PDA

View Full Version : Need input from all of the FET experts


Spoon
2006.06.16, 09:10 PM
I have compiled a short but comprehensive FET guide that will hopefully answer a lot of questions. If you guys could take a look and give me any feedback that would be great.

The final version will hopefully have more pictures and have a little more formatting to distinguish sections. I will also include a link to ByeBye's FET video. This will most likely be in some sort of web form and not a word document or pdf like it is now.

Let me know if I should include anything, exculde anything or if anything is just plain wrong. :o

Spoon
2006.06.20, 10:07 AM
11 views and no replies. Someone must have something to say...

briankstan
2006.06.20, 10:54 AM
you might just want to add that some "stock" forms of racing don't allow for upgraded or fet stacking.

great work.

imxlr8ed
2006.06.20, 12:26 PM
Well, wouldn't consider myself an expert but what the heck... here's some talking points... kind of in order. :o

Well, they are a part of any electronic speed controller... they recieve a signal on one (or two... like I said, I'm not an expert. ;) ) of the forward feet and open up the gate basically.

There are four fets on the board, forgot the steering. Helpful fact: the output side of a fet is always the side at the top of the lettering.

It's the amp draw that kills them, over-resistance on the motor is their main killer. Hooking up a motor that takes a larger amount of amps to operate requires higher rated fets.

Some fets can pull up their feet (unsolder themselves) too, it'll seem like a fried fet, but it's just a bad connection. I have seen fried fets with no blistering.

Personally, I like to think of fets as more of a gateway for the power, the more fets, the bigger the gate, the more power gets through. Thinking of them as a controller seems to be a little confusing to me, but like I said, they are the integral part of any new electronic speed controller.

Great job... it's alot to cover and anything anyone can provide as a simple tutorial for newcomers is a good thing ! :cool:

ruf
2006.06.20, 02:52 PM
"A stock Mini-Z circuit board has two FET's, one each for forward and reverse."

This statement can be construed a few different ways, but it is incorrect as I believe it was intended. The Mini-Z circuit board uses dual channel FET's in a push-pull configuration. For forward, the n-channel of one and the p-channel of the other are used, and vice versa for reverse. At any given point in time, both FET's are being used.

The term FET has become pretty standard to describe a chip that actually consists of two FET's, making the above explanation a little confusing. Also, FET is a pretty generic term. Specifically, we use MOSFETS - Metal Oxide Silicon Field Effect Transistors.

sciencemike
2006.06.20, 03:27 PM
My computer crashed so i cant open the pdf file right now, but ill read it soon. As for what i have read in the above passages, RUF is right. But there is also another error, well kind of.. The AD board uses 4 fet packages. Two on each side. So a total of 4 3010 fets run motor. I have pictures if you need them...

Spoon
2006.06.20, 03:48 PM
Great input guys, thanks for the feedback.

I haven't taken a close look at the data sheets so I didn't know it was a dual chip. I'll try to make that more correct without making it unnessarily(sp?) complicated.

Mike,

If you have pictures of any of the boards that would be helpful.

imxlr8ed
2006.06.20, 05:01 PM
IRF is pretty helpful in understanding more than you want to know ! :o

http://ec.irf.com/v6/en/US/adirect/ir?cmd=catNavigateFrame&punchInID=71

Here's the PDF of those 7317s I used to use... actually, I'm pretty sure a few of the others listed would work as well, still looking for the 3010's info.

Lotta learning to do myself... current questions:

Do fets only operate within certain recieved "signal" parameters ? (is this the ID channel column ?) Is the rating shown a minimum... ? Somebody smart... please help ! It would be very cool to know exactly what fets are compatible with the Zs... I have a feeling it's all been guess work up until now. (or has someone already gone through this process ?)

imxlr8ed
2006.06.20, 05:12 PM
Whoa... diggin deep now !

http://www.me.sc.edu/courses/emch367/Download/Func_bx/mosfet_edited_Brian.doc

imxlr8ed
2006.06.20, 05:29 PM
For example, how do I know if something like this would work ?

8 volts ! 3.0 amps... interesting !

imxlr8ed
2006.06.20, 05:42 PM
Aha !!! Think I found the 3010s... they're Hitachis apparently.

imxlr8ed
2006.06.20, 05:48 PM
Are these the 3004s as well ?

Sorry I'm dumping all of this on here... just trying to find where to get the 3010s, and if I can't get them, what is comparable?

sciencemike
2006.06.20, 06:33 PM
Yes imxlr8ed, the 3010's are hitachis. Not IRF. I had ph2t tell me that same bit of trivia a few weeks ago. BTW, i took my 3010's off my AD board, in lieu of a nelly. So they are up for sale, dont know right now for how much. But i have 4 of them..

