Sensorless Controller 36v 350w Shenzhen Ecrazyman Keywin

[zukster]

Sensorless Controller 36v 350w Shenzhen Sucteam Ecrazyman Keywin

I don't know what to call it. Keywin does not list it on his ebay site.

http://stores.shop.ebay.com/HuaQiang-North-Road__W0QQ_armrsZ1

You have to email him about it: ecrazyman@gmail.com

I just finally tested this thing and I'm pretty happy with it.

My setup is:

Bafang SWXH Geared Brushless Rear Hub Motor
2 x 20v Canadian Tire Yardworks Batteries = 40v Pack
and the 36v 350w Sensorless Controller
Keywin right-handed twist throttle

So far the sensorless controller works great at 40v (freshly charged
batteries are max 42v). I havn't tried it with 3 batteries yet.

It seems pretty much like the hall controller except the very first
part of the throttle, which people have been asking about.

It hammers on a tinny bit as opposed to the hall throttle. By that, I
mean I can here a click in the motor when I first start turning the
throttle. It's pretty minor and I don't think its a problem.

If its really feeding the motor too much power at startup - well, only
time will tell if its hard on the system.

Note that the batteries are waterproof but I haven't done the connections yet.
You can see the hall 5 pin block but nothings plugged into it. I'll prob just seal
it up rather than cut it off.

 

[floatingdog]

Thanks for the post. Great looking build. Give us some more thoughts after you get a few miles. I want to buy one of the new senorless Bafang models, so a sensorless controller is my only option.

Tell us about the controller. Did it come with the heavy looking wires and Andersons or did you do some mods?

Given the hall wire connection, do you think this is a dual mode controller? I.e., works with or without halls?

 

[dnmun]

the hall sensor plug is coming from the motor. the controller is sensorless so it doesn't have a capability for hall sensor input.

can you reduce the 'hammering' by letting up on the throttle or does it just die out as the speed increases. or is it from a dead stop?

 

[zukster]

I didn't know there are 3-pin Bafang brushless hubs. I just checked them out. If these are out
now, I suspect they'll all be 3-pin soon. Too many potential problems with those tiny hall wires,
sensors, and connectors. I originally bought the sensorless controller a few months ago for a
different motor for which the hall sensors stopped working, then ended up using it for this
bike because I accidentally hooked up 4 20v Yardworks batteries in a row (80v+) to one of
keywin's hall 36v 350w controllers - and smoked it. I've also had problems with those little
wires fraying and have had to repair them. So from a manufactures point of view, they'll
have less warranty issues from the 3-pin motors.

Does anyone have one of the new Bafang's that have the waterproof connectors? I'd love to
see what they use. I really hate Anderson's for this kind of exposed stuff.

When I first told keywin about the motor with the bad hall he said, "i think the hall of the
motor was burned ,but easy to solve is ,use sensorless controller or add a sensorless module
on on the controller will ok." So I think it might just be a module that is added to the normal
controller. I ordered the whole controller but I didn't open it up. It only has 3 wires for the
motor coming out. No hall. So perhaps the hall module is not included when you order it
whole (too save cost). He charged the same price as the other one (but he might have gave
me a good deal because I had a problem with the hall on the motor).

I'm sure you could figure out how to dual mode one if you really wanted to. I think I'll just
go with the sensorless versions from here on in.

Actually, maybe someone could even tell me if the controller I over-volted 80v+ might be
fixable. Here's a picture of the cylindrical piece that cracked. The burnt smell mainly
seems to come from this piece too, although other components could be damaged...

 

[zukster]

can you reduce the 'hammering' by letting up on the throttle or does it just die out as the speed increases. or is it from a dead stop?

Once the throttle is on, its gone. Just a click/noise in the motor when you first kick in the throttle.
I think hammering was the wrong word to use. It does not stutter under load. You can hear the click
mainly when you 1st engage the throttle at lower speeds. At higher speeds, it seems to be gone, but
maybe that's because the motor is already rotating so you can't hear it. Like I said, its really minor.
I was out for a 20km run today, and pretty much forgot about it. I was pulling a trailer and 5 year
old too.

