Making an esc better?
 

Hey guys,

This is just for the purpose of chatter and perhaps my gaining some insight. I've been cruising this forum for a few months now and it is apparent that there are some electronics savvy guys/gals here.

I've often come across comments like the Mamba Max would be better if....

Well, I'm sure that the guys at Castle had a product in mind when they conceived the MM controller that would fulfill a particular need. The number of MMs that I've smoked in the last couple of months has told me that my needs exceed the capabilities of the current MM. Of course, the MMM is coming out sometime in the next decade, but I don't want to have to wait.

What I would like to know is whether there is a way to "upgrade" a MM to enable it to run 5S lipo reliably and up the current handling capacity of the MM esc. I like the MM's programmability and its form factor, so I thought I'd ask. If it is at all do-able, I might be inclined to try. I have a couple of guinea pigs sitting on my bench.

Ive not heard it done with castle controllers but in the SAWs (top speed record running) boat world, there isn't a speedo available that will cope with the currents of a sub 10 second runtime, some folks have taken the schultze 32/160 and stacked extra powerboards on to increase the current handeling ability, if the MM has seperate boards for the brains and power it may be possible too.

I don't belive there is any way to increse the voltage it will work with bar desoldering alll the components and replacing them with higher voltage versions(if they fit), but with the time, knowledge and expense that would take you could probably build a new controller to your specs.

I'm not worried about the time, but the knowledge I am lacking in. That's why I was hoping that maybe a parts change would work. Then again, if it were that simple, someone would have done it already.

I just thought I'd ask. Perhaps, it was just a stupid question.

Keep in mind that Castle are working on the MMM. Consider they do this for a living, 8 hours a day, everyday, with a team of qualified engineers and a lot of sophisticated test equipment. Consider how long it has taken them already so far and it is not yet available.

Consider alternatives. :) JMHO...

Quark Plasma!

is it out yet?

No idea...

Well, the wait for the MMM continues. I was also thinking that Castle has a lot more resources than I do, but on the other hand, they also need to build the esc AND keep prices within reason.

I just wanted to know if I could do an upgrade.

There is no reason for CC to work faster on MMM version. And it is very easy to build prototype unit in R&D but it will takes tons of time before it can be released to public.

Can MM be upgraded - sure why not, can it be done within the same package - questionable, can it be done cheap - possible.

Here couple thoughts:
For MM to handle higher voltages you need to modify software as well, that's a little bit more tricky then just swapping fets or adding couple daughter boards. But we can always set it to custom voltage, right (thanks to flexible software).

Do you need high current at higher voltage - good question.
What is more important that with higher current (or the same 100A) and higher voltage (4S+) capacitors are growing in size, so current package is doomed, unless we increase price and let electrolytic caps go home :)

Well, bottom line if you want to upgrade one, send it to me I will take a look and maybe do something about. In the past I've done lots of upgrades to ESC from any brand and flavor.

Griffin,

That is a tempting proposition. I have two MM controllers coming back from Castle. When they get back, I could send one off to you to look at. I have one particular MM that doesn't perform as well as the others and may be a candidate for upgrading. What do you think that cost might be?

Oh, I did check out the specs on the Plasma. Am I crazy or does the esc look more like it's geared for on road use? It handles higher voltage, but only a 100A draw.

AAngel, you must consider that when voltage goes up, amps goes down, given the same wattage.

100amps at 6S is a heck of alot more power than 100amps at 4S. So, higher voltage controllers don't usually need so much current rating, unless you are running a huge motor.

For most extreme brushless users, the limit is a 1515 Neu or a LMT1950, and the right KV version of that motor on anything above 4S is probably not going to draw anything near 100amps continuous.

Also, I beleive that amps rating is per phase, which means its actually alot more than 100amps. The Mamba Max is 100amps in brushless mode, but 1000amps in brushed mode.

Wow, 1000 amps! :dft002: I didn't know that. That must be a burst rating. There is no way that those wires and/or PCB traces could handle that sustained without melting. Not to mention those poor FETs!

