upgraded FETs

Nice it looks like the power is more regulated, but the smaller componets and the resistors are limiting your current. Neuro's FET design is more effective and does not get hot at all. Warm but not hot. I think the transistors and resistors your have to regulate the power is causing the restrictive current output.
(At least it looks like small transistors setup in a regulate type circuit.) Post the part numbers of the smaller transistors maybe we can fix the heat problem. Post a diagram if possible. Oh are you using logic level Fet's.

I say this because I'm currently designing an external turbo for my mini-z and I want something that meets the following (high) expectations:

-Ultra low Rds(on)
-Good power disipation (D2PAK? SO-8?)(TO-220 is too big, lol)
-Easy to install (idealy run of the motor terminals)
-Ability to drive on 5, 6 cells.
-possible incorporate mosfet driver and 10V DCDC convertor cct.

I know this doesn't help you that much but as a fellow devotee to the black arts I thought I'd share......

greets from downunder. If you're on MSN, look me up ph2t@prahranpianos.com

ph2t. [/b]

Well ph2t I have finished Nero's version of the Fet upgrade with the SO-8 package and it seems to work ok. It doesn't have the MAX total current load as the TO-220 package Fets but works just as well. Didn't test it on a plasma or any other super mod'd motor
just one a SB hybrid with 25 turns.

Hope this helps.

[Jshwaa]

Nice it looks like the power is more regulated

There's no regulators on that board, only FET's and resistors.

but the smaller componets and the resistors are limiting your current.

The smaller components are FET's and the resistors prevent their drain-to-source current from frying them. Guess you'd need a circuit diagram to even know that, which puzzles me as to how you're an expert on this circuit having never seen one.

Neuro's FET design is more effective and does not get hot at all. Warm but not hot.

That may be, but mine isn't finished yet. I highly doubt that anyone's attempts at external ESC building has conquered some of the dilemnas that it introduces. For instance, how are the gates of the FET's being driven completely on and off, when the weak signal from IC1 isn't even half as strong(voltage-wise) as what's needed to drive an FET with VGS of +-10V?

Those smaller(TO-92) FET's are what I used in an attempt at conquering that dilemna. It didn't work as well as I had anticipated. The parts that get hot are the power FET's, not the supposed current-restrictive smaller FET's and resistors. Without those parts(TO-92's and resistors) the ESC couldn't have worked as well as it did. So I'm completely puzzled as to how anyone's(Neuro's, Ph2t's, etc.) could have had produced desired results. I haven't seen any more than what's been provided on the forums, so I don't know for sure, so don't get me wrong. I'm not trying to bash anyone's enginuity or knowledge on the subject.

Post the part numbers of the smaller transistors maybe we can fix the heat problem. Post a diagram if possible. Oh are you using logic level Fet's.

I know exactly where I went wrong and have parts on their way(by next monday,5/3) to build another one. I'm confident that the issue will be resolved and the ESC will then work at peak efficiency(200W, 74A, 20V, 0.028 RDS(on) for forward or reverse minus the motor load)

which puzzles me as to how you're an expert on this circuit having never seen one.

Jshwaa, did I upset you? Question your intellegence? I'm not sure what was ment by this on your post. If at all you find something your don't like just say so. I didn't mean anything at all in my post. It was just a guess at what was used in the design. No I'm no expert, after all you made it. I'm just saying with a "little help" from everyone maybe the problem can be handled faster. Who knows even better then the original. If you like working things out on your own that's fine too. Props to you for it seems that your are taking this to the next step. Even after me moving and being off of the boards for a while your post was the first new thing I saw here that was advancing this project.

Peace be with you...
Codedat

[Jshwaa]

Jshwaa, did I upset you? Question your intellegence?

Hehe. No. I just made a valid point. Sorry that it sounded harsh. You thought the smaller transistors and resistors were restricting current, as if they were in the path of current going to the motor. I just wondered how that could have been assumed. No biggie.

Hehe,

NP. Just thought that the givens were taken when you wrote your post.

[Jshwaa]

Sorry, *reaches out to shake hand*, my woman's got PMS, lol. Actually she's 4.5 months pregnant, even worse. We just found out we're havin' a girl, at her 20 week ultra-sound. Autumn Ann Richardson is going to be her name.

