Hightekbikes controller vs Crystalyte with 2806 hub

[stevero2001]

When I got my nine continents 2806 hub motor, I originally got it with a 20Amp first pedal controller with 4110 FETS. This is a sensorless controller. I was running this with 36V lithium. I have a steep hill I ride up and down everyday, and I noticed that when my coast speed exceeded the no load speed of the hub, which is around 30MPH in 26 inch wheel, it would suddenly lock up the controller, and act like a plug brake, only very uneven and harsh. This is not something I wanted. The only way to clear the problem was to turn off the power to the controller, and then it would freewheel again. After some back and forth with Justin at ebikes, we came to realize this is normal. It can be solved by putting a diode in line between the controller and the battery, but then you have the voltage drop and power dissipation of the diode. I’m a bit surprised more haven’t complained about this.

Another thing I noticed with this controller is there was considerable cogging, which was significant at speeds over 15MPH or so when running with the throttle off, but the controller on. What became apparent is, this controller is always in regen when coasting. This is something I found annoying, and would prefer not to have, as it slows the bike down, and means you always need to have a little throttle on, even going down hills. Another problem is, the phase wires would heat up noticeably at the anderson connectors, even with currents of around 17A, and the controller would get fairly warm. Also, the no load amp draw was around 2.5A at full throttle.

After all this, I decided to try a different controller. I went with the controller sold by hightekbikes, normally used with the AOETEMA motors. This is also a sensorless controller, but it starts instantly, unlike the Crystalyte controller pedal first. It’s no load current is around 1A, and when the throttle is off, there is zero cogging. It also seems to run a bit smoother, but it has a higher pitched almost ringing sound. I have it modified to put out up to 28A, although on hills, it stays under 25A, but even after a 10% 1 mile long hill climb at 20A+, the controller barely gets warm, and the connectors stay cool to the touch. It is clearly a more efficient controller. I have even rode 5 miles on flat ground at average 15MPH with the throttle off, so the wheel drag is not too bad.

I really like this combination, and the price was reasonable. The only thing I still find annoying is, when the nine continents is under load up a hill, it sometimes makes a scraping noise. Letting off the throttls, the noise goes away. When I get up to speed, the noises goes away. This seems like more than resonance noises. It sounds almost like the rotor is scraping against the magnets. The problem is, it’s not consistent. Sometimes I can be drawing 24A with little noise, and other times at around 16A, there’s the noise. I wonder if anyone else has experienced this. This is my second 2806 motor, and they both make the same sound.

[chvidgov.bc.ca]

Do you think the desirable effects of this controller would translate with Crystalyte motors as well?

[dogman]

Yes, though underkill, I'm real happy with how my 5304 is running on the HTB controller. I'll solder the shunt later if I ever get a better set of batts.

[stevero2001]

I would definitely try it with a Crystalyte motor. The big advantage is, no hall effects, so it's more reliable in wet areas, and you still have immediate start. I only pushed mine to 28Amp. It could go higher, but I didn't want to push it to much. It only draws 28A from a dead start, but settles around 24A when pulling a steep hill, and the torque is great. I can go up a 10% average grade at 14MPH with no pedaling. This is with approximate total weight of 220 lbs (me and bike). I usually pedal with the motor, and go average of 18 to 20MPH up the same hill.

As far as my noise problem on the 2806, it seems to be speed related more than anything, so probably a resonance issue. It seems to get the noisiest at around 18MPH. Above or below that, the noise goes away, so I don't think anything is really scraping, but it is annoying.

[stevero2001]

I just discovered that when I go down steep hills where I'm going faster than no load speed, that this controller will regen when you put full throttle, causing some braking, but it is a nice gentle deceleration. I noticed my watts up meter showed an increasing voltage when I this was happening, but amps show zero, since watts up won't show negative values, so I don't know how much current it is regening, but in any case, it only does it when I twist the throttle, so this is also a nice way to slow the bike down, when I don't want to go too fast, and get a little energy back to boot.

[dogman]

I could be wrong, but don't most controllers do that? But they don't actually put the power into the battery?

I have noticed the same thing, no throttle I reach a higher speed going down a steep long hill than I do if I put on just a touch of throttle. With just a touch of throttle, I seem to reach a top speed close to the no load speed. It makes the motor hotter than going downhill with no throttle too.

I saw the same effect with the WE brushed controllers too. It must have to do with back EMF, but I'll never fully understand that stuff.

[stevero2001]

I used to have a brushed BD36, and I didn't notice that. but It has a no load speed of around 35MPH, and I rarely went that fast. I'm pretty sure it puts something back in the battery, because the voltage increases when I'm going faster than no load speed, and drops back when I let off the throttle. In any case, the Crystalite always had a drag on the motor even with the throttle off, and this one doesn't drag at all with no throttle, so in cases where you want to go faster down a hill, that's a big plus.

One thing I noticed, that might be due to wind changes, it seems after I let it regen for awhile, it feels like the max speed is lower at the same current, so it seems like the controller was partially damaged doing this. It could be my imagination, since I don't have a speedometer, but I feel like I'm pedaling slower at the same current draw, than before. Anyway, now I'm not sure it's such a good idea to let it regen like that. It seems to make more noise around 20MPH than before too. a couple MPH above or below that, and the noise goes away. Around 10mph it makes a noise too, so it seems to be a harmonic at those speeds that causes the noise.

[stevero2001]

I tightened the spokes on my front wheel, and found that this has eliminated the noise I was getting at 20MPH, so it appears that the resonance at that speed was vibrating loose spokes. It still has a resonant noise at around 10MPH, but it is not as annoying. It just sounds like a low ringing sound, but without the grating, rattling sound I heard before, and since I rarely stay at 10MPH for more than a couple seconds this is acceptable. This explains why it seemed to be getting worse, as the spokes were loosening over time.

[jag]

I have a similar combination, Crystallyte 72V 4110FET and a 9C 2809.

Notice "roughness"/noise if giving full throttle at low speed when going up hill.
For me that is 25-30A at 60-70V (72V nominal pack).
The rougness goes away when over 20-25km/h or so and then one can "blirp" to full
throttle with no noticable noise/roughness, just a nice kick forward from the 2kW to the motor.

I thought the roughness might be due to the pedal first controller not being capable of
quite getting the signal modulation right at low speeds and high power.

I was curious if people with sensored controllers notice this?

[stevero2001]

I don't know, but I did not like the Crystalite controller. The Hightek controller is much smoother IMO, but it is only rated to go up to 48V, and I only use it at 40V max, 36V typical with 28A, but it gives me all the power I need. I might upgrade to 48V later for more speed, when my battery dies, but that will likely be at least another 3 years. Overall, I'm quite happy with the performance I'm getting.