[rmay635703]

If you want to use a Kelly you must oversize it at least 2x and run a voltage about 20% lower than its rated at.

If you do both of those things they usually work fine but don't expect a warranty.

 

[teverth]

Thanks rmay. This is my car: http://www.evalbum.com/1772
Has been running since 2008. After four controllers hit the dust I am obviously cautious and want something that will last...
You say Kelly has no warranty on their controllers? That sounds discouraging.

 

[onegreenev]

Setting Regen too high. At the speeds you get with the automobile and the currents you can really crank up the current to beyond what they can do. These controllers are designed for golf carts and industrial carts. Not for what cars can dish out. Same for many other controllers available. Regen controllers are not easy to come by. Best to stick with series motors and forgo the SepEx stuff. Not that they are bad but because no one builds an automotive quality SepEx controller.

Pete

 

[teverth]

Yes indeed. The Navitas had a 'great' logic: When it detected an over voltage condition it pulled the contactor. Needless to say that the only case of over voltage is regen braking. So instead of regulating braking down - as it could and should - it did the one thing one should NEVER do, pull the contactor under load. Bang, the voltage shoots through the roof...
The Sevcon handled that better but it simply was not lasting and burned out over time though perhaps bad caps or cascading failure of its MOSFETS.

Generally I think you are right: I should have started with a series motor... I keep wondering if I simply could drive the shunt field in parallel with the armature on a series type controller? Has anybody done that?

 

[major]

I keep wondering if I simply could drive the shunt field in parallel with the armature on a series type controller? Has anybody done that?

Yep, a couple of guys have. One not too long ago. Search for SepEx. If it comes to me I'll post it here. And as far as Kelly SepEx; I don't think it would be a wise move on your part.

I've had good luck with SepEx on both Curtis and Sevcon (Millipak) with 48V on IUVs (Industrial Utility Vehicles). Haven't done a high speed EV with one.

Ahh, it was mechman600. http://www.diyelectriccar.com/forums/showthread.php?t=81416&highlight=SepEx

 

[teverth]

Very interesting! It turns out on closer inspection that the HSE Kelly controllers don't do field tracing. In other words they provide a constant field current only! Dugh! Especially from what mechman600 said, that would be a show stopper.

Perhaps I ditch the SepEx motor completely and go for an AC35 or AC50 system.... cost.... arghhh

 

[teverth]

Just heard from Fany at Kelly that they plan to have a field tracing version of their high power sepex controllers in a couple of months... perhaps worth waiting for.

 

[puddleglum]

Just heard from Fany at Kelly that they plan to have a field tracing version of their high power sepex controllers in a couple of months... perhaps worth waiting for.

I've not used anything from Kelly, but according to Fany, their KDZ series controllers offer field tracing. She also noted that it is a tracing function not field weakening. They offer a 72v 550a model or a 120v 400a one. Others have said Kelly doesn't put out what they claim to, but in theory it should work. I was looking at them when I was thinking of going sepex. They claim a 1 year warranty, but getting them to honor it might be hard, I couldn't say.

 

[teverth]

Thanks, yes I have seen these but all say I need to double my current requirements with kelly. So I need to look at the HSE models. My max current will be about 400 amp but kelly strongly suggests that I use their 1000 amp model......

 

[onegreenev]

If you go with a Kelly I strongly suggest you go with their 1000 amp model as well.

 

[DavidDymaxion]

I have a Kelly HSE (sepex) controller. I couldn't get it to work on 48V but it worked great on 96V. I haven't used it much, but it has worked perfectly on 96V so far including full power accelerations (haven't verified max current, but the acceleration was good). I tried to set it up for regen but haven't verified it actually regens yet. It cooked a motor on the Salt Flats, so the current output couldn't have been completely wimpy. I still need to get my new and improved motor in and my lithiums (hopefully before September).

I've heard this about the field not changing, but on mine with a meter I could see the field weaken if you suddenly commanded full power. My manual said it is constant field until 90% PWM, and then above 90% the field starts to weaken (I think to about 1/2, I forget the exact number now). I imagine if you have certain voltage or current limits you might never see the field weaken for those reasons. My controller is about 2 years old.

Mine is on a huge heat sink, several posts here on DIYelectriccar say you need to do that will Kelly controllers.

Very interesting! It turns out on closer inspection that the HSE Kelly controllers don't do field tracing. In other words they provide a constant field current only! Dugh! Especially from what mechman600 said, that would be a show stopper.

Perhaps I ditch the SepEx motor completely and go for an AC35 or AC50 system.... cost.... arghhh

 

[joeblack5]

So , I would like to use a kelly control for a PM motor KDH14801E and provide a field control myself similar to Mechman designs and experience.

I am not getting a clear answer from Kelly ( Fany) about the regen ability of the control.
My questions to Kelly is how much regen current can the controller deliver when the vehicle roll down a moderate hill with the Pm motor generating 50 V and the battery pack ( not fully charged) is at 120 Volt.

It seems there must be limits to generating from very low motor voltages into higher battery voltages. I have asked Kelly about some literature but sofar that seems non existent.

Anybody has some real life experience? with Dc regen controllers from Kelly in PM or shunt motors?

Thank you all,

Later J

 

[major]

It seems there must be limits to generating from very low motor voltages into higher battery voltages.

The limit would be that of the controller's current limit. When the DC motor controller regenerates it functions as a boost converter taking the motor's generated voltage and converting it to a higher voltage to force current into the battery. So, to the user, it is irrelevant what the motor's voltage is. The controller will boost it as needed. Obviously as the motor approaches zero RPM, the generated voltage becomes too small to boost, but typically this occurs at like a few percent of rated RPM.

 

[joeblack5]

Mayor, Thank you,

I was hoping Kelly could quantify this a little better in graphs so that I get a better understanding about the possibilities.

Later J

 

[joeblack5]

This is the response I received from Kelly. We have gone forward and backward a little about this and it seems Kelly has no other way of explaining the regen ability.
Anybody else has had similar conversations with Kelly and got a better understanding out of it?

Thanks J




Hi Johan,

By default,the brake switch regen is 10% of the driving.
The brake analog regen mode is about 10%-50% of the driving.


Usually the regen will quit at high speed on downhill.


Thanks,
Fany

 

[joeblack5]

Bump,

According kelly these controller could not do regen on along hill?
I do not understand their answers as an example in previous post.

Anybody with experience with Kelly regen that can help me understand what they mean?

Just bought a wonderful evcort with broken soleq and have read about these cars on this forum.

May be it is just a little language barrier with Kelly?

Thank you,

later J

 

[onegreenev]

PM Sent for your review.

 

[DavidDymaxion]

As I recall, you can program for up to half of motoring current for regen.

I imagine several things could make regen quit on a long steep hill at high rpm.

• remember voltage rises on regen, so the high voltage cutoff could be triggered
• over current protection
• thermal cutback

Sorry, haven't done regen on mine, hopefully soon, but a local guy gets regen on his.

 

[joeblack5]

Great, so it sound like the Kelly regen is legit.

I will start looking for 144 volt 1000 amp kelly regen.


Thank you for your help.

later J