Anyway. What are some of the combined R(DS) values of some of your guys' h-bridge design? To figure that out, just look up the R(DS) value for both of the N-channel and P-channel FET's you'll be using, and add them. *NOTE* Make sure that you refer to the R(DS) value that corresponds to the V(GS) that you will be applying to the FET. For example, with the 4562's, there is a R(DS) rating of 0.035 Ohms for the N-channel IF you apply 2.5V PWM to its gate, however, if you apply 4.5V PWM to its gate, that resistance goes down to 0.025 Ohms. So gate voltage is very important in determining the overall resistance that the h-bridge is going to present to the drive current that goes through the motor. More resistance means when more current is driven through the channel, more power is going to be needing to be dissipated. If the power doesn't dissipate rapidly enough, then the FET burns up.

I got a combined R(DS) of 0.028 Ohms using IRF3205 and IRF4905 power FET's. I can acquire this because I'm going to be driving the gates using >10V PWM. And if I stack them, that R(DS) goes down to 0.014 Ohms. If I go 3 high, then that goes down to 0.009 Ohms. Just curious what some other people's designs were reaping and with what FET's.

[aaron]

Heya....

I've just been reading through the whole thread and thinking... 'nice setup' with the IRF4905 and IRL3705...

The interesting thing is that the same place ausmicro.com sources our IRF7389s from can also supply 3705 and 4905 in TO-220 packaging.

How much are you guys paying for a set (2 pairs) landed at your door? I may be able to undercut - and have immediate supply. Pay with Paypal or whatever

drop me an e-mail (aaron@ausmicro.com) or PM (here or on ausmicro.com)

A.

How much are you guys paying for a set (2 pairs) landed at your door?

Check out digikey.com, for US costs. Well that's one place but seems to be the cheapest in US as far as low qty sales. About $3.25 a piece for low qty. That's where I get mine. I bought 4 with some other parts paid like 27 bucks to advoid handling charge which whould have been more moiney.

BTW:
Jshwaa congratz on the baby. The 1st 10 years are easy. It's the next 20 you have to worry about

[aaron]

So about USD15 delivered isn't insane.....

Sweet

A.

Originally posted by aaron
So about USD15 delivered isn't insane.....

Sweet

A.

Nope Not at all

[aaron]

Well I have two sets up on ebay right now... If you do a search for the IRF4905 or IRL3705 you will hit them.

Thanks to e-bay and Paypal fees they work out being about USD17.50 at your door within 7 days

A.

[charliebrown]

Just an idea neurokinetik...


These FETs you described can handle up to 55V, right?

Well, I infer that a basic FET will get triggered by a gate signal to suck power from a power source (like, the 4 AAA batteries that plug into the car system), and the drain current will pull from that source. I was thinking, maybe you could reroute the source connections to a different power source, to create more power for the motor...like a REAL turbo. You could, say, hook a 12 volt power supply or a lithium battery supply to the source, and instead of sucking power from the batteries in the car....I'm gonna try this out for myself once I get the correct parts. The motor would be a small stove, but worth every degree celcius.

Now I need a way to get money...and NOW!

[Jshwaa]

Well, I infer that a basic FET will get triggered by a gate signal to suck power from a power source (like, the 4 AAA batteries that plug into the car system), and the drain current will pull from that source.

What I'm hearing is that you'd like to try using a separate power supply that the FET's will be turning on and off to the motor. Separate as in having one to run the electronics of the car, and the other solely for the main motor.

That is exactly the principle behind the relay mod(by Chino). He taps into a series combination of batteries(however long), at just the point where he can run his car without frying it(say in between the 4th and 5th AAA) and then connects relays to the end of the rest of the series combination, thus using the original motor wires as just a signal to turn relays on and off which turns the motor on and off. You can use as many AAA's as you'd like. He's getting speeds over 80kph using this method.

My regulator method works on the same principle as well, although there's no tapping of the series combination of batteries, and the regulator is choking the voltage down to 5V for the elecs. and allowing the "whole" supply to power the motor.

Either method will work in giving tremendous speed. And there's one thing that's common with both methods as far as the wiring goes......the power sources to the motor and the elecs. have to share the same ground. You can't have one battery for the elecs. and the other for the motor. That won't work.

[charliebrown]

Either method will work in giving tremendous speed. And there's one thing that's common with both methods as far as the wiring goes......the power sources to the motor and the elecs. have to share the same ground. You can't have one battery for the elecs. and the other for the motor. That won't work.

What do you mean by "ground"?

Chuckster...