Just want to post a good report of my experience with Kelly controllers.
It was cheaper to buy a new Kelly controller than have my old Curtis controller repaired.
I was advised to go for a higher rated Kelly to replace the Curtis, so I did, (as the ratings are for peak amps), but I am really pleased with it. Free software and cable, good price, free postage, and it works well. It was simpler to set-up.
The nightmare I was having with Curtis repair and config is over. Fany from Kelly was helpful by email, and the controller is better in every way!
That's no surprise to me :-)
However, without Fany C. I am pretty sure Kelly Controlls would be on the losing side, as their controller technology is pretty outdated by now... But Fany's excellent technical support makes up for some of that.
My rides:
2017 Zero S ZF6.5 11kW, erider Thunder 5kW
Just out of interest which manufacturer would you say is 'using the latest' technology?
I did look at the Sevcon's but they were expensive and again the programming issue.
What about Altrax how do they compare?
If you are asking about Alltrax I must assume you were shopping for a brushed DC-Motor controller, not for a brushless "permanent magnet synchronous motor" (PMSM)? I was referring to the latter, where Sevcon, Golden Motor FOC, Sabvoton and Emsiso already sell field-oriented ("sine wave") controllers that allow very smooth and efficient motor control, but are a hell of a lot more complex to program and first must be characterized to the motor they are to drive.
The simple block-commutating Kelly KEB and KBL controllers are a lot simpler to program but are very rough at low rpm and have lower efficiency when compared to a well-made FO controller. Plus the FO controllers can be designed to allow a form of "field weakening" even in permanent magnet motors to achieve higher idle RPM and thus also higher top speed with a given battery voltage.
But again, all that is irrelevant to you if you were shopping for a brushed DC-Motor controller such as a Kelly KDS...
My rides:
2017 Zero S ZF6.5 11kW, erider Thunder 5kW
"Very rough" is about right. Starting up a steep hill from a stop give the rider a "vibro-massage".
Also, don't get the "waterproof" option. It is achieved by filling the entire case with potting resin - a terrible idea considering the thermal strains the components are subjected to. My "waterproof" controller lasted only a couple short test rides. And Fani never reimbursed me for the cost difference between the "waterproof and the non-waterproof replacement, he basically said that I should be happy I was getting free shipping for the replacement.
By the way, my replacement Kelly is occasionally doing the same thing the original Kelly on the scooter did - occasional periods of rapid cyclic cutout of the regen.
Yes it was a KDZ Sepex controller, and I didn't go for the 'waterproofing' resin!
But I can honestly say what I got from Kelly was good value, and the whole process was a breath of fresh air compared to the Curtis.
I really think Curtis could learn from them that openly available programing software is the way to go.
If I buy another EV I will avoid SEPEX systems, I don't like having that many wires, and brushes! It seems a lot of extra messing around for small benefit.
People talk about the Regen benefits of Sepex but in reality the percentage of energy returned is very small.
Permanent Magnets saves more by not having to use any battery energy to power the field, which can amount to 20%+ of total.
@MERoller what is a 'FO controller' and how can any type of field weakening be programmed into Permanent Magnet motors?
I can't think of physically how that could be done?
A field-oriented controller tunes the two directions in which the magnetic field in the motor can go according to rotor position, rotor speed and rider input / torque required.
One is the classic torque-producing direction (either accelerating or deccelerating the rotor) and is smoothly increased and decreased according to rotor position (usually in a sinusoidal shape), and the other is the radial direction that produces no torque and is mainly active if current is applied to a coil while it should be going through a "commutation" zero-point. That apparently is the point at which a current puls is sent through the coil, thus inducing an electrical current in the permanent magnet, which in turn temporarily "weakens" the PM's magnetic field.
Of course this is far less efficient than with your SEPEX setup, which can easily weaken the rotor field strength by reducing the exitation current, but I am not sure if the Kelly KDZ do this at high rpm automatically (when programmed to do so), and maybe you don't need such high motor rpms anyway...
My rides:
2017 Zero S ZF6.5 11kW, erider Thunder 5kW
My impression is that non-Chinese motor controller manufacturers are oriented to serving large manufacturers and industrial/mining applications - so it makes sense to sell the programming software separately for a high price becasue the manufacturer is going to use it for programming or servicing many units and they specifically don't want the end-user to be programming the controllers themselves - there could be safety issues involved. They (specifically their lawyers) don't like home-builders and rather not deal with the small EV-conversion cottage industry. Same with non-Chinese lithium cell manufacturers.