Mike

imxlr8ed
2006.06.20, 06:51 PM
I do have a set on my AWD... was thinking about popping those if I can't find the 3010s for sale anywhere. But then again, I really don't have to worry about this until 2007... until the market gets filled with the new style MR02s.

Thanks though, maybe some else might jump on them.

Spoon
2006.06.20, 06:59 PM
Yes imxlr8ed, the 3010's are hitachis. Not IRF. I had ph2t tell me that same bit of trivia a few weeks ago. BTW, i took my 3010's off my AD board, in lieu of a nelly. So they are up for sale, dont know right now for how much. But i have 4 of them..

Mike

So you have an AD Nellified AWD? Impressive. I knew you had the AD band and I knew you had the Nelly but I didn't know they were the same car. It could be tho onlyone out there! :eek:

sciencemike
2006.06.20, 08:08 PM
Yeah thanks spoon, its also been the only 98 mm VDS AWD with AD band and NElly for about two months... But now all you jokers have the 98 mm's. Ill have to get somthing else to one up yall.

Spoon
2006.06.21, 10:21 AM
Yeah thanks spoon, its also been the only 98 mm VDS AWD with AD band and NElly for about two months... But now all you jokers have the 98 mm's. Ill have to get somthing else to one up yall.

Maybe a 106 mm AWD AD Band, stacked Nelly, Toyota Estima Minivan with lights and hydraulics. :cool:

Back on topic:

There is a lot of info coming through here...Excellent!

I think I will leave this article as a beginner oriented resource. There has been a lot of FET info popping up with the new 3010's coming on the scene. I think all of that info will probably get compiled into the FAQ.

Spoon
2006.06.21, 03:54 PM
It's the amp draw that kills them, over-resistance on the motor is their main killer. Hooking up a motor that takes a larger amount of amps to operate requires higher rated fets.



I think you are half right here imx. After getting some motor teachings from ScienceMike, hotter motors actually have less resistance. This would mean at the same voltage (4 AAA's) they are going to draw more current. So, yes, the amp draw kills them but I think it's because the motors have less resistance.

I=V/R. As R goes down and V is fixed, I goes up.

imxlr8ed
2006.06.21, 05:24 PM
I think you are half right here imx. After getting some motor teachings from ScienceMike, hotter motors actually have less resistance. This would mean at the same voltage (4 AAA's) they are going to draw more current. So, yes, the amp draw kills them but I think it's because the motors have less resistance.

I=V/R. As R goes down and V is fixed, I goes up.

Takes more amps to get them spinning though, something to do with the generated magnetic field I think.

I know enough to get by... but I'm certainly not an expert ! :D

I do know that touching the black and red wires together makes pretty sparks too !!! (well, al least once or twice ! :rolleyes: )

Slipstream
2006.07.10, 02:52 AM
Yes imxlr8ed, the 3010's are hitachis. Not IRF. I had ph2t tell me that same bit of trivia a few weeks ago. BTW, i took my 3010's off my AD board, in lieu of a nelly. So they are up for sale, dont know right now for how much. But i have 4 of them..

Mike


sciencemike are the 3010s still for sale? Let me know how much. Sent you an email through this board. Thanks.

tungfai
2006.08.06, 10:14 AM
Why no people use 8958A? I use it since 2002.
My feeling is the most best liner control FET. But need to stack 3X2 or 4X2 if use mod motor.

LBRC
2006.08.06, 04:55 PM
Why no people use 8958A? I use it since 2002.
My feeling is the most best liner control FET. But need to stack 3X2 or 4X2 if use mod motor.
If you mean these:

Fairchild FDS8958A
N-Channel
Max Rated Continuous Drain Current = 7.0A
RDS(on) = 0.040 ohms @ VGS = 4.5V, 6A

P-Channel
Max Rated Continuous Drain Current = -5A
RDS(on) = 0.080 ohms @ VGS = -4.5V, -4A

Note; since the P channel will blow somewhere around 5+ amps the higher N channel current rating is a mute point.

