New Kelly Sensorless Controllers?

fechter

fechter

Moderator
Staff member
Joined
Dec 31, 2006
Messages
16,806
Location
California Bay Area, USA
I stumbled across these on the Kelly website:
https://www.kellycontroller.com/shop/?mod=cat&cat_id=41,53,56

Their 72v 100A model would be cost competitive with a Castle Creations (but much larger)
They have models up to a 72v 600A model.

Has anyone ever tried one of these? This could be a good solution for the high powered RC drives. It may also be a good solution for the BMC motors that seem to have timing issues when using hall sensors.

Here's what they say about the 100A model:
KSL72100,24V-72V,100A,Sensorless BLDC Controller

Weight: 5 lbs.
Stock: 50 Beta Test Version
KSL controllers are produced for sensorless BLDC motors, especially for fans, water pump, air-conditioning, compressor, etc.
Parameters should be set correctly before use!
Features:
•Intelligence with powerful microprocessor.
•Synchronous rectification, ultra low drop and fast PWM to achieve very high efficiency.
•Current limit.
•Battery current limiting available. Doesn’t affect taking off performance.
•Low EMC.
•LED fault code helps user debugging.
•Battery protection: current cut back, shutdown and warning at high or low battery.
•Rugged aluminum housing for maximum heat dissipation. Rugged connectors.
•Thermal protection: current cuts back on low temperature and high temperature to protect battery and controller.
•Current multiplication: Take less current from battery, output more current to motor!
•Easy installation: 3-wire potentiometer will work.
•Stuck Rotor protection.

General Specifications:
•PWR Supply : 24V-72V
•Frequency of Operation: 16.6kHz.
•Standby Battery Current: < 1mA.
•5V Sensor Supply Current: 40mA.
•Supply Voltage, PWR, 18V to 90V.
•Supply Voltage, PWR, 8V to 30V for controller rated higher than 72V.
•Supply Current, PWR, <150mA.
•Operating Voltage, B+, 18V to 1.25*Nominal Voltage.
•Analog Throttle Input: 0-5 Volts. Producing 0-5V signal with 3-wire pot.
•Reverse Alarm, Meter, Main Contactor Coil Driver: <200mA.( NA*)
•Full Power Operating Temperature Range: 0C to 50C (controller case temperature).
•Operating Temperature Range: -30C to 90C, 100C shutdown (controller case temperature).
•Motor Current Limit, boost current: 100A/50A, depending on model.
•Motor Current Limit, continuous: 40A/20A, depending on model.
•Max Battery Current : Configurable.

Optional Features:
Optional Waterproof : The price is 38 U.S. dollars.
 
I wonder how it would handle high inductance motors. Maybe the throttle advance is slow enough that it works at 16k switching.


Cool to see some options.
 
I'd be interested to know if it would syncrhonize well with an RC motor (With the relatively quickly changing load and relatively high electrical RPM). I had a lot of problems on my crystalyte sensorless controller with that issue and I suspect anything not specifically designed for high electrical RPM and quickly changing loads would have a relatively high chance of problems.

I heard that there was a sensorless version of the keywin controllers coming out and given how well his sensored version work with RC motors, it seems like there might be a chance that the sensorless versions will. That critcally depends, though, on how quick the algorithm is and the speed of the comparators or ADC and the overall design.
 
Interesting news indeed...
fechter said:
•Current multiplication: Take less current from battery, output more current to motor!
Click to expand...
Very funny thing to put in their "feature" list. I don't know of any motor controller that doesn't do this!! :lol:


fechter said:
•Motor Current Limit, boost current: 100A/50A, depending on model.
•Motor Current Limit, continuous: 40A/20A, depending on model.
Click to expand...
40A continous motor current doesn't sound like very much for a modern high end controller such as this. They probably use cheap-ass FETs on their lower-end models (as we've seen recently in another thread)... I'm still wondering if it's possible to just drop in better FETs to get better performance without having to modify the FET drive circuitry?
 
