Controller Selection

16 posts / 0 new
Last post
Bruce_Wayne
Offline
Last seen: 14 years 8 months ago
Joined: Tuesday, September 18, 2007 - 17:45
Points: 67
Controller Selection

Just wondering if anyone in the community has any experience with curtis controllers? I know they've been around for a great many of years and hear them highly recommended for car conversions.

Curious to know if anyone has played with regen on the emax as well. Something like this looks like a good fit. Can anyone tell me what parameters you can program via rs-232 on this? Didn't mention that on page.

http://www.kellycontroller.com/shop/?mod=product&cat_id=41,22&product_id=304

Couldn't find the schematics for the emax either in the handbook section. I get redirected to an ilbcnu.org search page.
Anyone still alive out there?

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: Controller Selection

I have E-max wiring diagrams that I could send you via the e-mail on your web site if you like. Please respond if you want me to send them. I'm not sure if the yellow-green-blue/A-B-C motor phase and hall signal wire designations are "standardized" or not. If not, a bit of trial and error with the wiring connections is going to be involved.

Adapting a Kelly controller to an e-max is too big a job to give a comprehensive answer here. Just a few thoughts. I believe that Matt (nick=antiscab) has put a Kelly on his e-max, so he can probably help. The e-max can be rigged for regen. The controller can be set up to provide regen a number of ways. The simplest way being to use one of the brake lever switches - the other still used to operate the tail light and 12-volt power-shutoff signal to the controller. The regen braking will be an all-or-nothing setup, so the programmable %regen setting will have to be set by trial and error.

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: Controller Selection

According to this old thread:

http://visforvoltage.org/forum/3497-replacement-controller-kasea-ze2000

The correct phase and hall signal wiring is A=Blue B=Yellow C=Green. Good luck.

Bruce_Wayne
Offline
Last seen: 14 years 8 months ago
Joined: Tuesday, September 18, 2007 - 17:45
Points: 67
Re: Controller Selection

Pjd, that would be really great of you
to email those diagrams. Bruce [at] pcelementz.com

antiscab
Offline
Last seen: 10 months 1 day ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Controller Selection

Gday Bruce,

I have indeed used a kelly 72v 400A BLDC controller on my emax

PJD is correct, putting one on is a bit tedious.
kelly supply a pretty decent wiring diagram though.

parameters that can be changed in the kelly controller

Battery side low voltage limit (soft, doesnt cut out when voltage falls below this, just lowers current until voltage is above)
Battery side high voltage limit (regen backs off when this is exceeded)
Battery side current limit
Motor side current limit
Motor side voltage limit
motor temperature cut-out (the thermsistor used in the emax motor isnt compatible)
output frequency limit (speed limit)
max regen from brake switch (i wired mine to trigger with the rear light)
max regen from regen throttle (i tried putting a second throttle on the left handle bar, but never got it working)
max regen from 0-throttle

truly an impressive piece of equipment, IMO

to be honest, im not much a fan of *any* of curtis controllers.
They are *all* of analogue logic side control design, and offer very few features.
the interface between the accelerator and powerstage is very poorly thought out, IMO.

Is this talk of a new controller because you blew your original one up?
how is your kasea/emax going bruce?
I dont hear of fellow emax owners often anymore.

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

Bruce_Wayne
Offline
Last seen: 14 years 8 months ago
Joined: Tuesday, September 18, 2007 - 17:45
Points: 67
Re: Controller Selection

Goodday matt,

Excellent to hear from you again. I hadn't been riding it all that much this year mainly because of all the damn rain we've been getting on this side of the planet. I put about 1000 miles on it this year and the batteries still work like a champ. One weeknight while coming back from the local theater, I was coming home, wen't over some railroad tracks, and the the throttle cutout on me. I had to push it home 4 miles, hehe. The 12 volt parts are all working fine.

