Mamba Max FET Datasheet
 

I took my MM apart and was able to determine the part number of the FETs for those interested. If you flip it over, you can just make out the part number on a couple of the FETs: 22N03S.

And here is the datasheet...

I haven't had time to really pore through it, but I'm sure other electrically-minded people will like this (GriffinRU, Zeropointbug, etc).

my brain just a 'sploded

LOL. I just noticed that the Quark and MM use the same FETs! So any problems with heat, blowing up, etc is purely programming! VERY interesting!

Why's that?

okay, a good 70W on 60 C. How many are there in there?

Um, I forgot to count them. I got too excited when I was able to get the part number. Oops :dft001:

I think there are the same number on the bottom as on the top. IIRC 6 in a row X 3 rows (for each phase) X top and bottom. Someone with a MM in front of them can confirm the top count (at work right now).

yah! As soon as I saw the 22N03S I was like whow... :eek:

It must be a good FET then, if both are using it, two of the best controllers out (not that there are many for cars).

So you mean both the Quark and the MM have to same amount of the same FET?

This is kinda cool, in a way.

So technically they are the same except programming, and packaging. If I'm right about that, then 5s is doing no more damage to the MM than the Quark

Or the caps... caps should be twice the voltage rating of the max 'specs', IMO.

if i'm correct the quark has 25V caps but i think it has 4, with the mamba having three 25Vcaps, they should be about close

Actually, the Quark only has two. :002:

If i ever used my Quark without the case, I would solder new caps that are 35v, and larger uF cap. You can't use the case though, they would be too big.

What IS the issue with the Quark with 6s? Was it the caps? Or the FET's?

Wait, I just checked, the Quark has two 35v caps, 330uF a piece. They might have a temporary rating of 40 volts.

Ok stand corrected, thought i read somewhere they were 25V

It makes no sense in comparing the quark and the MM because they use the same fetts.

The entire setup may be completely, i mean COMPLETELY different. they may work with another switching frequency, another startup sequence, other software, other steering technique etcetcetcetc.. a powerfet is only a small detail of an entire controller.

Very true Serum. It's like saying two cars are the same if they use the same piston.

We already know they behave differently, so the programming is obviously different. The Quark is much smoother, but heats up more than the MM. The Quark also seems like it has beefier traces and stouter construction. Don't get me wrong, the MM is a nice controller (especially for the money), but I still like the Quark better for larger vehicles. And let's not forget that the Quark can run pretty much any motor. The MM tries to cram as much circuitry as possible is a smaller area so thermal concentration will be higher.

Personally, the Quark would be THE controller if it had a better heatsink (or maybe just get rid of those stupid thermal pads), computer programming function, and lower price.

And if it ran 6S like it's supposed to.

Not everyone is having trouble with it on 6s, but yes, still it should be able to do 6s LiPo without issue no matter what.

yes, I guess that too...

ZPB: Who isn't having trouble with the Quark on 6s? AFAIK, everyone has reported issues. The only time it seems to work is if you let the pack discharge a little so it's not at exactly 4.2v/cell.

I wonder what the result would be if you set the charging CC/CV threshold to something like 4.1v/cell or 4.15v/cell? Anyone want to donate a 6s pack for me to try? :)

Really, I thought not everyone was having issues?

I wonder how 7s A123 pack would do? Same 25.2 volts at fully charged. However, you CAN use a 3.45 volts to charge, I think you get about 95% charge, and takes a little longer (3.45 float charge voltage), so that's 24.15 right off charger. But as soon as you discharge a little bit out of an A123, the voltage goes down to operating nom. volts instantly. I bet a 7s A123 might work fine.

It's basically a problem of discharge curve, and voltage droop of batts. If batteries had a very good discharge curve (like A123, or better) and very little droop, you would need less cells to get the working voltage. It's better for everything really...

Altair Nano cells are going to be 2.4 volts/cell, which is nice, gives you more voltage options. They also would last a lifetime, LITERALLY, they have put 10's thousands of discharge cycles on them at 10C discharge/10C charge, they retain over 85% capacity still at that point (16,000 was the latest figure)

Nice info, Brian!

