MGM article on voltage under braking & lipo quality
 

Good article, at least for me. See attachment.

Good find! The results make sense and reflects what I have found in some personal tests. The solution to use an ESC with at least 1s-2s higher rating than the voltage used is something that I currently do, and has worked well. Actually, it is a good practive with anything electronic for greater lifespan.

But I guess it depends on the manufacturer too; do they rate their ESC voltage/current close to the max component rating to increase sales, or do they rate their ESCs with some "headroom" already built in? I would assume that the more expensive ESCs fall into the latter category, while cheaper ESCs fall in the former category.

Option b, using a controlled shunt device, is what the MMM uses. However, this is not a perfect solution because if they are overworked and fail (fail open), the user most likely will never know since its presence isn't obvious during operation.

Super informative! Much appreciated :yipi:

"sneaks off to order better connectors and shorten his cables".

Now that we are at it... where do you get super connecotrs at a reasonable price?

They even say I shouldn't use Deans! :oh:

Well, what now? Those super-connectors got super-prices...

What I don't understand is that at the bottom of page 5 they recommend 6s2p packs to lower the inner resistance.

I just looked up the inner resistance of the hyperion g3 35c packs:
3s4200: 4milliohm
6s2200: 14milliohm

Wouldn't two 3s packs in series have a lower resistance than two 6s packs in parallel(6s2p)?

Actually, they wouldn't. Two 3s4200 packs in series yields 8 mohm (4 mohm + 4 mohm), while two 6s2200 packs in parallel yields 7 mohm (14 mohm || 14 mohm).

If i am reading that right, Then you should figure 1/3 over the working voltage you plan to run. I did see something similar in my eagletree runs, not in voltage but in amps. One or two runs showed 200+amp spikes where it was braking, this was on a 9XL fegaio. I thought it was a glitch and deleted it, wish i had kept it to show, those runs were showing 120 to 140 amp spikes on acceleration on a V2 micro eagletree. I never seen those spikes witth my Neu though, Is it possible that the quality of the motor has an effect?

Maybe, but I would tend to think that the Neu is more efficient so less power is needed for the same performance? Also, Neus, being 4 pole slotted, has more torque.

Weird though that a manufacturer is saying you can't use their esc's at the max rated voltage if you're running motor brakes. Have there been any problems with mgm esc's that anyone knows of?

Well, before I got these G3 35C 4200 packs, I have been using a 7s1p, and 7s2p A123 packs in my revo running a 1512/3D/F on a Quark.... and when I modified the Quark last week I noticed that the internal caps leads were blown/melted right off the board! Only the externally soldered on caps were doing something!

After I did the cap bank mod, I for surely found the brakes to be different, the motor sounded different under braking (think the buzzing noise).

So now with this Quark mod, and some very low resistance packs (G3/35C), it should be golden now.

I never actually looked at the whole article, were they talking about ESC destruction and/or battery damage? I am assuming the former.


BrianG, I am wondering if I should make another larger cap bank on the Quark in a reciever box (asthetics) with the 'no-spark mod? It can't hurt right, lower the system resistance further, can only be good for everything I would think.

It's possible that this is a recent discovery, or one that was known but not proven till recently. Like i said i thought mine was a glitch

BrianG, I forgot i also switched to a quark esc as well. Is it possible that the quark has the same gate like thing as the castle monster?

MGM seems to be weird that way... they sell you a overpriced product, and then they make articles and disclaimers about their product. :oh::lol:

crazyjr, I believe the eagletree doesn't read/record negative currents, I think it must have been either a glitch or just a current peak.

Electrically, I'm sure any bit helps. But realistically, you run into placement/mounting issues. Also, if the caps are soldered somehow to the ESC, the weight of bus bars and the caps themselves would place considerable stress on the PCB. And, at some point, you get very little return for the work/expense. If you already have twenty 500uF caps (10,000uF), adding 4 more to get 12,000uF isn't helping a whole heck of a lot.

I'm not sure, but I don't remember seeing a TVS device on the Quark. Of course, I never looked specifically. The ones I've seen are roughly the size and shape of a typical FET that you'd find on an ESC (square/rectangular and flat), and it has two leads.

It might not record currents below zero (has anyone even tried setting the graph low point to a negative number?), but there should be a voltage peak that may be at or above the battery voltage.

I've got 6 caps on the Quark as it is, mounted very sturdily and safely. I was thinking about putting 10 cap bank in a receiver box wired up to the Quark with short wires. But, like you said, probably not necessary... as 2 caps has held up this long, and the 6 cap bank made such a difference.

http://www.rc-monster.com/forum/show...uark+regenesis

The eagletree micropower logger does record positive amps (not negative) with an increase in battery voltage while braking.

http://www.rc-monster.com/forum/atta...1&d=1246396048

Latest log recorded yesterday with 8s2p A123 while making speed passes. Notice the amp spikes (20-25A) while the green rpm curve is falling. I never recorded amp spikes as high as described in that article.

