I've got FETs in my SE!

Mostly dealing with electronics and pcbs and modifying them, all the things that make electrical-engineers tick and the rest of us cringe in fear.
Post Reply
VooDooMafia
bitPimp
Posts: 1266
Joined: Sun Aug 17, 2003 12:34 pm
Location: Canada
Contact:

Post by VooDooMafia »

hue35 wrote:Ain't no pimps needin' to be played in the name of science... that's just too damn complicated. It seems like you have a lot to offer in solving the problem at hand, and it will definitely not go unappreciated if you continue to lend your brain.
Wise words from a wise man.

Hell even when someone in the hood does a bad job there are people right there telling them it was a good effort then pointing/helping them improve.

This place is not one to just keep things inside all bottled up, if theres an idea out there you will see if from the regulars here and alot of the new pimps aswell.

The internet was made for the sharing of info and this is one site that you will find that is all about the free clear info :)
Image
I Don't care what you say PIMPIN aint easy, you Broke A$$ Mother F*&Ker
ph2t
bitPimp
Posts: 1979
Joined: Tue Mar 25, 2003 6:33 am
Location: taco city

Post by ph2t »

cyborgzero wrote: Well, being introduced in this thread as "turkey" sorta pissed me off to begin with.. PHt doesn't even frikken know me and is calling names behind my back?
silly nugget, like I give a shit......

this thread has gone from a great techo development thread into a pissing match.......

the next comment I make will be to do with the tech and not the bullshit, it's just a waste of time.

From cyborgzero's suggestions:

1) Dude, I've tried this already, unfortunately it's a no go. The two resistors in question, when shorted out (ie: a wire bypasing them) don't add any extra voltage to the gate pins on the n-chan fets for fwd or rev. Also, you read the value of the resistor incorrectly :smile: , the resistor going to the gate pin on the reverse fet is 300 Ohm [301] not 1K Ohms [103], something easy to do, specially with surface mount. You are right though, the gate voltage needs to be raised to at least 2V to be feasible. (See my previous post re: Gate voltages vs. Rds(on)). As an experiment I manually put 3V on the gate pin of the fwd fet and the car just flew! So I now know the goal of what we are all working to.

ph2t.
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

ph2t wrote:From cyborgzero's suggestions:

1) Dude, I've tried this already, unfortunately it's a no go. The two resistors in question, when shorted out (ie: a wire bypasing them) don't add any extra voltage to the gate pins on the n-chan fets for fwd or rev. Also, you read the value of the resistor incorrectly :smile: , the resistor going to the gate pin on the reverse fet is 300 Ohm [301] not 1K Ohms [103], something easy to do, specially with surface mount. You are right though, the gate voltage needs to be raised to at least 2V to be feasible. (See my previous post re: Gate voltages vs. Rds(on)). As an experiment I manually put 3V on the gate pin of the fwd fet and the car just flew! So I now know the goal of what we are all working to.

ph2t.
Dude, im dyslexic as hell lol esp when my eyes are crossed in my microscope. ;)

Hrm, well, I am trying to increase the source voltage going into the driver chip on the upper left that is connected to the gate drivers...

There are two sources of power, one is 2.9 volts or so, and another is 1.67.. I am thinking the 2.9v is the source for the gate drivers, and the 1.67 is internal operation voltage.

So, what we are dealing with seems to be that we take 2.9v, knock another .7 volts or so off of it lost to the chip inner workings, and then you get a gate drive that ends up 1.5 or so pulsed peak to peak.. I am not sure about eh pulsed part, but I will verify it on a scope.

So, get more power to the drive side of the chip and you are golden. We can totally bypasss the rest of the BS.

Rob
ph2t
bitPimp
Posts: 1979
Joined: Tue Mar 25, 2003 6:33 am
Location: taco city

Post by ph2t »

you've got a scope? you lucky bastard! :D

This is an idea I was toying with as well. The three pin chip on the pcb is a dcdc converter, prolly the ME301 chip that was made popular on the 1st gen zz, bits, etc.... So the output is 3V for a "less than 3V input". If we could use another DCDC converter to up the voltage to say, 5V then hopefully as you said the output would be higher than the 0.9V that we are getting. I'm hesitant to do this 'cause I don't know the operating range of the IC's on the pcb and I've only got the one zz se!

ph2t.
User avatar
mad_haxor
Posts: 16
Joined: Sun Dec 07, 2003 9:38 pm
Location: da' hood

Post by mad_haxor »

ph2t I agree with you on the bias resistors. they shouldn't make any difference in this circuit, since we are dealing with dc instead of ac, and fets are voltage controlled instead of current controlled like bjts.
also I would kinda be scared to run 5 volts through the PC board.
"been pimpin since.... been pimpin since....been pimpin" (Babypowder)
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

mad_haxor wrote:ph2t I agree with you on the bias resistors. they shouldn't make any difference in this circuit, since we are dealing with dc instead of ac, and fets are voltage controlled instead of current controlled like bjts.
also I would kinda be scared to run 5 volts through the PC board.
Okay, you say the bias resistors don't matter.. Well, they *shouldn't*, unless you have a voltage divider type scheme on the gate, which I have seen before.. Then, they do matter and you can't overlook that when reverse engineering a circuit, because the gate voltage becomes a function of the ratio of the two bias resistors. But, so far it looks like we have a straight through design with no pull up or pull down resistors.