 

[biohazardman]

The capacitor is toast but easy to replace. Difficult to say if more damage was done. You can replace the cap, they are very inexpensive, then try it to find out. As for the clicking try getting the bike moving a bit faster and add the throttle a bit slower and see if that helps. My guess is that 80 volts is a bit much for the motor you are using. Does not look like the controller liked it either both might do better with 48v. Good luck.

 

[marshy]

Given the hall wire connection, do you think this is a dual mode controller? I.e., works with or without halls?

I've ordered a "sensorless" controller from Keywin that he claims can also be used with sensors. Apparently he's supplying it with the sensor connections to the controller than I can use if I wish. Not sure if there are any other mods required to make the switch, but I should have the controller within the next few days and can provide more info then.

I was looking for a controller that I could use with sensors and fall back to sensorless in the event of problems with the halls or those fragile little wires. Good ol' Keywin came to the rescue.

 

[dnmun]

can you read the printing on the side of the cap? i have some 100V1000uF but that looks like 63V. lemme know, or if you order another one from keywin, get him to put a cap in the box when he ships. maybe 2.

 

[zukster]

can you read the printing on the side of the cap? i have some 100V1000uF but that looks like 63V. lemme know, or if you order another one from keywin, get him to put a cap in the box when he ships. maybe 2.

Hey - thanks. I just came back from my workbench from writing down the numbers. Also, can they be
upgraded to 100v 1000uF, or do you have to stick with what it came with? What it is:

63v 470uF (I suspect that the u is for micro, but I don't know how to write the greek letter mu )
80 II LH
+105 degrees C
CE NHG

Maybe I should just try replacing it before ordering a new controller... I'm not much good at detailed
soldering, but it looks like an easier piece to change. Where would I get on of these? I'd rather not
have to mail order such a small part. I'm in Vancouver, BC, Canada. Humm - I bet Justin at ebikes.ca
would have one. Any other suggestions?

 

[zukster]

The capacitor is toast but easy to replace. Difficult to say if more damage was done.

The bike actually ran for a while after my wiring mix up. About 15 minutes.

I had two 60v packs with me and a bunch of cables that put 2 batteries in series. I accidentally
added a 40v pack + a 40v pack = 80v with one of these cables, instead of a 40v + 20v pack like I
usually run with. The controller runs great and was in use almost every day with a 60v pack.

Keywin told me before that the controller was good up to 65v. The batteries never charge
passed 21v so my pack was never over 63v. This was no problem. That must be absolute max for
the controller if that capacitor says 63v. Must be a bit of margin for error though.

I just emailed Keywin regarding getting another controller. He said that the current one he is
selling is only good up to 48v. So I have asked him if components have been changed on the controller
compared to what I originally bought that now makes the limit 48v now... or is he trying to get people
to use the controller well below the safety threshold (ie. CYA).

I'll let everyone know.

What matters to me is that if I used both a 40v and 60v pack on the original controller,
for six months, often pulling a trailer and 5 year old, then that's what I call reliable.

I want to stick with the little controller if it can do the job. Its light, small, cheap, and by
nature it limits the current that goes to the motor. Over that six months of use, often
going up steep hills with full-on throttle, for 10-15 minutes straight, nothing has ever
overheated - with either battery pack. Also, my son's been using the same setup to and
from school (10km) for about six months too. I haven't touched his bike since I put
it together. Just charge and go.

As for the clicking try getting the bike moving a bit faster and
add the throttle a bit slower and see if that helps.

Like I said, I barely notice it. And it doesn't happen at all once the bike is at 5km/h
or more. I'm pretty fussy; if I had any reason to think that the hall controller was
better, I'd use it. I think these sensorless are fine, so I'll be using them from now on.