Would it be possible two wire to esc' in parallel to a motor?.... or is that just crazy talk. I was thinking connect the battery to both esc', then both esc' to a single BL motor. I'm guessing it would result in a puff of smoke & flames, but it would be interesting to try with a couple of those cheapo Bl systems you see on ebay- or try modding them both together to work as a more powerful system = experiment on them rather than a good MM...... [/crazy idea]

Ok, let me know when ready.

1000A, I will question that but in configuration when 3 phases goes to one brush and other brush directly to battery it is possible, but where you can find batteries and brushes to handle that current continuous, unless you use winch motor and 3 Optima batteries :)

Do not ever try to use sensoreless ESC in parallel!
You can use brains from one and parallel output stages.

Griffin,

What is the result of running two output stages in parallel? Would I get roughly double the voltage and current tolerances? Perhaps I should send you two MMs.

Double the current, same voltage...

Thats the rating...no where in hell will any brushed motor really pull that without turning into a metal glob though. I believe the Novak GTX supports like 760 amps (or so they say...).

Brian correct I was also thinking about replacing output stage with something more heavy-duty, like Direct-FET or D2-Pak 7...
Custom PCB and hopefully MM can drive them, otherwise custom driver as well, which can be done with ir2101/ir2102 or simular.

lol, pretty much redesigning everything except the brains. If the power board is running at higher voltage, won't the feedback pulses be too high for the controller? Or do you somehow clamp them down?

OK, this is getting WAY over my head. All I want (and I don't think I'm asking much) is a controller that will handle up to 6S, though I can live with 5S, and a decent current handling capability that doesn't suffer from heat problems and is smooth at low throttle.

lol, aangel dont worry I think brian and arthur are talking over almost everyones heads.

AAngel, that would be a Quark, besides the heat problem, though that can be cured with the RCM heatsink. It handles 5S (says 6S, but it doesn't really like 6S) and is very smooth.

lol, well given the title of the thread, it was bound to get a little technical. :)

Really, all the good ESC manufacturers should get together and make something that really works well. Castle's programmability and versatility (being able to run brushed/brushless), Quark's smoothness and zero cogging, and MGM's high voltage capability (10s/36 cells!!). Throw in a switching BEC built-in and they could market it as an "All Stars" ESC. And while they're at it, add a port to daisy-chain optional power boards for added current capability. :)

I agree. Aside from the heat and no solid 6s issues, it truly is a winner IMO. I really think that the heat issue has to do with how it is programmed. It is just so much smoother than the MM even though it uses the same FETs and number of them. I think they sacrifice a little efficiency for a better running ESC, but that's just my theory...

so long as schultze was not involved (loose connectors) I'd buy it

I specifically left Schulze out of the list. What would be the point of making the ultimate ESC that suddenly bursts into flames? :)

lmao, but yet so true. Right now I would just be happy with 6s and the ease of the MM programming. Anything more then that is just a bonus for me.

That would really be awesome, and then all the companies can just split the earnings between each other.

Yeah, I thought the Quark would be the ticket. I can live with 5S, but when I started reading about the heat problems, I wasn't so sure. I consider excessive heat to be a sign of a bad design, especially in light of the fact that the esc won't do 6S like it's supposed to. I don't mind doing the mod to help keep it cool, but that would render it a one shot deal. If it blows, you're just about screwed.

I figured I'd sacrifice some smoothness for reliability and cool running, so I opted for the 16016 Compro. I bought it almost a month ago and still haven't been able to run it. After a week of trying to program it, I decided that it wasn't me and on Mike's recommendation I sent it to him, so that he could make sure that I wasn't screwing something up. Mike's had it for over a week now, but I haven't heard from him. He must be busy because he hasn't responded to my emails inquiring about it. I'm starting to get the feeling that the Compro is going to have to go back to MGM for repair, which really sucks. This is one of the reasons that I'm going to start ordering everything from my LHS. At least if I get something defective from there, they'll swap it out for me. It's what they've done in the past.

So...I've been running my MM controllers. Apparently, the truck that I'm running is much heavier than most are using the MMs in and the controller is having a hard time. It's a Losi Muggy running big 40 series masher tires and goes about 13lbs with the 8Ah pack. I've blown three MMs running them in the Muggy. Two on 5S and 1 on 4S.