ZapPat said:
40A continous motor current doesn't sound like very much for a modern high end controller such as this. They probably use cheap-ass FETs on their lower-end models (as we've seen recently in another thread)... I'm still wondering if it's possible to just drop in better FETs to get better performance without having to modify the FET drive circuitry?
Click to expand...

Is it as simple as comparing gate capacitances?
 
swbluto said:
Is it as simple as comparing gate capacitances?
Click to expand...
That can be very important but other parameters can be important too, depending on several factors:

- Switching speed (closely tied to gate cap. but can vary for same cap. from MOSFET to MOSFET)
- Junction-to-case thermal resistance.
- Case-to-sink thermal resistance.
- Threshold voltage (if available gate voltage is fairly low).
- On-state thermal resistance
- Pinout (almost all 3-pin MOSFETs are the same but there's an occasional misfit that could really ruin your day :) ).
- Case size (TO-220, TO-3P, TO-247, TO-264, any of several SMD variants).
 
zukster said:
Keywin / ecrazyman sells these too.
Click to expand...

Link? I'm always interested in picking up new controllers to check out.
 
Don't pick up this one to try out. There are two serious problems right now making them really suck for bikes. The first is that the throttle won't shut off the motor. So at zero throttle, the motor is still spinning. This is fine for a pump or fan running off a PLC, but on a vehicle it's REALLY annoying to have to flip a switch to start the motor, and then flip it again to shut it off. The other problem I'm having is that when I quickly advance the throttle, the motor goes to 100% on its own for maybe 5 seconds...then decides to follow my throttle signal again. This scared the $hi1 out of me the first time it did it....I'm glad I was running on 24V at the time or it would have been bad.

Oh, they also can't get torque control working correctly. So it will have the same terrible throttle feel as any other RC ESC.

Vincent at Kelly said he's working on the torque control, and the runaway motor problem...but he didn't say anything about fixing the motor start sequence to make it more like a normal RC controller.

So far I've tried the following ESCs...Castle HV, Hobbywing Platinum, Hifei 120HV, Schulze 160A xxl, Jeti Spin 200, Kelly KSL48-100. I'm still waiting for one that does proper torque control. Right now they either have really nasty "bang bang" feel to the throttle...or some sort of throttle ramping feature which is a lot safer but takes all the thrill out it :(
 
liveforphysics said:
zukster said:
Keywin / ecrazyman sells these too.
Click to expand...

Link? I'm always interested in picking up new controllers to check out.
Click to expand...

I haven't tried mine yet but will soon and will post how well it works. I got the 36V/350w one
which keywin sets up so it can take up to 65+ V if you ask him.

ecrazyman@gmail.com

His ebay store is http://shop.ebay.com/e-crazyman/m.html?_nkw=&_armrs=1&_from=&_ipg=

I don't see it on here. It may be new and still under development so you will have to email and ask
about potential problems as discovered above with the Kelly's. Of course, a certain sensorless controllers,
or sensorless modules, may work better with some controllers/motors that others.

Keywin also sells a sensorless add-on for the controllers he sells. I burnt out the hall sensors on my
Bafang, so that's why I bought one. I my case I just bought the whole sensorless controller because
I didn't want to mess around with soldering in an sensorless module,
 
CNCAddict said:
Don't pick up this one to try out. There are two serious problems right now making them really suck for bikes. The first is that the throttle won't shut off the motor. So at zero throttle, the motor is still spinning. This is fine for a pump or fan running off a PLC, but on a vehicle it's REALLY annoying to have to flip a switch to start the motor, and then flip it again to shut it off. The other problem I'm having is that when I quickly advance the throttle, the motor goes to 100% on its own for maybe 5 seconds...then decides to follow my throttle signal again. This scared the $hi1 out of me the first time it did it....I'm glad I was running on 24V at the time or it would have been bad.

Oh, they also can't get torque control working correctly. So it will have the same terrible throttle feel as any other RC ESC.

Vincent at Kelly said he's working on the torque control, and the runaway motor problem...but he didn't say anything about fixing the motor start sequence to make it more like a normal RC controller.

