bitPimps.lixlink.com

interesting man, what you thinking? a duty cycle sweep? some sorta overdamped setup?

ph2t wrote:interesting man, what you thinking? a duty cycle sweep? some sorta overdamped setup?

The former. I figure I can put an 8 pin SOIC microcontroller with internal osc. in there, connect it to the VCC and VSS pins for power, then insert it in place of those D9D transistors so it intercepts the control inputs and run the forward motor control with a quick PWM duty cycle sweep when going either from backward to forward or stop to forward.

I figure about a quarter second or so ramp from zero to full power should avoid most of the back-flip action. Might also need to check for full reverse to stop or forward and ramp those too (sometimes I flip when going full reverse to stop).

I haven't quite worked out the control scheme yet though. I want to try to keep track of what the last control state was (reverse, stop or forward), and how long it was in that state, so I can make a reasonable decision about how much of a ramp it needs to perform. Going from high-speed reverse to stop would require a ramp, but going from low-speed reverse to stop would not. Going from high-speed reverse to forward might be complicated. But then I always try to overcomplicate things.

I will have two more control inputs, I could use one for a enable/disable switch so the chip could be turned off, and the other I could use for a trim pot. The trim pot would allow me to control how steep the ramp is. It would be cool to adjust the acceleration amount such that it could hold a 45ish degree wheelie until it got to top speed

Anyway, for the first version I'll keep it simple, so I can get a feel for what it needs to do. Probably just backwards to stop and stop to forwards ramps of about a quarter second.

codesuidae wrote:I'm going to try to put in a circuit to tune down the torque on startup, make it a bit easier to run a fast motor on high traction surfaces without flipping over. That'll give me another chance to take a look at that label.

Kinda like an electronic slipper clutch?

That might be a good idea, these things get pretty silly with the faster motors.

Ideally it would have to be adjustable. That would be so cool to dial in the amount punch off the line with the turn of a little pot.

crazydave wrote:Kinda like an electronic slipper clutch?
[...]
Ideally it would have to be adjustable. That would be so cool to dial in the amount punch off the line with the turn of a little pot.

Yup, thats the idea. The main problem is that without an idea of how fast the wheels are turning, its a little difficult to guage how much power should be delivered to the motor at a given time. The trim pot seems to be the best solution there. Since the board isn't propo the controller can make some reasonable guesses as to how fast the wheels are turning based on how long the motor has been on, and that can be scaled by the value of the trim pot.

This is unedited XviD compressed footage of a FET modded ZZMT on a wood floor.

The clips are about 3 and 4 minutes long, low bit-rate, no sound. Intended to show how the MT behaves with a high speed motor and FET mod. Watch carefully and you'll see a number of interesting tricks.

As unedited video, these aren't terribly exciting, but I wasn't going to waste much time coming up with an edited version.

Try this and this.

Let me know if the links don't work.

codesuidae wrote:

crazydave wrote:Kinda like an electronic slipper clutch?
[...]
Ideally it would have to be adjustable. That would be so cool to dial in the amount punch off the line with the turn of a little pot.

Yup, thats the idea.

Sorry about replying to myself.

Is this a stupid idea? Should I just put on a wheelie bar and forget the electronic control?

meh - in my opinion, the getting it done part is genious - the amount of work involved vs the end result don't seem to balance, but...
that's not to say it shouldn't be tried or investigated. I mean how many things new and cool have come about from screwing around with a few simple ideas?
screw it - I say go for it man, prove that it can be done and show us how.

CaboWabo wrote:meh - in my opinion, the getting it done part is genious - the amount of work involved vs the end result don't seem to balance, but...
that's not to say it shouldn't be tried or investigated. I mean how many things new and cool have come about from screwing around with a few simple ideas?
screw it - I say go for it man, prove that it can be done and show us how.

Yeah, but what might sound complicated to us, might just be weekend of fun and excitement for this guy. He's obviously one smart cookie.

yeah, these bloody techo freaks, I just don't trust the dodgy buggers.....


croikey!!


ph2t.

Yeah, ph2t, they got them shifty eyes.

Did one with stacked FETs.Here's the top.


...and the bottom.


Stacking the FETs made a difference, but really it's too much torque by this point.

crazydave wrote: Stacking the FETs made a difference, but really it's too much torque by this point.

By this I presume you mean you tested with single FETs, then again with stacked FETs?

What ere the performance figures for both trials?

I bet code that in the ZZMT case, stacking mosfets would have an even greater difference between single and double stacks vs. the same setup in a BCG or ZZ.

The Vgs drive is 3V in a zz/bcg. This is a borderline mosfet drive at best.

Stacking mosfets increases the input gate capacitance between the Gate pin and Source (known as Ciss on most datasheets). This decreases the already borderline drive's ability to charge the mosfets.

In theory, mosfets don't enter their "enhancement mode" until the input drive has gotten over this initial "hurdle" so to speak. Once the Gate pin is fully charged, it then opens the bridge to allow current to pass from drain to source (nchan example).

Given all this, practical testing on bcg pcb's shows that whilst there is a difference between single and double stack, it IS marginal. A triple stack though is MORE noticable than a single. See here for more info on that: http://bitpimps.lixlink.com/pages/phpFo ... php?t=3594

So if all this is possible with a crappy 3V gate drive, mate, the 4.8V gate drive the ZZMT provides will even perform better, and should show a greater margin of performance difference between 1 - 2 stacks, 2 - 3 stack, etc, etc....

I can solidly state that their is a clear relationship between increased torque (however measureable) and lowering the Rds(on) value of the mosfets in the drive train. Be it through mosfet stacking OR the substitution of lower Rds(on) mosfets.

Using lower Rds(on) mosfets is the better peformer than mosfet stacking though. Evidence (in propo systems) has shown that throttle response is cleaner on non-stacked mosfet h-bridges vs. stacked mosfet h-bridges....

cheers,


ph2t.

oh yeah, excellent work dave! impressive!

Stacking the flipped mosfets is a pain in the arse, I commend your patience on that one man...lol...

ph2t.

ph2t wrote:oh yeah, excellent work dave! impressive!

Stacking the flipped mosfets is a pain in the arse, I commend your patience on that one man...lol...

ph2t.

I commend all of you.........

...........for MAKIN' my brain hurt!!!

I am officially lost with all of this "gate" and "V" and "Rds" and "N-channel" this and "P-channel" that and a bunch of other terms.

Could someone make a SIMPLIFIED list of terms and their applications towards RC cars, and put it in the resource section? I think it would help alot of people understand a little more of what is going on with all the mods that are getting performed.

I'm getting the ideas here, but you guys always lose me with the pictures and the terms.


Whatever self-sustaining device your working on...Keep it up.

He lost me with the Rds.

I don't have a speed checker so I can't judge performance like that, and honestly, I don't know how those speed checkers work, but i don't feel top speed is relevant to me, since I rarely use it, I'm more interested in the low-mid power band.

Anyrate, I go by the, "If it makes Dave happy meter", and stacked FETs definately got me excited. The difference I'm seeing, is before with Ornage gears I'd have to back up before going into forward to pull wheelies. With the stacked FETs I can see it lifting the wheeks slightly coming from a dead stop. So I know it's doing something, and now that ph2t mentions it, I never could see much difference when stacking FETs in bits.

This one I did for someone else, and I just sent it off, but I'm gonna do up my second monster in the next day or two. So if you want me to get any reading off it to compare, just let me knoe where to get them from.

And thanks ph2t. .
I'm always proud of myself when I do a job like that. I stick them in my wife's face and say, "Look at that, that there is some good soldering", and she just rolls her eyes at me.