I am only concerned about going over 3v because there are alot of chips now that can only tolerate 3.5v.. 5v digital chips are not the standard anymore.
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

ph2t wrote:you've got a scope? you lucky bastard! :D
ph2t.
Yeah! I used to take alot of work home with me where I could work on it without being interrupted..

But, yeah, the DCDC converter is a good idea, BUT, i am just a bit concerned about what else might be affected.. Im trying to keep the rest of the car as stock as possible..

When you increase the voltages it tends to increase efficiency, so for a small car like ours, a 1-2v bump up would prlly make it run longer too.
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

okay, here are some specifics:

1) the gate input is pulsed DC, pulses are 2.5v peak, SO increasing the gate voltage won't do monkey squat.. Its RIGHT where it should be. What we were measuring was RMS voltage on our voltmeters of around .93.

2) I am going in a COMPLETELY different direction and I am going to try to increase the pulse width to the H-bridge. THAT is the REAL reason why when you put *steady state* 3v the motor screamed. ;)

If you jumper the 2.89 supply line to the bridge you get the same results. :)

So, there is a way to increase the PWM by simply adding a capacitor from the gate to ground and turn it into a lowpass. Pretty simple. ;)

Duh, didn't know why I didn't think of that before. :)

Rob
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

ok, update;

P-chan fets on the top side (black stuff): Okay, the gate voltage on those bad boys is a mere 1v peak. That explains why when you swap the bottom N-chan fets on top, you only get a mild speed increase, and when you add the ones on top, the car picks up quite a bit, and when you double the ones on top, it really gets a good kick but its still not superfast.. Its just a small touch faster than a stock ZZ now.

WE have ruled out the voltage on the N-chan fets: they are not the problem at a 2.5v peak pulse.

So, there is only one thing left, and that would be to increase the pwm amount.

The N-chans are already pretty well on.

Rob
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

Ok, to continue this saga: ;)

I ran the FWD P-channel fet off of the gate of the REV N-channel, giving the P-channel a ZERO gate voltage, thereby turning it completely on when the FWD is on... I removed the 750 (75 ohm) towards the front of the car and ran a wire directly from the gate on the FWD P-chan to the gate of the REV N-chan. Now that FWD P-chan is DEFINITELY ON all the time. ;) Before it only went down to around 1.6v, now its down to ZERO AND its not being PWM'd, its on in a steady state. :-D Ya baby... I tried to run the reverse this way, off of the gate of the FWD N-chan, but it gets stuck on and I haven't dug into it deeply enough to figure out why, besides I don't use reverse all that much and its still fast enough to do quick switches and donuts out of reverse, so I don't really need it.

All I can say is wow.. Another good kick in speed came from that simple little mod.

Now there are still those two pesky current limiting/biasing resistors that need to be dealt with. ;) 104 and 102 from what it looks like, on the top and bottom of the H-bridge feed.
potentiality
Posts: 9
Joined: Fri Dec 26, 2003 7:30 am

Post by potentiality »

Are you going to have an updated diagram? Or are you still thinking of showing a final version after the transistor change?
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

yeah, once this is all done I will put up a completed diagram of what to do..

In advance, go ahead and do the mosfet mod.. Either way thats gotta be done..

Then I will show how to reroute the gate of the FWD 6401 P-chan fet to get decreased resistance and increased speed.
potentiality
Posts: 9
Joined: Fri Dec 26, 2003 7:30 am

Post by potentiality »

Personally, I'll wait, for then I can order all the components in one shipping. Plus even after reading the thread.. the modifications done since the original post (such as your rerouting to have the FWD fet on all the time) are stil beyond my technical know how without diagrams.

In the meantime. What tip and wattage soldering iron should I seek (the ones I have now are way to big.. I did some smr before on some g. cards and motherboards, but the soldering work always looked like dooky)?

BTW Thank you for all you electrical buffs who have put so much time into this so far.
cyborgzero
bitHood
Posts: 25
Joined: Thu Dec 18, 2003 8:48 pm

Post by cyborgzero »

well, if you got the $$$, a Metcal SP200 iron with the 406-248 1/64th needle bent chisel tip is the BEST, second best would be a weller EC4001.. You can get them used from Ebay for fairly cheap.
viperdout
bitGangsta'
Posts: 482
Joined: Sun Nov 23, 2003 1:49 pm

Post by viperdout »

ph2t wrote:dude, microamps is known for fet mods 'cause he sells the service to other people, that's all. he didn't come up with the mod, the krauts did!
he did invent this new SE fetmod that he won't tell us about, hence all the discussion in this thread.

ph2t.
Who are they?
Post Reply