 

[zukster]

I've ordered a "sensorless" controller from Keywin that he claims can also be used with sensors. Apparently he's supplying it with the sensor connections to the controller than I can use if I wish. Not sure if there are any other mods required to make the switch, but I should have the controller within the next few days and can provide more info then.

That sounds pretty cool. Especially for comparing. Switch back and for from hall to sensorless under the same
test conditions. Keywin must have ran the hall wires out of the case for you. I wonder if you need to use a
switch or something though. Maybe the sensorless EMF module feeds the hall signals to controller. In which case
you'd want any output from the sensorless module swithed off completely if using the hall signals from the motor.

Keep us posted!

 

[dnmun]

i have some 63V 470uF caps here in my parts pile. took them off some other controllers when i upgrade them to 72V lifepo4.

the reason it says 63V is because that is considered the highest voltage it can sustain.

you may have not blown it up until later after your brief overvoltage to 80-90V, if you were running at 63-65V like you said.

maybe they will send allow me to mail it if i put it in a big envelope. lemme know.

 

[zukster]

i have some 63V 470uF caps here in my parts pile. took them off some other controllers
when i upgrade them to 72V lifepo4. the reason it says 63V is because that is considered the highest voltage
it can sustain. you may have not blown it up until later after your brief overvoltage to 80-90V, if you
were running at 63-65V like you said. maybe they will send allow me to mail it if i put it in a big envelope.
lemme know.

You wouldn't mind doing that? I'm up here in Canada, eh. I also have a keywin 30a controller that I
tried b4 it was too much for the motor/batteries (you probably remember that thread - I toasted one
of my first Bafang motors ) I guess my other option would be to either try to use it but limit the
current coming out of it (have to read up on that) or pull a cap off of it, but that seems a waste of
a good controller. You can use a bigger cap right?

 

[zukster]

I just tested the sensorless controller at 60V and the initial throttle-on stutter is more
pronounced. I'll have to try it for a while at 60v before I decide if its smooth enough
for my general use. It was definitely fine at 40v, but perhaps at higher v than that and
you're better off using a hall controller, as mentioned by others.

If manufacturers like Bafang stop making hall motors, what are all us over-volters
going to do Open up the motors and run in our own sensors? Not me anyway.

 

[marshy]

Given the hall wire connection, do you think this is a dual mode controller? I.e., works with or without halls?

I've ordered a "sensorless" controller from Keywin that he claims can also be used with sensors. Apparently he's supplying it with the sensor connections to the controller than I can use if I wish. Not sure if there are any other mods required to make the switch, but I should have the controller within the next few days and can provide more info then.

I was looking for a controller that I could use with sensors and fall back to sensorless in the event of problems with the halls or those fragile little wires. Good ol' Keywin came to the rescue.

I've now got my ecrazyman sensorless controller (with the "optional" sensor wires) and done some limited testing. Observations so far with an unspoked SWXH motor (no wheel, no load) at 36V:
- motor runs backwards unless blue and yellow phase wires are switched (hall sensor wires disconnected)
- there is some minor "hammer" or stutter on startup from complete stop as reported above, nothing serious I assume but makes me cringe a little. Performance is smooth once motor is turning.
- with hall sensors connected in a suitable combination it also runs fine...however, it still has the startup stutter. So there doesn't seem to be any benefit actually connecting the hall sensor wires. Note that I only tried a few hall sensor combinations, perhaps there is another one that performs like a sensored controller, i.e. without the stutter. I'll check with Keywin on this point.



-

 

[dnmun]

sounds like the zero crossing on the sensorless is causing the mosfet drivers to fire before the signal comes from the hall sensors. which means the timing of the zero crossing needs to be adjusted back so the phase is not triggered before it should be.