So far I've tried the following ESCs...Castle HV, Hobbywing Platinum, Hifei 120HV, Schulze 160A xxl, Jeti Spin 200, Kelly KSL48-100. I'm still waiting for one that does proper torque control. Right now they either have really nasty "bang bang" feel to the throttle...or some sort of throttle ramping feature which is a lot safer but takes all the thrill out it :(
Click to expand...

Hello Dave. I saw your posts on the RCGroups DIY controller thread stretching all the way back to 2005! I thought it was interesting you had an interest in powering an electric scooter all the way back then. Have you had any progress?

Anyways, I think the "bang bang" or non-thrilling throttle ramp is pretty much a necessity with sensorless control and "heavy loads" that EVs impose on them - you might remember from one of quax's posts that industrial applications had sensorless startup that usually took somewhere upto 10 seconds implying non-thrilling acceleration; If they could safely get that time down, I'm pretty sure they would as it would broaden the potential applications. I'm assuming you're talking about starting from an absolute stop; from 2 mph onward on my electric scooter (A slight push from a stop), it seems the castle controller operates just the way I want it to.
 
Yes, I've been doing this for a long time now :) The start algorithm doesn't bother me very much, it's the acceleration I am referring to. My motor is only about 0.007ohm, and my batteries are also very low resistance. This causes serious problems when the ESC is run in "speed" mode; which is the mode that every sensorless drive uses (so far). If your motor is high resistance, then you won't notice the brutal acceleration. However, my goal is extremely high performance...but it needs to be controllable. So the low resistance motor is key, along with a good set of batteries...but the ESC needs torque control. As in 50% throttle = 50% torque. Right now 50% throttle = 50% voltage. This is a problem because even a 1V drop across my motor will be about 142A of winding current. So just inching the throttle upwards results in a bucking bronco type experience. I am having thoughts of designing my own ESC, and I'm trying to open up a dialog with some other DIY esc designers out there...but so far there has been little progress. I want to sell my bike motors soon, but this will be difficult with the state of controller design as it now stands.
 
This is my own custom design brushless motor. Not sure what you mean by RC?
 
No, all the ESC's work great with the motor. It's the application that has problems. The ESC's are all designed for spinning props or some other low inertia load, not a bike with a person onboard 8)

I don't like sensored ESC's very much, they have problems of their own. A hybrid sensored/sensorless would be best, but that is also just a dream at this point.
 
Hmmm, kelly is marketing this for pumps/compressor, etc. not for vehicles. They may or may not make this work for us, but that's not the target market and it's 100% up to them. But, yes the KSL will spin any motor at 0% throttle. When you flip the throttle switch the motor starts spinning, no matter where your throttle is. That is how it is designed.
 
I'd like to see the melexis in a real world application. Most of their demos are very low inertia which isn't a very good test IMO.
 
I've yet to learn what RC is too. Cyclone Taiwan suplied me with a kelly controller(for 36v 960w motor) was useless (sent it back)now he wants me to have another 100A max one 30% more torgue than the 650 what converted to 960w he says, hell i'll stick to what i've got till she blows as waste of money although good to have a spare. It was KBS36051L.
 
You must log in or register to reply here.

Similar threads

E
Kelly BLDC motor controller ESC 24-72V, peak. 500A, continuous 200A IP67, Express EU Shipping
Replies
0
Views
2,468
e-electrics
E
Z
    • Brainy
Device to shut off charger at a pre-set voltage?
Replies
9
Views
1,767
slitespd
slitespd
dchgarage
Kelly Controller KLS7230S (Sold)
Replies
0
Views
743
dchgarage
dchgarage
magudaman
Sold!!! Kelly KBS-X 72051xs 24-72v extra high speed version -$50 Sold!!!
Replies
3
Views
1,452
magudaman
magudaman
digitalbilly
Garmata EEB battery box says it can hold 400pcs 18650 cells. I think I can fit more.
Replies
7
Views
1,453
digitalbilly
digitalbilly
Back
Top