ScooterMan101 said:
I have a Lyen Controller 9 fet 3077 mosfets , from what I can see on the outside it is exactly like the controllers vendors like eM3ev.com and others , sell that list them at 48 volt 30 amp controllers.
They all look the same on the outside, regardless of what they are on the inside.
How do / Can I exchange a couple of parts inside to make it work on 36-72 volt and 40 amps ?
More voltage requires changing any capacitors and MOSFETs that aren't up to the max voltage it'll see. I recommend a margin over what you need, at least 20-30% (to account for voltage spikes and component aging over time), but some people ahve run stuff on teh ragged edge successfully. Probably 100v FETs and caps will work. Use caps that can take at least 103C; they'll last longer in the heat inside the controller than the lower-temperature-rated units. I personally prefer Panasonic or Rubycon; have not had problems with those brands, when ordered from known-good sellers like Mouser, Digikey, etc. FETs there's a lot of choices nowadays; dunno what's popular (IRF4110s probably fine, but they're also widely counterfeited, so have to be careful where you buy from).
It also probably requires you change the fat resistors in the low voltage power supply section (where the "keyswitch / ignition" wire to a higher resistance, proportional to the change in voltage. Otherwise, the current flow thru them will be higher, and they'll heat up more, and possibly burn out (some poeple have had them burn the PCB itself, and come unsoldered, floating around inside the controller shorting things, while keeping the controller from operating becuase the MCU can't get power).
Higher Voltage is more important to me than higher amps for the motor I will be matching this controller up with.
so up to 40 amps would be good enough .
HIgher amps is often easier to get than higher voltage, becuase it seems quite a few controllers will handle 10-20% higher currents (or evne more for really short periods), and some can simply have their shunts modified, replaced, or added to, to get higher currents. Some are even just programmable to higher current limits.
Shunt mods can be very unpredictable in their results, though, becuase you can't really know exactly how you are changing the shunt resistance. WHne you do this, you make the controller unable to "know" what the current really is, so it can't protect itself from overcurrent anymore.
If you have extra shunts out of an identical controller then you could parallel them for more current. If the original had four, and you added a fifth, you'd increase current by 25%. If it had two and you added one, it'd go up by 50%. If it had one, and you added one, it'd double the current. Theoretically, anyway--actual results do vary, sometimes from controller programming, sometimes from controller powerstage design, or even the motor attached, the load, etc.
I added a third shunt to a generic 15FET, and went from about 33A max, to 80A max--way way way more than predicted! Something else is probably also affecting it, most likley the shunt I added is much lower resistance than the existing ones (but I have no idea what resistance it or they are). I haven't blown it up yet, because it sees taht peak only for a second or two as I startup from a stop, and it's also spreading the load of the trike with a second motor and controller (albeit an unmodified generic 12fet of similar rating as the other started out as).
The main thing about higher current is that it makes more heat, so making sure that heat quickly gets out of the FETs and the controller is very important. The more you "overclock" a controlelr for more current, the more important it is to ensure the FET heatsink bar that they're all bolted to is fully and flatly contacting the case, and that both are very smooth and flat so there are no gaps between them when bolted together, and that whatever heatsink paste you use between them is applied very very thinly to only fill any scratches/etc to remove airgaps.
There is a electronics supply warehouse that sells many parts at good prices just around 30 miles from me so getting higher voltage capacitors is easy and last time I was there I thought I saw some for under $8 each.
The caps should only cost a couple bucks each, to maybe four. Noname / "generic" brand caps would be like a buck each for 100v 1000uF.
I have opened up a couple controllers now and have no idea which of the round cylinders are for voltage or amperage or just capacitors, and what a shunt looks like , and what exactly to change out .
All of the round cans are capacitors, and only affect voltage.
The shunts look like fat wires about half an inch to an inch long, that just "jumper" across a gap over by where the big fat main battery power wires come into the controller board.
FETs are for both current and voltage.
( Amberwolf I read a post by you where you said one still works on your SB and a controller from someone else blew the first time you tested it,
Yes, there have been a few blown ones. One was blown when I got it because the original owner had accidentally reversed power to it; it was fixable with a new shunt and caps, and it still works even now. One that blew when I tried it out was a modded 18FET; I forget who by, maybe Methods? was a long time ago--but it fried trying to get CrazyBike2 going; I think I was trying to get it to wheelie.
It just couldn't handle the overcurrent from whatever mods had been done to it; it had no way to control the current becuase it couldn't know how much current there really was, so POOF.
The one I added the shunt to still works...so far.
But it (and the other generic) are getting replaced by a pair of Grinfineons as soon as I cna build the new wiring harness for them (and depending on results I may mod *those* for more current, too, but only if they won't let me haul the loads I need to with quick enough acceleration--I'm much more interested in perfect reliability on the trike than performance).
Keep in mind anyone reading this that the Controllers that I have that I want to mod have the plug for a Cycle Analyst, and I have the Cycle Analyst that I will be using with the Controller that will be modded , and I know how to set the amp limit in the settings on the Cycle Analyst .
Therefore I am willing to do a shunt mod as well as higher voltage If the 9 , 3077 fets can handle it .
The CA will no longer read current correctly either, so its amp limiting will be wrong too, just like the controller's. The only way the CA could properly read the current (and thus do any limiting correctly) is if you knew exactly what the new shunt value was, and programmed that into the CA, *or* if you buy the external (SA) shunt from Grin and used that instead, just for the CA to know exactly what the current really is.
FWIW, it is probably cheaper to buy new controllers that handle the voltage and current you want, than to mod the ones you have--generic controllers are pretty cheap these days, and all the FETs and caps can cost up to twice as much as the controller, depending on what you need it to do vs what it did before, and what features you need. Especially if the higher current/voltage turns out to be more than what something else in the controller can handle and it takes out other parts than those you modded or replaced already, and you have to repair those too. (and if you use new ones you still have the old ones as working spares....)
A quick google finds a number of 72v 40A controllers for as cheap as $40 with free shipping (DOMINTY on Amazon), but I have no idea of the quality. The generic I shuntmodded that still works was about that price, I think.
At a guess, genuine IRFB4110s will probably cost around $5 each for 10 or less of them, possibly including shipping. So there's $50 already. There's probably 2 to 4 caps that will need replacing, if they're $2 each, there's another $4 to $8, higher if they cost more ($20 to $40 at the prices your local place charges). The resistors in the LVPS are probably a dollar or two.
So guesstimate $60-$100 for parts to upgrade each of your controllers, not including oopses in assembly or testing.