Verses:
Renesas 3010
N-Channel
Max Rated Continuous Drain Current = 6A
RDS(on) = 0.045 ohms @ VGS = 4.5V, 3A

P-Channel
Max Rated Continuous Drain Current = -5A
RDS(on) = 0.130 ohms @ VGS = -4.5V, -2.5A

IRF 7317
N-Channel
Max Rated Continuous Drain Current = 6.6A
RDS(on) = 0.029 ohms @ VGS = 4.5V, 6A

P-Channel
Max Rated Continuous Drain Current = -5.3A
RDS(on) = 0.058 ohms @ VGS = -4.5V, -2.9A

IRF 7389
Even though these are worth considering I'm leaving the specs. out because just listing their Drain Current and RDS (Resistance Drain to Source) ratings for comparison with the others would be misleading. The 7389ís have noticeably greater heat dissipation abilities and faster switching characteristics that put them close to and even sometimes above the 4562 under the right circumstances when installed in a Mini-Z. For comparison I would list them as very close to but slightly under the 4562 for most applications.

Vishay 4562 (The Current King)
N-Channel
Max Rated Continuous Drain Current = 7.1A
RDS(on) = 0.025 ohms @ VGS = 4.5V, 7.1A

P-Channel
Max Rated Continuous Drain Current = -6.2A
RDS(on) = 0.033 ohms @ VGS = -4.5V, -6.2A



Then why? Because for price IRF 7317ís are a much better deal, and even easier to come by, and when price isn't a consideration the Vishay 4562's walk all over them.

Although after looking at them they are quite close to the relatively hard to come by 3010ís and might be something worth considering as an allowed substitute/replacement for future stock class racing.

herman
2006.08.06, 10:20 PM
anybody heard of 0113 fets?
so from what i gather the 4562 is the best?
and as for the boards, ad band board utilizes 4 3010 fets tp power the motor? how about the latest boards that are just beginning to use the 3010 fets or should i say mosfets, are they utilizing 4 too?

thanks for any reply...

tungfai
2006.08.07, 12:02 PM
I don't think only to focus on the rating of the voltage and the current. Already stack FET, how come need to concern the current? Still not enough?
I already used over 1k pair of FET before. All the comment is better then 7389 and 4562.
You better check check what is the best of 8958 first. 1 character is no other fet better then it.

tungfai
2006.08.08, 11:38 AM
Anyone have try to stack Turbo FET on the PCB? I mean 1 is P-ch. Another is N-ch.

imxlr8ed
2006.08.08, 11:57 AM
Yes... we have done turbo stacks, not sure what you are getting at here ?

imxlr8ed
2007.01.05, 10:00 PM
Duh moment...

Which side is forward again ? If I'm just replacing one 3004 with a 3010, which one would it be. Sai and Gasman have done this, and it works fine... but I'm sitting here trying to remember which side is which. I think he said the fet closest to the direction of the crystal.

Just want to be sure...

LBRC
2007.01.05, 10:17 PM
See attached photo

imxlr8ed
2007.01.05, 10:21 PM
Uhhhh... that's not what I'm asking here, which fet should I replace to get this thing faster in FWD. I know which side of the fet is the output side.
Thanks for the fast reply though.

Ninja
2007.01.06, 12:36 AM
I wouldnt recomend only replacing the FWD FET as a chain is only as stong as the weakest link, in this case, the 3004 FET. and just for reference. The FWD FET would simply be the one on the side that the +(red) motor wire hooks to.

imxlr8ed
2007.01.07, 11:38 AM
I know of some that have done this, and have not had any issues... I've raced against them, and I know it works fine. I know where you're coming from though... I always thought it better to do both as well, but when I started hearing the single swap idea from a bunch of World Cup drivers, I figure it's gotta be ok to do.

LBRC
2007.01.07, 02:45 PM
Canít seem to find the thread where this was discussed before, maybe it was over at Ripperís.

It actually works pretty well, since yes the H-bridge is only as strong as itís weakest link which in the 3004ís case is the wimpy 3.5 amp 0.34 ohm P-channel, so when you replace the forward P-channel with the 3010ís 5 amp 0.13 ohm P-channel you get a significant improvement when going forward since you are now using that improved P-channel with the 3004ís better half the 5.5 amp 0.11 ohm N-channel which ends up being a nice well balanced combination. Reverse is not improved though since it still uses the 3004ís less than stellar P-channel but then since very few people go haling around the track WOT in reverse itís rarely much of a handicap, although I do seem to remember a video somewhere where it would have been.

Depending on your car motor and radio you want the 3010 on the positive output side when going forward, which should but is not always the red wire. If you get it wrong itís simply a mater of swapping your motor wires and reversing your TX, unless you use a KT-5 that is.

imxlr8ed
2007.01.07, 03:35 PM
I kinda figured it was a P channel, N channel difference in the 3004s that allowed this mod to work. (nice to actually see the numbers though !)

Thanks ! :)

LBRC
2007.01.07, 06:12 PM
The resulting P and N channel forward specificationsre are sort of ethically pleasing, the closeness of the numbers look like they belong together.