For CA HC: can single phase be used instead of hall magnet?

So I'm considering adding a CA-HC to my bike's mix. But one question eludes me: can I use a single phase from the controller instead of the supplied hall magnet speed sensor? In theory is "should" work as that is what the standard 6-pin CA-DP uses. Or should I just purchase a CA-DP and hack the 6-pin adapter to work with an external shunt? (basically, building my own CA-HC)

Any advise appreciated.
Michael

MJSfoto1956 said:

So I'm considering adding a CA-HC to my bike's mix. But one question eludes me: can I use a single phase from the controller instead of the supplied hall magnet speed sensor? In theory is "should" work as that is what the standard 6-pin CA-DP uses. Or should I just purchase a CA-DP and hack the 6-pin adapter to work with an external shunt? (basically, building my own CA-HC)

Click to expand...

Unless there's something unusual about the HC version I'm not aware of, you can use any hall sensor signal from the motor in place of the wheel speed sensor.

But you can't use a phase from the controller, because the high voltage from it will blow up the CA's speedo input, and probably the MCU inside it.


However....since you've already added a CA-DP connector to your controller, why do you want the HC version with it's own external shunt? You can simply use the regular CA-DP to do this, setting it to High Range for the 0.35mohm value, and put that value into the CA setup.

Then in your DP connector wiring in the controller, hook the speedo wire (green?) to whichever hall signal input you want to use it on. (doesn't matter which).

(jsut be aware that there are some bugs in the High Range shunt mode that aren't in the Low Range mode, like throttle ramping times being very slow; they'll affect any form of the CA3 that uses a shunt smaller than 0.75mohm, as that requires High Range mode. They'll get fixed at some point, I'm sure...but until then...).




Side note: One reason I don't like using the speedo signal (or the throttle signal) in the DP connector is becuase that connector also carries batteyr voltage. If anything happens that causes the battery voltage to short to the low-voltage signals, it will fry the CA. (and whatever else is connected to the other end of the low voltage signal). That happened to me, when the wiring sheath on the CA-SA shunt split, and then the wires inside the sheath, right at the junction into the molded shunt, were twisted enough that when insulation failed on the battery positive and the speedo wire, it fried the hall in the motor, and the CA itself. I was lucky that the controller did not fail from this, as that's also on the same hall signal wire.

Additionally, the same thing can happen if you have a failure in the motor-to-controller cabling or connections.

Wouldn't be a problem if the CA had hardware protection against this, but it doesn't, and neither do the halls in the motor, or potentially the controller input.

amberwolf said:

...However....since you've already added a CA-DP connector to your controller, why do you want the HC version with it's own external shunt? You can simply use the regular CA-DP to do this, setting it to High Range for the 0.35mohm value, and put that value into the CA setup. Then in your DP connector wiring in the controller, hook the speedo wire (green?) to whichever hall signal input you want to use it on. (doesn't matter which)...

Click to expand...

Oh, it was merely a hypothetical exercise. Your explanation was perfect -- thanks!

M

I'm sure you could make a simple interface circuit that would take a phase wire signal and make it compatible with the CA speed input. For sensorless motors this might be desirable.

Some sensorless motors have a hall sensor for speed built in, and this could be used (it's designed for it).

Some sensorless controllers have a board that translates the phase signals into hall signals, mounted at the PCB where the halls signals would've gone in there. Most of them are cheap "generic" controllers. I've seen several LiShui (Li Shui, LS) controllers with this (I have at least one or two), and I think some of the Lyen sensorless-capable controllers had this board too, as did other "infineon" sensorless controllers of the time.

Nowadays that stuff is probably built into the main board, or even just done in software inside the MCU from voltage-divided phase signals or similar.

But if you have one of those old controllers laying around, you could pull the board and reverse engineer it to make your own.

amberwolf said:

Some sensorless motors have a hall sensor for speed built in, and this could be used (it's designed for it).

Some sensorless controllers have a board that translates the phase signals into hall signals, mounted at the PCB where the halls signals would've gone in there. Most of them are cheap "generic" controllers. I've seen several LiShui (Li Shui, LS) controllers with this (I have at least one or two), and I think some of the Lyen sensorless-capable controllers had this board too, as did other "infineon" sensorless controllers of the time.

Nowadays that stuff is probably built into the main board, or even just done in software inside the MCU from voltage-divided phase signals or similar.

But if you have one of those old controllers laying around, you could pull the board and reverse engineer it to make your own.

Click to expand...

I have a CAv2.3 and use the yellow speed sensor wire connected to a hall sensor wire going into the motor but it only reads accurate when i release the throttle. Is there a known or possible reason for this. I use a HC shunt. Here is the set up. It's a mid-drive motor on an LMX bike.

brumbrum said:

I have a CAv2.3 and use the yellow speed sensor wire connected to a hall sensor wire going into the motor but it only reads accurate when i release the throttle. Is there a known or possible reason for this. I use a HC shunt. Here is the set up. It's a mid-drive motor on an LMX bike.

Click to expand...

A middrive motor cannot give you the wheel speed (bike speed), unless the gear ratio from motor to wheel is always 1:1 (or you can use the ratio you do have to get a number of poles that the CA will allow in setup, that will give the right speed), and there are no freewheels in the system.

