Speed sensor in back wheel of Powabyke Classic

[ChrisOfBristol]

That's a great help! I can see that I have the speed sensor and throttle in the right sockets, and I appear to have rewired them in the right order, red/black/green, red/yellow/green.

Your power leads appear to be in the order brown/nc/green/yellow/red/nc. Mine is brown/nc/yellow/green/red/nc/red. I don't think that's important since green and yellow are both power output and the end contact isn't connected to anything, so although my wiring means that the outer red lead is wasted, at least it isn't connected to anything it shouldn't be! I'll rearrange them to match yours just in case.

I'm getting nowhere with testing the speed sensor, so am looking at the semiconductors. So far I have tested the voltage regulator chip 7812 which is OK.

 

[WheezyRider]

That's a great help! I can see that I have the speed sensor and throttle in the right sockets, and I appear to have rewired them in the right order, red/black/green, red/yellow/green.

Your power leads appear to be in the order brown/nc/green/yellow/red/nc. Mine is brown/nc/yellow/green/red/nc/red. I don't think that's important since green and yellow are both power output and the end contact isn't connected to anything, so although my wiring means that the outer red lead is wasted, at least it isn't connected to anything it shouldn't be! I'll rearrange them to match yours just in case.

I'm getting nowhere with testing the speed sensor, so am looking at the semiconductors. So far I have tested the voltage regulator chip 7812 which is OK.

I did replace the input power cable with 12AWG wire going to an XT60 connector. It was a long time ago, but I think one end went to the keylock and then I soldered the other wire onto a a very short bit of the original wire on the power connector block. I'd have to open it up again to remind myself exactly what I did.

 

[ChrisOfBristol]

the speed sensor is circled.

I have had no success in testing the speed sensor. Might you have time to test the resistance between the pins on yours? If mine matched yours I could probably assume that it was working.

I measured 524R Black to Red and 1772R Green to Black. All other combinations were infinite resistance.

 

[WheezyRider]

I have had no success in testing the speed sensor. Might you have time to test the resistance between the pins on yours? If mine matched yours I could probably assume that it was working.

I measured 524R Black to Red and 1772R Green to Black. All other combinations were infinite resistance.

Ok, I'll have a look. From memory, the values were in the kOhm range, 524 sounds a low.

 

[ChrisOfBristol]

Ok, I'll have a look. From memory, the values were in the kOhm range, 524 sounds a low.

Thanks Wheezy!

 

[WheezyRider]

For my sensor, I get >20M for red probe on red, black probe on black, 50k for red probe on black, black probe on red. For red probe on green and black probe on black I get 64k and when the red probe is on the black and black probe is on the green I get 132k.

On the main power connector block, the yellow and green wires seem to be the negative power lead to the motor and they are connected together. So I don't think the order of the wires will make a lot of difference. The positive input from the battery goes directly to the motor via the key lock switch. This is taken to the motor by a red and blue wire, but they seem to be shorted together too. Probably not an ideal set up to have the live line directly connected to the motor...

If you get stuck, you could call in at Powabyke if you are in Bristol?

As vfr said, they used to sell a device that would defeat the operation of the sensor. Maybe it would be 555 timer based, but the question is how it would trigger the controller. It wouldn't just be a case of a sensor line picking up a positive pulse train, I imagine it has to drag the green wire close to ground to detect a pulse.

Might be worth looking at how old fashioned Hall sensors worked, before they became integrated with an amplifier and digital switching circuit.

Hope that helps.

As for me, I would rip it all out and modernise it. It makes quite a nice bike with all the junk taken off of it and a decent motor/controller/battery.

 

[vfr400]

Maybe it would be 555 timer based, but the question is how it would trigger the controller. It wouldn't just be a case of a sensor line picking up a positive pulse train

When you operate the throttle, there's an AND condition that the sensor is pulsing. The 555 would switch on and pulse as soon as you switch on the controller and remain pulsing until you switch off, in which case the AND condition is always met. I'm sure that's the same as how the PTDISABLE device works.