I took the controller apart and saw a rather thick pathway on the far right of the pcb burned open. I used some metal tape and repaired the pathway and I had my throttle back again for about 5 miles on and off. Now the throttle is dead again. I don't feel it's worthy of being repaired as long as there is something much more thoughtfully designed out there. The wiring job on this controller is horrible, I will definitely have to re-do most of that anyway. They didn't use enough bond for the 18 gauge wires going to the controller. Anyone can see this. Good god, $3 dollars extra for a molex male female connector? It amazes me this bike worked at all with this factory crap on.
The speedometer is flaky too.

My pack voltage generally stays around 53-56 volts dc. I don't see why I need a controller rated for 400 amps. There's no way I'm pulling that much. I've never checked but I highly doubt it. The 48 volt 100 amp controller w/regen should suffice fine for what I'm doing. I'm curious where you managed to put the other 4-5 cells on your bike to get the 72 volts. I have absolutely no more space unless I mutilate the frame and do some welding. I'll be right back to post some pics of my bike. I would like to see some pics of your setup and get some more input from you very much. Remember I've got the board gary and rich fechter worked on and the 16 lifepo4 cells in series and a stock 10 amp emax charger(that still works like a champ a hundred charges later.)

I don't want to spend anymore than maybe 250 bucks fixing this controller issue. Hopefully this will remedy the soft clicking in low speeds as well that has always been there since I've had the bike.

Oh, I have three spare cells sitting around doing nothing right now, so if there is a way to fit more cells on my bike I would love the speed increase, as this would help going uphills right now. I just don't know of a cost effective way of doing that right now, unless you do ;-).

Bruce out.

Bruce_Wayne
Offline
Last seen: 14 years 8 months ago
Joined: Tuesday, September 18, 2007 - 17:45
Points: 67
Re: Controller Selection

Here's a shot of the large trace that got repaired.

repair.jpg

A shot of the crappy wiring job inside the controller. They didn't leave enough slack on the last wire so I had to splice it and make it longer.

IMG_1458.jpg

Haven't wanted to spend the extra money having the bad design removed and painted over yet, but I'll get there.

IMG_1454.jpg

About 2000 miles or so on the new pack and bms with no problems there.

antiscab
Offline
Last seen: 10 months 1 day ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Controller Selection

i cheated for the last 4 cells and used smaller format YESA cells.
i also took to some of the plastic body panels with a hack saw.
the original emax frame just didnt have much in the way of battery space, im afraid theres no easy way to fit more than 16 cells.

i agree, 400A is overkill.
i bought that large so i could re-use it in a larger motorbike conversion (i am now planning on using a 120v 400A controller instead).

to get exactly the same performance as you get now with using the boost function, id suggest getting a 200A unit.

all motor controllers (except some OEM) are rated in motorside current.
the original emax controller, with boost mode at full throttle applies around 200A to the motor, up till 10mph or 15kmh.
current then falls off due to battery side current limit (which is around 73A).

another reason for the 200A model, is that its the smallest one that has a sufficient continuous current rating.

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: Controller Selection

Matt,

I assumed the current rating for the Kellys was the battery-side current. But you are right about the low continuous power rating. My e-max controllers, which I modified to raise the current battery-side current limit to 95 amps, continue to work well - about 13,000 km on one and 5700 on the other.

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: Controller Selection

Bruce,

If to you change your mind about repairing the existing controller, you could try soldering a heavy jumper wire (10AWG) from the Bat+ wire, over the bad solder and burned board, to the FET bus strip at the corner of the board. This assumes that the solder burned out because of a bad solder joint and not due to a short somewhere else.

One lesson I learned working on those controllers is that you need to find something more powerful than a 40W soldering iron to do any of soldering on those heavy traces or bus strips the on the power-side of the board - including replacing any FET's.

If you aren't interested in repairing your controller, I may be interested in it if the price is right...

Paul D.

antiscab
Offline
Last seen: 10 months 1 day ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Controller Selection

I assumed the current rating for the Kellys was the battery-side current.

