Cogging...

[portyansky]

What would cause cogging? Would it be from over geared, under geared, or either or?

Medusa 2000-60, MMMv3, Tekno Revo 17/36 (gonna go even higher with a Slayer tranny with CD). Im not sure if its cogging or not. I havent realy had a chance to test it out the way I want to. As soon as we get some better weather, I will know more. But still would like to have an answer to my above question.
Thanks in advance,

Porty

[TexasSP]

Most of what people think is cogging is something else. A good description of what is happening would help us help you.

Without that it could be, poor connection, sub par batteries, drive train binding, true cogging (doubtful), over gearing, wrong programming on MMM, bad motor, etc.

[BrianG]

^^ +1. As far as I'm concerned, true cogging is very undesirable because the motor is somewhat stalled and very large currents are possible.

However, the MMM with firmware 1.20 does what looks like cogging, but is really more like "chattering". The main difference is that there isn't that huge surge of ESC/battery killing current. If the gearing is too high where a low throttle input does not allow the ESC to get a proper back-EMF reading from the motor, the ESC seems to try to go, but backs off before things start to get bad. This repeats quickly and sounds like cogging.

The two ways I can think of to reduce or even eliminate this would be to gear down and/or increase the ESC "start power".

[What's_nitro?]

OT: That term "chattering" interests me, Brian. Does that mean I can't burn up my MMM by hitting the throttle too hard from a standstill?

[BrianG]

Well, I wouldn't want to put it to the test, but I do believe that is the purpose.

The algorithm is much "smarter" with v1.20. Before v1.2, if I put the the truck on the bench, set the throttle trim on my tx so the ESC was going forwards slowly without me pressing the trigger, and I stalled the motor by grabbing the wheels (not advised without gloves - trust me), there would be a surge of current while the ESC attempted to send more and more power to get it to turn. If the motor was seriously stalled in the real world, this could lead to a "bad" situation. However, with v1.20 fimware, stalling the motor does NOT cause large amounts of current, just a little chattering while the ESC tries a little to go, sees it can't, and lets off the power, and then tries again repeatedly. Much different and much better IMO.

Do I recommend gearing really high and seeing if this will prevent that? No. That's just tempting fate.

[What's_nitro?]

Does the MMM have an internal current limiter?

[BrianG]

IIRC, I do not believe so, but Patrick has said there is a thermal shutdown. But if the increase in temps happens too fast, this circuit doesn't have a chance to shut things down before catastrophic failure.

Besides, a current limiter would have to measure current to work right, and that would involve either measuring voltage across a shunt, or some type of feedback device, both of which would introduce resistance, size, and expense into the circuit. The only other way I can see to do this without those losses would be to monitor the input battery voltage and back off the throttle if there is too much of a dip in voltage.

[What's_nitro?]

I can't find a picture (because they don't make it anymore), but I have an ammeter for my car stereo (made by Autometer) that uses some type of inductive coil around the main power wire to measure the current. It doesn't add resistance. It would add size/complexity/cost, however.

Monitoring voltage drop in order to limit current would be tough due to the variance in C-rating of different cells.

[kostaktinos_mt]

Originally Posted by What's_nitro?

I can't find a picture (because they don't make it anymore), but I have an ammeter for my car stereo (made by Autometer) that uses some type of inductive coil around the main power wire to measure the current. It doesn't add resistance. It would add size/complexity/cost, however.

Monitoring voltage drop in order to limit current would be tough due to the variance in C-rating of different cells.

exactly... something like the eagletree or equivalent... if integrated into an esc, it would make a nice blackbox [logs for everything] and would help the main software to know when to stop pushing [to prevent damage] or when to punch hard [to prevent 'chattering']... not to mention that being capable of adjusting the average current drain to your likings is what would be a REAL torque-clutch... 100% usefull for racing, as traction would be dialed without the inaccurate time-delay routines that are used today, and without sacrificing topspeed too...

i am not sure if the feedback circuit would be fast enough to catch up with current surges or voltage spikes though... these things happen really fast, and the sampling frequecy will need to be equivalently fast [eagletree does 10/sec only], probably to the scales of pwm fet switching [8khz and up]...

i think this step will be realized sometime; just as it happened with voltage monitoring of the input [mandatory for lipo application; wasn`t needed for nimhs back then]...

[BrianG]

What's_nitro: An inductive clamp (which is what you're talking about) probably could be used, but I'm not sure if it would be small enough to fit without taking up too much room. In these vehicles, weight and space are at a premium, which is why these ESCs are so small/light to begin with.

kostaktinos_mt: I believe Eagletree does use the shunt resistor method. No biggie just for playing around, but I don't think I'd want anything that introduced losses into my rig. I do agree it would be nice to have something like that, but I'm sure the price would reflect the additional circuitry. The time-delay, or integration, of the punch control setting is not perfect, but sure is better than nothing. And I do agree about the clock speed of the device needing to be fast enough to measure those milli-second pulses.

Doesn't eagletree sample faster than 10X/sec? I think that's the default, but IIRC, you can speed it up but at a loss of storage room.

[kostaktinos_mt]

hi,
sorry for the delayed response; it`s a long story...

brianG, the eagletree uses a hall sensor [not a shunt].
http://www.eagletreesystems.com/MicroPower/details.htm [4th line]

the max sampling rate on my v2 was 10hz, and i think it`s the same for the v3 [read the page of the link above].

i don`t think a hall sensor is expensive or physically large either... maybe a hall-sensor-equipped esc will probably need a stronger processor if a serious sampling rate is to be fully implemented along with the mathematics involved to close the feedback circuit fast enough [output current adjusting pwm correlative to throttle input; while at the same time looking for rotor position and all other stuff for proper commutation].

don`t know... maybe some will say it`s useless, but imo, a real current limiter [current controlling closed branch actually] can really help in a LOT of things [from reliability to traction control and efficiency tuning]...

on the other hand, brushless tech in rc is still a baby the way i see it... taking a look on what was happening 2 or 3 years ago and contrasting it to what happens now, makes me believe that...