Mosfet SWAP?

[Tek]

I've got a scooter that needs a controller upgrade. It's got a box with 60V and 30A ratings. I'm going to upgrade all power wires on the bike but I'd like to mod the controller.
Currently it is equipped with 12 RU75N08 Nchannel FETs.
Anyone have experience like this and what fet's can I upgrade easily with?
Could I use IRFB4110 or the like and get improved power capabilities just like that?
Any and all input is helpful.
Thanks!

 

[Punx0r]

Your current FETs don't look great: http://www.dzsc.com/uploadfile/company/49818/20119229434196.pdf

75V 80A and 8 milliohms RDSon

The spec on the IRF4110 is much better, but the IRF3077 is even better if you want to stay with a 75V FET (60V controller rating).

Swapping one type of FET for another is a bit of an unknown as there are so many different characteristics for these devices so they don't all function quite alike. Such swapouts usually work fine though

 

[dnmun]

do not try to swap out the mosfets.

the Ruichip mosfets are good parts so there is no need to replace them. obviously they are limited to 75V but unless you exceed the 75V they should do fine.

if you go to higher voltage it would be easier to just buy a high voltage controller.

 

[Punx0r]

That really depends how far the OP wants to push the controller. 60A would probably be fine on the Ruichips FETs. However the 3077s are 120A rated Vs. 80A and 2.8mohm Vs. 8mohm

 

[dnmun]

i would never recommend anyone, especially a newbie with little experience, remove mosfets from the controller and replace them.

it is just better to buy a controller built with better mosfets already installed.

i don't think he would gain any voltage going to irfb3077 and he can buy a controller with the 3077 already installed anyway. cheaper than buying them from digikey too.

 

[friendly1uk]

Swapping the fets won't turn the power up. If will just raise the limits when you do turn the power up, which is a different procedure.

The controller regulates the power, and to do so it measure what is going on by looking at volts drop across the shunt. You need to lower the voltage across the shunt by thickening it up with solder, Or putting resistors across it.

Actually... I'm guessing as I dunno what controller you have. There could be a software method, though I doubt it.

 

[Tek]

Thanks guys! while not a newb with the soldering iron, I haven't much EV experience. I could change the fets but first I am changing the battery linking cables and all the battery cables into the controller with 2x the diameter wires. We'll see if that makes a difference.Should have about 4x less resistance, so I should get some increased hillclimbing and accerelation. Thanks for the guidelines on the controller end.

 

[friendly1uk]

Be a lot of work for very little gain pal. If the thicker wires allow more power to flow, the controller will stop it from happening. It regulates power flow, not under sized cabling.

 

[dnmun]

not only is it very hard work, but almost impossible to resolder the new FETs in place after you ruin the original traces while unsoldering the original set of mosfets even when you can reach the legs of the mosfets to cut them off first. just reaching the legs with dikes is almost impossible.

then when you assemble it with new mosfets, the solder shorts out the legs somewhere and you have a just a big poof when it gets powered up again.

just not worth it.

 

[Thud]

Been there/done that....learned a lot.

But If your hot to assemble your own xiechang controller, you can buy "kits" from ecrazyman on ebay.
Kewyn will have every thing populated except: the power mosfets....he will ask you if you want 63 or 100 volt capacitors.

(On the 6fet units, you will need to calculate the droping resistor to power the 12volt regulator...the 12 fet & up units have a dc-dc converter that will run it on most common bicycle voltages)

They are easy to assemble & he will even sell you a harness for plug in cycle analyst operation.

The boards have evolved but this is a good referance(even with corupted photo's)
http://endless-sphere.com/forums/viewtopic.php?f=16&t=7361

Re building one is a fun learning experiance.....but you can buy them set up, tested & working from em3v with exactly what you want including connecters for prolly less money than sourcing the parts.

 

[Skalabala]

Thud buddy, you don't mind posting a link for the kits from ecrazyman?
I could not find any

 

[Thud]

It is an unadvertised service,
Just send a polite e-mail to ecrazyman (@) Gmail dot com with a reqest of what you would like.
It may be a day or so for a reply, but he'll get you what you want & for very reasonable pricing.

 

[liveforphysics]

The Ti nextfet series is really what you would want if you're going to go through the effort of swapping.

They are the current cats pajamas TO220 that is both cheap and available.

 

[Tek]

Thanks for your advice guys, I will try and upgrade by myself...as I like this kind of thing. I also wouldn't be buying stuff from anyone as I live in Shanghai and if it all comes down I'll be making the 45min drive to Savboton factory and buy one of theirs.

 

[amberwolf]

FWIW, on CrazyBike2, I removed most of the connectors between batteries and controllers and motors, and went the direct-soldered route. Some places I also removed lengths of lesser-gauge wire, and shortened some of the larger gauge stuff where possble.

I do feel a teeny little bit of improvement in acceleration on it, though not enough to quantify numerically. It does affect the power consumption, because there is less voltage sag at the controller now, so I see nearly 10% higher Wh/mile readings on my work commute (whcih has a LOT of stops and starts) than before.



Changing out just the FETs wouldn't likely give me any more power, though, and could actually make things worse: If the gate drive circuits in the controller aren't good enough to charge a different FET for some reason, the FETs could turn on slower than they need to, and power could be wasted in the controller as heat, causing things to perform worse than before (or worst case, actually fail). It's probably not a likely scenario, but it is possible.

Best case, different FETs would have lower RDSon, which would let more of the power get to the motor, and have less heat in the controller--but this is even more marginal an improvement than taking connectors out of yoru wiring, most likely.