All kelly controllers (along with every other controller) are rated in motor-side current for two reasons:
the motorside current will peak first, the limit of which is determined by the FETs on the powerstage
most of the losses within the controller are proportional to the motorside current.

When the motor back-emf = pack voltage, then motor-side current and battery side current are the same, so a 400A controller can draw 400A from the battery.

for the original Emax, the requirement was for a controller that mimics a CVT transmission. This meant the controller had to be able to go into current limit on the battery side at a very low speed when at full throttle.
in original form, the Emax controller is a 200A controller, with a 73A battery side current limit.

When we do the shunt mod, motorside peak current doesnt change (as this limit is set in hardware), but battery side current limit does increase.
This means that the speed at which the controller goes into battery side current limit is increased.

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: Controller Selection

Thanks for the explanation.

So, if I understand this right - using the e-max for an example, a stationary or very slow turning motor will be drawing 200 amps. This will decrease to the battery current limit at some speed. But if the back-emf equals the pack voltage, how does any current flow at all? I had assumes such a thing is approached at the top motor speed.

Also, what do you think the battery amp limit for can be safely increased to? I had initially pretty much removed the battery amperage limit by using a copper shunt resistor jumper during my first phase of fiddling around. The torque and acceleration was very impressive and I recall measuring about 170-180 amps at the batteries when powering up a hill at about 45 kph. I was afraid I'd burn the motor out and settled for about 20% higher than stock - or 95 amps.

How is the motor current defined (and measured)? Is this the forward-or-backward amps in any one phase, or the absolute value of the current in all phases at any point in the motor rotation, or is it an RMS value like AC current?

Thanks in advance for any answers...

antiscab
Offline
Last seen: 10 months 1 day ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Controller Selection

So, if I understand this right - using the e-max for an example, a stationary or very slow turning motor will be drawing 200 amps. This will decrease to the battery current limit at some speed.

yes, though i actually simplified this, explanation below.

But if the back-emf equals the pack voltage, how does any current flow at all? I had assumes such a thing is approached at the top motor speed.

perhaps i explained this a bit poorly.
You are correct, the back-emf component is less than pack voltage when the motor is driven.
top no load speed is when back-emf*sqrt(2) = pack voltage.

motor voltage = speed*constant + amps * internal resistance
(Back-emf)

The controller is trapazoidal, however it is driving a sine wave load.
so the relationship between pack voltage and maximum controller output voltage is:

Max controller output voltage = pack voltage / sqrt(2)

The output voltage is measured between any two phases.

in the past, for ease of explanation, i havent included the sqrt(2), though i should have
The sqrt(2) is there because the pack voltage must be higher than the sine wave voltage at all times.
The ratio between RMS and peak in a sine wave is sqrt(2).

Also, what do you think the battery amp limit for can be safely increased to? I had initially pretty much removed the battery amperage limit by using a copper shunt resistor jumper during my first phase of fiddling around. The torque and acceleration was very impressive and I recall measuring about 170-180 amps at the batteries when powering up a hill at about 45 kph. I was afraid I'd burn the motor out and settled for about 20% higher than stock - or 95 amps.

i had my controller set to 130A for a long time before it blew.
for TS 40AH cells, thats about the highest id want to go.

at this and higher settings, i found the motor mount wasn't nearly strong enough to deal with the counter torque.

assuming both those limits don't exist.
The next question would be do you spend much time in battery current limit?
If you spend a significant time in battery side current limit (climbing a hill for instance), an increase in current limit will lead to a hotter motor.
motor temperature rise (above ambient) is determined by your average current.
at some temperature, the hall sensors are the first to fail (when this happens, sometimes it takes out the controller aswell).

for my commute, i spend very little time in battery side current limit (no major hills, and few accelerations), so increases in current limit didn't make a measurable difference in motor temperature.

