Ku65 controller max voltage allowance

[patpatbut]

Hi all

I am tempted to increase the top speed by replacing my 36v battery to 48v lifepo4 battery.

My controller is ku65 but I am not sure if it can take 48v lifepo4 battery without blowing it.

I remember the full charged 48v battery has around 52.4v not sure the capacitor in the controller can take that voltage.

Does anyone know if the ku65 can take 48v lifepo4 without problem?

 

[Deleted member 4366]

Those that tried appear to have been successful. Technically, you're over the limit. 12S Lipos 44.4v (actually 48v) is safe.

 

[patpatbut]

Those that tried appear to have been successful. Technically, you're over the limit. 12S Lipos 44.4v (actually 48v) is safe.

Thanks Dave. Originally the plan was getting lipo but I am confident to handle them safely.

That's why I need lifepo4 which saves me a lot of hassle.

Do you know what is the capacitor inside the ku65?

Thanks
Pat

 

[Deleted member 4366]

KU63 has 50v. I've never opened a KU65, but it should be the same. You can get two KU65s from BMSBattery for about £30 including postage, so you could order a couple, which take about 10 days to come, and then try it. I've heard of about three people running these type of controllers on 48v even though over the limit, but you never get to hear the long-term success. Maybe they went pop on a hot day.

 

[Old_Dave]

Yep.. the KU65 is also 50v working.

 

[Jeremy]

Bear in mind that, for reliability, and as the commutation capacitors in the controller see a voltage that is slightly greater than the battery voltage, it is wise to derate the working voltage limit that's marked on them.

The capacitors actually see the battery voltage plus the commutation ripple voltage (generated by the switching currents in the controller and acting across the capacitor internal resistance). Typically this ripple shouldn't exceed a couple of volts, if the controller isn't being over-run and if the capacitors are of reasonable quality.

For reasonably acceptable reliability (i.e. a working life of a thousand hours or so) then the capacitor should be derated by 10 to 20% in terms of battery voltage to capacitor voltage rating.

You can certainly sometimes get away with running right to the limit, or a volt or two above it, but you're very much in unknown territory with regard to reliability.

Worth noting that if the commutation capacitor fails then it's a near dead certainty that several FETs will fail, as those capacitors are all that's keeping the voltage across the FETs at a safe level.

I've run 15S with 63V rated capacitors and got away with it for a few rides, but I decided to replace the capacitors with 100V rated ones, as the cost was relatively small and fitting new capacitors only took 15 mins or so. I think it took me far longer to find suitable low ESR capacitors that would fit into the case than it did to fit them once they'd arrived.

 

[patpatbut]

Bear in mind that, for reliability, and as the commutation capacitors in the controller see a voltage that is slightly greater than the battery voltage, it is wise to derate the working voltage limit that's marked on them.

The capacitors actually see the battery voltage plus the commutation ripple voltage (generated by the switching currents in the controller and acting across the capacitor internal resistance). Typically this ripple shouldn't exceed a couple of volts, if the controller isn't being over-run and if the capacitors are of reasonable quality.

For reasonably acceptable reliability (i.e. a working life of a thousand hours or so) then the capacitor should be derated by 10 to 20% in terms of battery voltage to capacitor voltage rating.

You can certainly sometimes get away with running right to the limit, or a volt or two above it, but you're very much in unknown territory with regard to reliability.

Worth noting that if the commutation capacitor fails then it's a near dead certainty that several FETs will fail, as those capacitors are all that's keeping the voltage across the FETs at a safe level.

I've run 15S with 63V rated capacitors and got away with it for a few rides, but I decided to replace the capacitors with 100V rated ones, as the cost was relatively small and fitting new capacitors only took 15 mins or so. I think it took me far longer to find suitable low ESR capacitors that would fit into the case than it did to fit them once they'd arrived.

Thanks Jeremy. If the KU65 is 50v capacitor, how easy to replace it with 63v one?

Can you show me how to do it? I can do soldering.

 

[Old_Dave]

From memory it's 4 capacitors that could do with being replaced, all of which have easy access


Sent from my iPhone using Tapatalk

 

[Jeremy]

I've not replaced the capacitors on the KU65, but it should be straightforward. The main thing is to choose the right size and type of capacitor. The type is important, as these commutation capacitors have a tough job to do and need to have a low equivalent series resistance (ESR). They also need to be small enough to fit inside the case.

There is a big variation in ESR between capacitors, even those from the same manufacturer. For example, the Rubycon ZL series are pretty good, but the Rubycon YXH series are poor. These capacitors also need to have a high ripple current rating. Even at a controller current as low as 10 to 15 A the capacitors will see a ripple current of maybe an amp peak.

The best bet is to measure the dimensions of the existing capacitors, see how much bigger you might be able to go in terms of both diameter and height (and still fit in the case) and then go searching through the on-line catalogues at places like Farnell and RS looking for suitable low ESR, high ripple current capacitors that will fit.

Worth also noting that the low speed, high torque load FET failures discussed recently in another thread was almost certainly exacerbated by the rather poor capacitors that some controllers have as standard. A high ESR capacitor will allow high peak ripple voltage at the FETs, increasing their stress level when they are already working hard to deliver the high peak motor current demand.

 

[NRG]

Ive replaced the caps on two on my xiecheng controllers, it take a bit of care. First the PCBs don't take much heat, you can easily lift a track or solder pad if heavy handed. Secondly they use through-hole plating making component removal a bit of a challenge. Use a good square tip soldering iron, 25w or so and heat each leg/pad of the cap in turn and tease or rock the capacitor side t side until the legs are free...you may need to add more solder to each pad and then only clear the holes with a solder sucker after the cap has been removed.

For caps I recommend Panasonic FK types or for a slightly lower ESR the FC types, very good caps for this application.

 

[NRG]

Here's a photo of one 6FET controller during FET replacement, I replaced all the caps but they key ones are the three upright ones in the front of the picture, the one laying down and the large upright one on the r/h side. Oh, I forgot, these caps need to go in the right way 'round they have polarity so get the + and - correct otherwise they will pop!