 

[zukster]

Here was the wiring instruction Keywin gave me:

1)controller hall sensor to motor hall sensor

red to red
Black to black
yellow to blue
green to green
blue to yellow

2)controller output to motor

yellow to blue
blue to yellow
green to green

3)please refer the attached photo connect the wires

Battery bold red wire:battery + bold black wire:battery - thin
orange(or red) wireower lock (connection to battery + will be power on)

throttle red :+5V black:GND green:speed signal

brake connect the yellow and black will be braking

other: the single thin yellow wire is positive voltage brake (connect to battery+ will be braking)

 

[zukster]

I've been trying the controller more at 60v lately, and its definitely more quirky.
When going downhill fast, then applying the throttle fast, I sometimes get the
hammering too. The start-up hammering is minor once you get use to it and make
sure you're going a bit before hitting it hard. What awesome torque at 60v

I'm probably going to switch back to the hall version however, because there are
no issues at all with it. For motors where the hall is damaged - the sensorless
controller will be a great backup.

I should also mention here that Keywin is now saying that running it at 60v is
unsafe and I should only do 48v. At first, back in April 2009, he told me that
60v was okay, so I'm trying to figure out which it is, or if something has
been changed on newer ones.

Maybe one of you electronics wizards can comment on this?

At first he said(re sensorless controller I bought in April),

Hello Jeff ,
because i moving my lab in these days ,so i am very busy ,6 gears total price is 45 usd ,the 36v/350w controller able working at 65v, but the 60v battery max voltage over 69v ,that is unsafety ,thank you.

And now he said if I order another one,

Hello Jeff ,
nice to meet you again ,if you only need one controller ,the price is 42 usd included shipping ,but it is working at max 48v battery , working at 60v battery is unsafety ,also the sensorless controller working at 60v battery is unsafety ,thank you .

then in a follow-up email,

Hello Jeff ,
i think you mistake the battery with voltage ,the controller can working at 60v voltage but can not working at 60v battery(max voltage is 13.8*5=69v),thank you.

So am I missing something here? Each Yardworks battery measures less than 21v after charging.
20.7v, 20.8v, etc. So my total pack is about 62.5v with fresh batteries.

 

[dnmun]

like he said, 60V true voltage is not too much, but 5 SLA in series would be 69V and that is too much for the 63V caps. you are running at 62.5 or less so you would be ok. i also think it may just be marginal to run 5 SLA if they are not too hot. all imo

 

[zukster]

Well 6 months at 62.5v means something.

Any idea what the engineering tolerance is in a capacitor? Like if it says 10v, is there like
a 10% margin, so you're prob okay up to 11v or something like that?

 

[marshy]

sounds like the zero crossing on the sensorless is causing the mosfet drivers to fire before the signal comes from the hall sensors. which means the timing of the zero crossing needs to be adjusted back so the phase is not triggered before it should be.

dnmun, is this likely to be an adjustment I can make myself (e.g. firmware update, pot tuning, etc) or am I stuck with it?

I asked Keywin about this a while back, no response yet. Ultimately I don't mind if it only works sensorless (just disappointing), but for others dual operation might be critical.

 

[dnmun]

capacitors have specs of plus/minus 20%. so 75V fits inside there.

i would expect there is a way to adjust the values of the resistors and capacitors in the feedback circuit, but way over my head.

 

[Jeremy Harris]

The +/- 20% tolerance is ONLY on the capacitance value. The voltage rating is a "do not exceed" figure. If it's rated at 63 volts, then that is the maximum rating.

You might, possibly, get away with a few volts over the rating for a while, but you can be certain that the life of the capacitor will be shortened.

Jeremy

 

[dnmun]

learned something else new.

 

[zukster]

The +/- 20% tolerance is ONLY on the capacitance value. The voltage rating is a "do not exceed" figure. If it's rated at 63 volts, then that is the maximum rating.

You might, possibly, get away with a few volts over the rating for a while, but you can be certain that the life of the capacitor will be shortened.

Jeremy

Well if I'm at 62.5V (just checked freshly charged pack again), and the cap says 63v, that few extra volts gives me the comfort zone I was looking for.

Thanks for the info guys.