I know the FETs on the powerstage will do 40A continuous each before they desolder themselves.
so thats 80A for each pair.
because they are on only half the time, they can do 113A for half duty ( sqrt(2*(80^2))=113A )
so thats 113A phase current.
which is 195A imaginary between the phases, motorside current ive been referring to in the past. ( 113A*Sqrt(3) = 195A )

since the controller already limits motor-side current to this level, its not likely to fail, unless driven at this current for longer than your battery pack could maintain it.

How is the motor current defined (and measured)? Is this the forward-or-backward amps in any one phase, or the absolute value of the current in all phases at any point in the motor rotation, or is it an RMS value like AC current?

the definition for motor current i have been using is current between the phases.
This is just a mathematical value.
What you would actually measure is current on a single phase (since its a balanced load, they will all be the same, just displaced).

so where i have said 200A motor side current,
i actually mean the phase current is:
200/sqrt(3) = 115A RMS

The current is indeed AC, but definately not a sine wave. it is an RMS measurement.

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

Bruce_Wayne
Offline
Last seen: 14 years 8 months ago
Joined: Tuesday, September 18, 2007 - 17:45
Points: 67
Re: Controller Selection

Ok, Excellent info guys. I don't know much about dc motor controllers so I'm gonna have to learn as I go. I was always under the impression the motor controller just controlled how much current should be pulled from the battery pack, and that same amount of current went right to the motor. You're telling me the pack and motor have way different loads? 75 amp as opposed to 200 amp. That's interesting. I thought it was much simpler than that.

I understand the regen as the ability of the controller to simply switch polarity of the motor leads at will, no? Regen should not be rocket science here. I just have to be mindful of the regen current rate. So, if I were to get the 48 volt 300 amp model, I could program separate motor and pack current rates into the controller? Say, max it out at 300 amp motor current, and 100% pack current, I should get a quicker bike no? More torque? Probably not more speed though. Have you played with this? Right now the bike does a little over 40. I'd like to get it up to 45. That would be fabulous. Matt, can you put a video of your bike in operation so I can hear the acceleration of the motor with the kelly controller on it? You should hear a slight high pitched whining, not soft clicking like I get with the original controller. I will upgrade it just to get rid of that clicking. It's just not right for an EV to sound like that.

Bruce out.

drdwl
Offline
Last seen: 13 years 8 months ago
Joined: Saturday, October 24, 2009 - 12:22
Points: 3
Re: Controller Selection

can you send me a copy of the emax wiring diagram. all of a sudden the red wire to the power light on the console shorted out - now the bike doesnt work at all - no power anywhere. do you think the controller was damaged (doesnt seem to be shorted- not smelly either).

doug

Kelly Controls LLC
Offline
Last seen: 8 years 7 months ago
Joined: Saturday, September 19, 2009 - 00:23
Points: 86
Re: Controller Selection

Sorry,we have updated the website.So I don't know the specific controller model you were talking about.
1.We have added a sketch to indicate how to connect Kelly controller to the computer on website.
http://www.kellycontroller.com/ConnectHelp.php

2.It is difficult for common customers to find a correct hall/phase combination between the controller and the motor.There are 36 combinations to try.
it is a hard work.But we provide hall/phase timing diagram for customers to find the correct combination easily.Needless to say you need find an oscillograph prior to do the test.If you want to get the timing,please contact me sales@kelly controller.com

3.By default the controller work with brake switch regen mode.What you have to do is release throttle and turn on the brake switch for the regen to occur.
If you want to get a higher regen current,you need choose brake analog regen mode.You have to turn on the brake switch to start the regen,then vary the regen with 0-5V signal.
Of course you need install brake pedal with a microswitch which can be used as brake switch.Please enable brake sensor type in GUI if you install it.

Fany Chen.

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • xovacharging
  • stuuno
  • marce002
  • Heiwarsot
  • headsupcorporation

Support V is for Voltage