These experiences have lead me to my esc wishlist, which I stated above. I've gotta tell ya, I'm so close to trashing all of this stuff and going to helicopters it isn't funny. By the time I get my Compro back, I won't need the esc as the MMM will probably be out. Either that, or the Teken MT. Geez, I've been whining a lot lately.

What about some sort of a mechanical controller? Like an AC light dimmer switch controlled by a servo. OK, I know I'm reaching now.

As for all of these guys making one good esc, I'd love to see that happen; but you know that it won't. What would be nice is if some new, small, fresh blood firm came about and designed what we were looking for from the ground up.

BTW, has anyone been able to figure out why it is that the Quark won't do 6S?

Like I said before, the Quark seems to sacrifice a bit of efficiency in favor of smooth running. To me it's worth it. I'd much rather add a little extra heatsinking than have an ESC that stutters, but that's just me.

If you are running a Neu on the MM and don't mind a little experimenting, try adding an extra capacitor or two to the battery inputs on the MM. They have to be low ESR types. Figure on two 220uF and 35v if you can fit it, or one more 330uF if not. Solder it directly to the + and - battery solder pads on the ESC. I'm going to try this someday, but I don't have a "bad" running system to see if it helps.

I suppose it can be discouraging to spend all kinds of cash on something that doesn't work up to expectations. Stick with it if you can because once you get it running, you'll be happy.

Mechanical controllers would be a bad idea. Those are usually servo driven and rely on resistors to drop some of the voltage to the motor so it goes slower. And dimmers are different; they use triacs on the 120v AC line and chop some of the sine wave off so the light/fan gets less power (I can go into more detail on this if you want, but don't think you really want to hear it, lol). They rely on a sine wave though, so it wouldn't work in R/C. IMO, the use of many paralleled FETs is the way to go - it's just up to the manufacturer to use them appropriately in their designs.

I do have to give credit where credit is due. From what I could see, inspite of not being able to program it, the Compro is going to be cog proof. It starts a bit abruptly, in that even easing on the throttle results in the vehicle just going, rather than easing into a roll; but it never cogged, no matter what the gearing was and it ran down right cold.

I could solder up some caps on the MM, but what would this accomplish? My assumption would be that it would help out the batteries, if they were unable to supply the current demanded of the system, but I'm running Kokams and they have no problems supplying current.

And yes, it has been discouraging. This whole ordeal with the Compro has about assured the death of my passion for rc cars/trucks. I would have just ordered another one to use while I was waiting for mine to get fixed, but something told me that it would be stupid to do so. If I do stick with this stuff, I'll be ordering all of my more delicate equipment from my LHS; and if they can't get it, I won't mess with it. I found out last week that they can get Quarks, so at least I'll have a line on the S&T, Castle, and Tekin stuff. I'll probably pay more for it, but I'm willing to pay for the service.

Soldering extra caps on the power line (as close as possible to the PCB) may help with the Neu cogging issue. The idea is that the Neu is a highly induictive motor and needs more to absorb the back-emf pulses. If you have a Neu and some caps, great. If not, then I wouldn't bother.

I am getting some blown ESC's from Mike, so updates will come later...

Griffin, while you are at it, can you look at the possibility of hooking up another MM in parallel to the master so you can have double the capacity and cooling? Of course, the "brains" of the slave would have to be disabled/disconnected...

If I get MM for testing then sure why not.

I guess it could be done with any controller, not just the MM, but it looks like the MM would be the easiest the way it's built...

What I'd really like to see is a tiny controller with just the brains (with USB programmability) with a short umbilical to power board modules. Keeps the power and heat away from the brains and maybe cheaper/easier to replace a burnt power section. I doubt anyone will ever make a modular ESC like that though...

Well, once my Quark gets here, I may just have an extra MM to offer up to the RC gods.

Thats the future plan, like industrial controllers, 6-pin header with opto-isolation and no limits.

So, you did end up getting the Quark. I thought you were going to wait to see about the MGM warranty thing? Maybe I missed a post somewhere...

Griffin; yeah, that'll be nice. Probably be at least a few years before that comes about though...