Is it safe to run 24S lipo / 100V battery on 100V fets/caps?

[miuan]

Although much has been mentioned in other threads, there had always been some degree of disagreement on this topic.

There are a lot of controllers running 4310/4110 or other 100V fets and 100V caps, but they are mostly rated for max. 72-74V nominal battery voltage. Still, many use them with 24S lipo without problems.

For me, I own a 2807 and run a 24S lipo on Lyen 9x4310 at 30A (stock shunt). I like the setup very much but I notice the controller gets quite warm at times. I don't think the controller will blow from the heat, but it may be the voltage spikes near (or slightly over) rated voltage, combined with heat, that may leed to a breakdown, possibly shorting a phase and sending me down at 60kph.

To avoid this, I could just purchase a HS3540 and run it at 74V with sensors at 35A, getting similar performance to my 2807 setup, but it'd cost me a lot of money, plus time to add sensors and rearrange my battery pack. Or I could order a new controller with 4115 fets.

All experience is welcome, I believe it will be very helpful for other ES members too.

 

[icecube57]

You can but its not suggested. You are running a 9C motor which is a low inductance motor. Its not going to generate the high current and voltage spikes that some of the other motors can. Voltage is one reason why controllers blow. They have caps in the controller to absorb the voltage spikes. If a cap goes bad you have no protection. Another thing that kills controllers is amperage. A cold controller allows for bigger amp spikes thats when the resistance of the components is at its lowest. So play gently with you controller at first. Knock the surface charge and get an amp hour or two down befor you start laying into it. For me high inductance motors like the 5302 5303 and 10x6 and 12x5 9C draw alot of current. When you go up in voltage you should back down in amps especially on the 12 fet controllers. The closer you are to the max rated voltage you need to derate the mosfets max amps. For these 100v controllers alot of people play safe and stick to 20S. Its alot safer in the long run. I went 18s.

 

[neptronix]

As a rule i would not run anything at the absolute edge of it's specifications if you want it to last.

Be safe and get a 12FET that's rated a bit above what you're looking for.

 

[texaspyro]

No

 

[icecube57]

The 4115 mosfet controllers can take a higher voltage but cant dish out the amps like the 4110 can.

 

[methods]

icecube makes some good points

I have been running 24S from the beginning on 4110 fets. I have blown up a controller or two - but I have blown up a controller or 5 running only 12S on 4110's. It is simply a matter of understanding and managing your system as a whole. What you are running for capacitors is probably more important than what you are running for fets - it is the caps that protect your fets

- Match your voltage to your KV - running 100V on 4110's with a 5302 in a 26" wheel is going to end bad. With a 5305 - things are fine

- Don't stall the bike - hearing your motor go DUK-DUK-DUK-DUK while you slowly climb up a hill = craps shoot. Even my wife has blown 4110 controllers on 12S due to lugging. Running the motor in the low efficiency zone not only overheats you motor but it really stresses the parts in your controller.

- PWM makes heat, not necessarily high current - turning down your current limit != turning down the heat! Again - it is about a matching Voltage to KV and no matter what you do running a high kv motor on high voltage is always going to kick your controllers ass. Rule of thumb - if you cant get within 20% of the freewheel speed on the flat then you dont have a good match.

All my bikes run 24S on 4110's
Is it "safe"? I dont know.
Does it blow up? No.... Not while I am at the controls - and I have the current limits maxed out. The thing a lot of new guys dont realize is that guys like me dont hammer it all the way down the street with it pegged on the current limit.... to me that would indicate a poor match. My bikes are on the current limit for a couple of seconds and then they have a linear taper down to something in the ballpark of 1.5W cruise.

The 4110 is an awesome fet. It has gobs of margin as it is produced by IR. You can step up to the IRFB4115 (150V) but you pay a big toll in mosfet performance.

Whats the worst that can happen anyway?
It goes POOF, you fix it, you learn something. 99% of the time you can fix a controller in 5 minutes by just swapping a few fets. If you have good technique it is a trivial procedure. Learning something while fixing a controller is ten times more valuable than owning something that is "reliable". I would rather have a true understanding of something than a true trust in it that it "Just works".

But then that is me... and I have pushed my 80lb bike more than a few miles. I consider that time to think

-methods

 

[miuan]

Thanks methods, I tend to agree for the most part. My controller is a Lyen, so should have good caps. I ride it quite relentless, applying power whenever the traffic requires it, but mostly I just cruise at 50-60kph, and the climbs are short on my commute. My motor has a no load speed of about 90-100 kph, and reaches 70 kph max. on straights, so it's not exactly 80%, but at these speeds, reaching 80% is not very easy. With a 30A limit, I ride with PWM most of the time, hence the warming. Increasing the current to 40-45A would probably help get rid of most PWM time at high speeds, but then I think I'd just kill those nine little 4310's. What happens if the controller blows at speed? Will it short the phases and brake the wheel?