At what freq. ESC running do you know? Also, is it difficult to measure or identify series resistor (in FET's gate) value? I think I have a clue why both of them have problems at higher voltages.

Meanwhile looks good on paper, but input capacitance a little bit high.

Why would input capacitance on an esc be too high? The more the better...

Do you mean esc caps?

He means the capacitance of the FETs.

Yups..

Oh, does that lengthen the slew state? Seems to me it would?

No, I don't know the resistor value nor the frequency. As it is, I could barely make out the number because just the top of the FET was peeking out. Judging by the compact cluster of FETs on the power board and the number of connections between the two boards, the gate resistors are probably on the "brains" board - and there's no way I was gonna risk a brand new ESC by unsoldering the two PCBs apart - especially since they are multi-layer PCBs. :) It would be nice though if someone did and took some very hi-res pics of all sides of all PCBs. Maybe someone who has a bunch of these things and one or two are questionable anyway *cough*glassdoctor*cough* ?

I just thought it was quite interesting to note that the Quark and the MM use the same FETs. So, any "issue" specific to one ESC is solely due to programming, and possibly a choice in support components.

Since it appears you will tear it apart and look anyway maybe I should tell you what our fet is? I think the spec sheet will impress you. Far better than what you are looking at currently for large scale applications.

Who would have thunk the monster guys were this techy...We have pro drivers that do not really even know what a fet is, or care as long as when they squeaze it it rips.

Tekin Prez

LOL, yeah, we'll rip things apart - as long as it doesn't damage the ESC of course. Hate wasting money that way. :)

I'd be happy to know the FET component number. It would save the need to tear it apart. :)

Some of us are techy, some just want to plug it in and go. There's quite a mix here.

I have heard that the newer MambaMaxes have a different FET... but I have an older one I could get some pics of... but then maybe CC would think I'm nuts and won't fix it them for free anymore. :D

Steve at radiocontrolzone.com has had the heatsink off the MM and even replaced all the FETs he damaged in the process.... maybe he has more info. He also moded the BEC circuit with 6v Novak components.

Prez.... you know, some of us would make great beta testers. And some of us can hang with your team drivers on the track too. ;)

GD: Yeah, I'd be curious to know the older FET number if you can manage it...

The MM I have is very new, so it probably has the newer FET.

Good point there buddy! So I guess it will go in order of status, right? As in Mods first. :)

Need to protect ourselves a little. The others will find this fet also I am sure. I will tell you the continuous drain current is higher, Rds on is lower, gate charge is lower and pulse current is unbelievable. Try over 250A vs 106A. Spendy little buggers however.

Tekin Prez

And here I was thinking an EE like myself would make a nice beta tester. :D

From TheSteve:
.

He said that these are better than the original MM fets.

he thinks this pic shows the same fets... ???
http://www.fahrenheit-145.net/Fahren...W/cimg0001.jpg

Sooo, what is that thing? Is that the MM taken apart? Old one?

No... not a Mamba... I'm not sure what it is. Looks like a Schulze... TheSteve posted it on the other forum.

Its not a problem to find a better FET, its how to make money while using it :)
Welcome to the Forum!

personally i'd happily pay the extra for reliable contollers,still the weak link with our hobby,with the bigger vehicles atleast,1/10 is no problem.

Those capacitors in that pic are too far away from the board, they need to be RIGHT up there.

OK, I took apart the MM from my Jato and found that it uses a different part number for the FETs: 4108N

You can find the datasheet here.

And, while I had it apart, I popped the heatsink off (very easy to do really if you're gentle) and snapped a few pics if anyone is interested:

http://scriptasylum.com/forumpics/mm_1.jpg

http://scriptasylum.com/forumpics/mm_2.jpg

http://scriptasylum.com/forumpics/mm_3.jpg

http://scriptasylum.com/forumpics/mm_4.jpg

http://scriptasylum.com/forumpics/mm_5.jpg

http://scriptasylum.com/forumpics/mm_6.jpg

http://scriptasylum.com/forumpics/mm_7.jpg

http://scriptasylum.com/forumpics/mm_8.jpg

COOL,
Nice job Brian