Basicly they say use an 8s ESC when running with 6s lipo and keep resistance as low as possible (connectors, batterys, cables ...) to minimize voltage spikes while braking.

Does the ET truly not log negative current, or does it just not show it because the graph starts at "0"? Could someone set the min value of the graph at some value and see? I could do it, but I'm at work :shhhh:

Interesting there, I have never seen a positive voltage rise after braking, but then again I do not brake that often, and I do not pound on the brakes, I'm quite easy on them.

Don't forget adding caps to the ESC, I am not sure how many you would need to reduce these voltage increases to a decent level, but probably you can't have enough... you would need something of capacity with very low esr.

I will do some tests sometime soon with my Revo under hard braking and the eagletree, see what we get for comparison.

I found a better graph i posted some time ago (while i was running 10s1p A123) where i pointed out braking Amps and the corresponding rise in voltage. The green curve depicts motor rpm.

http://bt2007.braintrust.at/ds/rc/ma...remsdetail.gif

@BrianG
When opening the eagletree logfile with excel you won't notice any negative value. There is no way you can differentiate between amps flowing to the motor and amps flowing to the battery when using an eagletree logger (other then looking at the voltage).

@zeropointbug
The Kontronik Jazz 55-10-32 ESC is rated up to 43V (10s lipo). With 8s A123 (25V average) i'm well below it's rated max voltage leaving enough headroom for braking voltage spikes. Even with 11s A123 (36v average) i do not have issues while braking with the ESC. I too tend to run my ESCs below their rated max specs to avoid premature failing.

Yah, I have ran 7s1p and 2p packs for 2 years now through my Quark with no issues... and when I got my G3 lipo packs, it started to cog alot, and do weird shutdown hiccups. This is an ESC that should never have even been designed for 6s lipo, according to alot of ppl on this board, including myself.

After I did the modifications to the esc, it has never ran so good, and I mean awesome. We will see how it holds up in the years to come....

good find; but honestly, what I can't understand is that these expensive esc's are not free of trouble. I have one of these cheapo-plane esc's, modded with some bigger caps. never cogged, no matter how bad the abuse is, it works.
so when I see people spending 200,- for just a controller, which burns after 2 weeks, I wonder if that stuff is worth the money...

they say, expensive must be betta!!! sure??:neutral:

Myself, I don't care as I could not spend that money for the controller...:tongue:

I think part of the deal with MGM is that each controller they make hardware wise is generic across the board and the programming is what makes it car, boat, or plane. Whereas CC and others make specific controllers for specific uses. I believe it's silly to do it this way. They might as well just down rate the controllers when being used for cars versus planes etc. One thing I did with my MGM though is run an external bec. The firmware and interface still left a lot to be desired, and it heated up quite a bit more/quicker than my MMM. It was a solid controller though although I do not feel it was worth the extra money for me.

That's interesting because most of the reports I've seen say that MGM actually heat up less. On second thought, I don't see why it would be either way, most ESCs use similar, if not the same, FETs. And heat comes from the rdson value and slew rate of the FETs. I'm sure some comes from copper losses too, but again, they are all built basically the same way. Curious.

Yah hehe, built in a similar fashion, but rated completely different I know! Example:

MM vs. the Quark controller

They use the very same FET's yet CC was wise enough, and not cheap to rate it at 3s lipo and 100 amps... Quark was TWICE cells rating and 25% higher current rating, and we all know that when both controllers are run on 4s lipo and same setup, they run pretty much the same temp, if not, the MM being cooler. And yes, the Quark has some higher voltage caps (a dollar extra cost IF that), so why does one controller cost twice as much as the next one? With the same, if not worse build quality, same core components, etc....

But, I can't complain too much, as my Quark is working wonderfully on 6s now.

Quark is super smooth, compared to the MM back when it first came out. Quark has a nice metal case, compared to the plastic case of the MM. Quark fet board is heatsinked on both sides, whereas the MM only on one side. I personally liked my Qaurk 80B in speedruns over the MM. But I just recently blew the Quark up with some overvoltage. But when it comes to power, Luciano's 1/10 ESC is mini monster compared the MM and Quark when it came to power capability.

I have some 10S A123 datalogs that shows the voltage rise during braking as well. It's harder to see in my lower voltage setups.

I don't think some poopy heat pad/spreaders on each side is good... the top FET's are not heatsinked, only heat spreaders to absorb the peaks in heat output from the FET's.

hey if you cant tell the difference good for you. I tried a million different cheapo chinese air escs before buying a monster. I can't stand the motor startup algorithm used in air escs. it's optimised for a propellor driven vehicle and isn't made to stop and start like cars often do *shudders* ill never go near another air esc.

air esc in a surface R/C = crap

a more concise version of the point I was making above :lol:

With the right software, why wouldn't an air-esc be just as good as any other car esc?

as I understand it, the loads that a surface vehicle puts on the esc are very different to what an aircraft does. if it had car software so it responded like a car esc, it would let out the smoke a lot more easily.