 

[ChrisOfBristol]

I'll comment on the easy points first and respond to the technical stuff when I've absorbed it.

On the main power connector block, the yellow and green wires seem to be the negative power lead to the motor and they are connected together. So I don't think the order of the wires will make a lot of difference

Agreed.

The positive input from the battery goes directly to the motor via the key lock switch. This is taken to the motor by a red and blue wire, but they seem to be shorted together too.

They are shorted together, but mine are red and black which makes me think that the cable is probably an old UK standard 3-phase cable.

As for me, I would rip it all out and modernise it. It makes quite a nice bike with all the junk taken off of it and a decent motor/controller/battery.

It would also require replacing the charger (to match the batteries), and the throttle (as the standard one works backwards electrically) and the left-hand brake lever (the cutout switch works backwards electrically), I'd probably replace both brake levers so they matched. There are no gears, so they would need adding. I don't see a need for a stand on a bicycle so I would discard that. The frame is a strange shape and very heavy to take the weight and size of the lead-acid batteries - so I wouldn't even want to keep that. That leaves the front wheel and the rack as original. It reminds me of the story of Trigger's broom or perhaps Victor Meldrew after a depressing visit to the garage saying that he'd been advised to keep the nearside wing mirror and replace the rest

 

[ChrisOfBristol]

For my sensor, I get >20M for red probe on red, black probe on black, 50k for red probe on black, black probe on red. For red probe on green and black probe on black I get 64k and when the red probe is on the black and black probe is on the green I get 132k.

Thanks very much for getting this out and testing it. I'll put my multimeter on a higher setting and try again. Doesn't look promising though.

 

[ChrisOfBristol]

I've spent a while fetching the power semiconductors out and testing them and they both pass the tests described on this useful thread: semiconductor test (Note that my posts there refer to a different electric bicycle.)

For my sensor, I get >20M for red probe on red, black probe on black, 50k for red probe on black, black probe on red. For red probe on green and black probe on black I get 64k and when the red probe is on the black and black probe is on the green I get 132k.

I'm assuming that you used the diode setting on your multimeter.
For my sensor, I get NC for red probe on red, black probe on black, 517R for red probe on black, black probe on red. For red probe on green and black probe on black I get 1732R and when the red probe is on the black and black probe is on the green I get NC.
In addition for red probe on red, black probe on green I get NC, for red probe on green, black probe on red I get a short flash of 1500R but that varies. These values are quite different, but the conducting connections are the same, so it's difficult to draw any real conclusions.

I am inclined to assume that it's had it and that there is no way of getting another sensor and look at:

When you operate the throttle, there's an AND condition that the sensor is pulsing. The 555 would switch on and pulse as soon as you switch on the controller and remain pulsing until you switch off, in which case the AND condition is always met. I'm sure that's the same as how the PTDISABLE device works.

I reckon it should be possible to short circuit all this and insert the appropriate 0 or 1 signal to tell the power to switch on. I think it might be worth getting a wire with a resistor in it, connecting it to +12V and poke around to see if I can find that spot.

 

[ChrisOfBristol]

Eventually...
Having tested everything else including the power transistors, I found that the sensor resistances were wrong. I was fairly convinced that it was the cause of the problem, and the next step would have been to replace it with a Honeywell SS41. However I needed the space back and was rather fed-up with the thing, so I have now sold it as is, with the offer to help fix it if needed.
The help I received on here was invaluable - thanks @vfr400 @WheezyRider.

 

[ChrisOfBristol]

There is a variety of very reasonable priced packages available which include a controller and a throttle. For anyone with similar problems with this electric bike, one of these is probably the best answer - they don't need a speed sensor and they avoid the problem of the Powabyke throttle electronics working backwards. The brake cutout could just be disconnected as it isn't a legal requirement. If needed, the switch in the brake could be replaced by a push-for-on type.