upgraded FETs - Page 2

neurokinetik · #16 11-13-2003, 07:06 PM

Well, experiment phase 1 is a success. I replaced the stock FETs with IRF7389s, and everything works!

Now it is time to remove them and begin phase 2...

neurokinetik · #17 11-13-2003, 09:40 PM

Well, phase 2 is built, but not working at this point. Time to troubleshoot. The FETs I have chosen may not have the proper characteristics for this circuit.

Here's some pics showing what I have attempted.

The first pic shows the add-on board with the new FETs:

The second shot shows a closeup of the soldering of a CAT5 wire to connect the ESC to the add-on board.

The last shot shows the entire electronics of the XMod with the add-on board attached.

mbartell · #18 11-14-2003, 10:13 AM

great to hear about stage 1!
I can see where stage 2 would be a problem.... that's a lot o' wires.
How do the mini z turbos do it, with only 4 wires?
They must be a lot more complicated than they look.

neurokinetik · #19 11-14-2003, 10:25 AM

Well, it depends. The "twin turbo" setups probably use 8 wires as well. Most of their turbos are only for the forward direction. I could concentrate on that as well, if I wanted.

What's important is that I have confirmed for myself that the main drive circuit of the XMods ESC is virtually identical to that of the Mini-Z. That opens up alot of possibilities.

BTW, the wiring isn't that bad at all, it took me about an hour and a half to build that entire setup from start to finish.

Anyway, if I get this working, it won't really be a stage 2, it'll function more like a stage 20...

FESCSteve · #20 11-14-2003, 11:07 AM

Is there no room to stack the IRF FETS?

mbartell · #21 11-14-2003, 11:09 AM

nope, they would stick thru the bottom of the chassis. Good for airflow, but bad for anything hitting it.... Unless you flipped over the board, or re-oriented it in some way.

neurokinetik · #22 11-14-2003, 11:57 AM

Flipping the board might be a possibility. You could also use a decent heatsink on top of them if you did this. I've got the perfect one from cooling cache RAM chips on overclocked Slot A Athlons. I have six of them from that project, and one of them will exactly fit the two FETs side-by-side.

Nathan · #23 11-14-2003, 10:41 PM

The mini-Z turbos send power directly from the batteries to the motor. They connect to the stock FETs and are simply a set of 4 more powerful ones. This way they are easy to install and bypass stock electronics entirely. I'v been thinking of building a simple one with RadioShack MOSFETs. This way I don't need to solder or mess with the PCB, just solder wires to the batteries and use those black connectors for the motor and motor output.

neurokinetik · #24 11-14-2003, 11:51 PM

Quote:

Originally posted by Nathan
I'v been thinking of building a simple one with RadioShack MOSFETs. This way I don't need to solder or mess with the PCB, just solder wires to the batteries and use those black connectors for the motor and motor output.

You'll still need to connect to the board... How else are you going to get the signals to the MOSFETs to turn them on? Also, check the Specs on the MOSFETs that RS sells. They are IRF510, which are only capable of handling 4A continuous. The IRF7389 handles almost double that, and the MOSFETs I am using handle 10x what the IRF7389 handle.

Anyway, I've been doing some testing, and it appears that my circuit is fine, but one of the two types of MOSFETs I chose is not suitable, and I think it is the N channel. The circuit is not working because the drains on the N channel FETs are not staying at ground when they are supposed to, while the P channels go to 5V. What you end up with is 5V on both the positive and negative sides of the motor, and no voltage difference means no rotation.

What I have learned about this motor drive circuit:

IC5 controls the reverse voltage, and IC6 controls the forward. When you want to go forward, the gate on N channel FET in IC5 goes to 3.5V (not enough to turn the transistor on), and the gate of the P channel FET in IC6 goes to ground, which turns the transistor on. The output (drain) of IC6 goes to 5V, and spins the motor forwards.

The issue with my circuit is that I need to change my N channel FETs to logic-level ones, so that they are not turning on when they are not supposed to. Once this is done, the circuit should work just like the stock one, but with more than 10 times the current capacity.

Here's a couple of pics of the revised circuit. It's down to seven wires now, as I noticed that pin 1 on both IC5 and IC6 are grounded.

SKYHIGH · #25 11-15-2003, 11:43 AM

I am very new to xmods and mini-z tech stuff. I read for quit a while last night about stacking the FETs, but you have a seperate board and is that what the MOSFET your are talking about is? Also where are you going to install the extra board

neurokinetik · #26 11-15-2003, 04:56 PM

Quote:

Originally posted by SKYHIGH
I am very new to xmods and mini-z tech stuff. I read for quit a while last night about stacking the FETs, but you have a seperate board and is that what the MOSFET your are talking about is? Also where are you going to install the extra board

The little 8 pin chips that the speed controller uses are the MOSFETs. Each chip contains a pair of MOSFETs, one N channel, and one P channel. Most people parallel a bunch of these chips by soldering them on top of one another to maximize their current capability. Instead of doing that, I am using much higher current devices that replace the chips with two MOSFETs each. These devices give the same or better capability as stacking up 10 of the small chips, which would be both unwieldy and not really possible where the board is located, even if you were to flip it over (the driveshaft ends up in the way). The small board will probably be attached to the roof of the car, like the light module is.

Obviously, with this setup, you won't be able to run the light module, but if you are really going for maximum performance, you aren't running the lights to begin with.

SKYHIGH · #27 11-15-2003, 05:02 PM

With the seperate board and FETs would you be able to run a 5 or 6 cell set besides the lack of room

neurokinetik · #28 11-15-2003, 05:10 PM

With this setup, I can run up to 55V.

Well, at least the motor drive circuit will be able to handle it. The rest of the board OTOH... :lol:

Actually, if I replace the steering servo FETs with IRF7389s, I could run as much as 30V with little risk of frying things.

Microfied · #29 11-17-2003, 01:02 PM

How much will this FET upgrade cost me and where can i purchase them?

FESCSteve · #30 11-17-2003, 01:39 PM

What is the stock FET rating? The reason I ask is that I am running 8.4V on a stage 2 yellow motor (next to highest speed motor - lots of current draw) and the controller seems to be taking it fine.

At this voltage, I can't drive it through turns as fast as it can go - it simply does flips. I wonder what the gain on adding the higher current capacity fets will be? I know that the on resistance will be lower, but you will probably negate any benefit with the long wires to the FET board. And I can't see anyone wanting to go much faster than 20-25MPH with one of these.

Of course, there are the hard-core speed freaks out there I suppose, but you reach a limit where the power simply is not useable abymore without increasing size or weight to keep it on the ground.

I guess I need to take some current measurements to get a baseline. Remember that as you increase voltage, the current goes up too.

Good luck on the project though - I am interested in the final product.

S