OMG so basically:

a^2 + b^2 = c^2
16^2 + 80^2 = c^2
256 + 6400 = c^2
c^2 = 6656
c = 81.58"

16^2 + 48^2 = c^2
c^2 = 256 + 2304
c^2 = 2560
c = 50.596"

80^2 + 48^2 = c^2
c^2 = 6400 + 2304
c^2 = 8704
c = 93.3"

Max. Length
d = sqrt(a^2 + b^2 + c^2)
d = sqrt(80^2 + 48^2 + 16^2)
d = sqrt(6400 + 256 + 2304)
d = sqrt(8960)
d = 94.657"

Answer to the Ultimate Question of Life, the Universe, and Everything is 42. why didnt the article just say that at the beginning? OMFG!!!

well, I know that there is no such thing like a proper break in these things. why? dammn, the're made for planes... comments like that= crap.

thats why I will get myself a proper car esc. there are some new ones coming out right now here in germany, and when they have proven to be good, I will get one.

BUT:
the reliability and the cogging-free throttle is - at least at the one I have - just great, nothing less. So I wonder, why can't other esc's of big brands be at least the same reliable? I've payed about 40,- euro and only water was able to kill them, indestructable stuff.

That is something I can't understand, and that was the point of my post.

florian

its not the brakes that is the problem. Go and do some jumps. You will notice sometimes if your throttle isn't perfect, the nose might dip a bit in mid air, and to correct it and land on the landing ramp smoothly, you blip the throttle just a tad and the gyro effect brings the nose up a little to stop it from nose diving.

Using a car esc, this is possible, because they allow the motor to spool up quickly enough to give a fast rpm change like that. This fast rpm change is when peak current is drawn - on the road in a truck castle measured peaks in excess of 800A. These peaks arent visible on eagle tree etc because they are only momentary, but as far as the electronics are concerned, they still happened. The FETs in the car escs can handle thes momentary rushes of current.

In an aeroplane or a heli, you dont want the prop to spool up to full speed in a fraction of a second, so they limit how quickly in can spool. in the same jump scenario above, you blip the throttle to bring the nose up and you've hit the ground on your roof before the rpm increases significantly. This way the Air escs can use much lower rated FETs because they dont experience the extreme current peaks that car escs do. This is also why they dont cog, because they are designed to spool up slowly, because thats what you want in aircraft. this also has the effect of much lower peak current draw.

I have some graphs showing a voltage rise w/ my MGM 9032 and 10A A123 a while back. Nothing huge, but noticeable.

I always loved my MGM. Tons of power and it barely ever got warm. 110F at most. HV did help a lot. Very solid controller, water proof/resistant coating also.

It should be understood too that at the time, there was very little choices for controllers. You had the cheap MM that worked pretty well for 4S (and was a relief when that came even,) a pricey Novak that thermaled easy and wasn't fast, but if you wanted more than that you had to get a $$$ controller. Things have changed really quickly. Go back to threads from even 2007 and see how it was. 2005 may well have been the BL equivalent of the 1920's Flatheads for gas powered engines.

thanks for that reply, very interesting.

to adjust the car during "flight" is truly tricky, as especially the breaks don't react immeadetly. by now I got used to it, almost. but throttle-response I would actually compare with the GM genius controller I used to have. My car is a truggy-style buggy, on 5s / 1700kv losi motor, about 4,4 kg total. I drive it on tracks as well, so handling with the cheapo-esc is good. except the breaks, which really s..k.

I used to have an etti controller (2 yrs ago), which was no good for car use at all, even for me... I also tried a scorpion commander 120a esc, which supposed to have a car-software. unfortunately mine didn't have that, and I have to say that I wasn't happy with the throttle-curve at all and there was cogging; acceleration was kinda lame, can't describe it. it also went up in smoke after a few rides, (got it replaced and sold it). the cheapo-one is much better...

I have once read that most of the chinese-plane esc are based on jeti-controllers. and somewhere else I have read that in such large esc's older/different fet's are used. these are larger, but like that can handle heat better and are more "rugged".

Anyways, as I am planning to get a car-esc (I think the new GM genius 120R), I just pray that it's reliable like my cheapo-esc. I just got used to worry-free driving.
florian

I have a 8s mgm esc. I was planning om running it on 6s, so no problem there. But I also want to run it on 8s occasionally. The esc has 4 50v caps. Now I'm wondering if it's safe to run the esc to it's full specs after reading the article?

You should be able to just don't brake!

I've always pushed the voltage limits of my ESC's but never had issues probably because I've always used mechanical brakes on my MT's and 1/8th scales. I've used up to 15% motor brakes, to get some front braking power, but thats about it.

I am running 6s on the Quark (the Quark has issues with 6s), and I use only 45% brakes MAX, the EPA is set to 45%, and I usually use only 1/2 of the full 45%.

The way I brake, is usually a small amount when at speed, then I gradually depress the brakes more when it slows down, to keep the same braking action. I have never really used hard brakes ever, it's just too hard on things, drivetrain, motor (heat), and esc obviously....