose]

IR can't let this stand for very long. Bet we get a new release from them relatively soon.

Avnet Express has them, Minimum Order Quantity is 1. 1/$5.19ea 10/$4.06 100 /$3.56

http://avnetexpress.avnet.com/store...rtsearch&CMP=KNC-Octopart_VSE&c=USD&r=NA&l=-1

 

[grindz145]

I'll be happy to order 100 of these if we can get a minimum. I'll throw down for 24 to begin with. Any other takers?

 

[Thud]

THUD said:

I am down 50 of em to play with on my last few infinions.......any one care to group buy with me?

tell me when you need a payment bro.

 

[bigmoose]

I'll take 25 or a bit more to make the order.

 

[grindz145]

Alright that makes it easy, I'll leave the field open here for a little while longer then order a bunch.

 

[grindz145]

Element14 has them for 2.78 at a 100 piece order, with a max of 150, so I guess I'll just put the order in now.