So, as long as your bike has no freewheels between the motor and the wheel, meanign the motor is always backdriven by the wheel, just like a DD hubmotor in the wheel would be, then it can be used to measure speed, as long as you can get the right number to put in the CA for #poles to make it come out right for your gearing.

Now, to find out why it doesn't work while throttle is active, you'd need to know what's different between the two states, and "how" it's inaccurate.

Meaning, is it always off by the same amount when throttle is active, regardless of motor speed and regardless of throttle amount?

Or is it a variable amount?

If it's variable, you could have a bad ground or a ground loop somewhere, that's causing voltage across a ground that prevents the CA from correctly recognizing all the pulses being given off, so the speed is too low.

Or it could be noise in the signal or in the ground that's causing spikes that cause the CA to see too many pulses, and the speed is too high.

Etc.

amberwolf said:

brumbrum said:

I have a CAv2.3 and use the yellow speed sensor wire connected to a hall sensor wire going into the motor but it only reads accurate when i release the throttle. Is there a known or possible reason for this. I use a HC shunt. Here is the set up. It's a mid-drive motor on an LMX bike.

Click to expand...

A middrive motor cannot give you the wheel speed (bike speed), unless the gear ratio from motor to wheel is always 1:1 (or you can use the ratio you do have to get a number of poles that the CA will allow in setup, that will give the right speed), and there are no freewheels in the system.

So, as long as your bike has no freewheels between the motor and the wheel, meanign the motor is always backdriven by the wheel, just like a DD hubmotor in the wheel would be, then it can be used to measure speed, as long as you can get the right number to put in the CA for #poles to make it come out right for your gearing.

Now, to find out why it doesn't work while throttle is active, you'd need to know what's different between the two states, and "how" it's inaccurate.

Meaning, is it always off by the same amount when throttle is active, regardless of motor speed and regardless of throttle amount?

Or is it a variable amount?

If it's variable, you could have a bad ground or a ground loop somewhere, that's causing voltage across a ground that prevents the CA from correctly recognizing all the pulses being given off, so the speed is too low.

Or it could be noise in the signal or in the ground that's causing spikes that cause the CA to see too many pulses, and the speed is too high.

Etc.

Click to expand...

I think the difference In speed readings is a constant, as far as i can observe (just a guess). So starting here....
Motor has 4 pole pairs.
Gearing is 91/14t (no freewheeling, only direct drive) which is 6.5:1 ratio.
Is there a formula to use to get the correct #pole for CA? Such as 6.5x4=26

EDITED: using 26 as pole# gives a fairly accurate reading on the CA compared to a gps reading; this is when throttle is engaged. However, the speed reading when throttle is released is about 3x higher than the accurate reading. I need to find out if this is a consistent reading difference between the two. So how can i get the two readings aligned?

brumbrum said:

So how can i get the two readings aligned?

Click to expand...

Have to find out what's causing the extra pulses first.

The motor isn't spinning any faster with throttle off than on, right? (shouldn't be possible)

Since the controller wouldn't be sending phase signals to the motor, then it can't be phase-hall interference.

If you have access to an oscilloscope, you can look at the actual speed signal going into the CA, and see what the extra pulses look like (vs the actual signal pulses).

If they're really small, like short spikes, a small capacitor from the speed signal to ground, as close to the CA as possible, might fix it, though you may have to experiment with the value. Could start with 0.1uF.

If they're basically the same as the actual signal pulses, then something else, probably inside the motor, is causing the hall itself to be triggered more often without power than with it.

amberwolf said:

brumbrum said:

So how can i get the two readings aligned?

Click to expand...

Have to find out what's causing the extra pulses first.

The motor isn't spinning any faster with throttle off than on, right? (shouldn't be possible)

Since the controller wouldn't be sending phase signals to the motor, then it can't be phase-hall interference.

If you have access to an oscilloscope, you can look at the actual speed signal going into the CA, and see what the extra pulses look like (vs the actual signal pulses).

If they're really small, like short spikes, a small capacitor from the speed signal to ground, as close to the CA as possible, might fix it, though you may have to experiment with the value. Could start with 0.1uF.

If they're basically the same as the actual signal pulses, then something else, probably inside the motor, is causing the hall itself to be triggered more often without power than with it.

Click to expand...

Unfortunately i have no access to an oscilloscope. The speed reading is most definitely 'exactly' 3x faster when off throttle. The motor does not spin faster off throttle or have any freewheeling capability.
i may have to look at how i have wired the bike together. Could this be causd by a wiring anomaly?

brumbrum said:

Unfortunately i have no access to an oscilloscope. The speed reading is most definitely 'exactly' 3x faster when off throttle. The motor does not spin faster off throttle or have any freewheeling capability.
i may have to look at how i have wired the bike together. Could this be causd by a wiring anomaly?

Click to expand...

Yes, if there is an issue with the ground wires it could make a loop and introduce noise on the signal. It should not change when you are off the throttle. That is a really strange problem and I haven't seen that one before.

fechter said:

brumbrum said:

Unfortunately i have no access to an oscilloscope. The speed reading is most definitely 'exactly' 3x faster when off throttle. The motor does not spin faster off throttle or have any freewheeling capability.
i may have to look at how i have wired the bike together. Could this be causd by a wiring anomaly?

Click to expand...

Yes, if there is an issue with the ground wires it could make a loop and introduce noise on the signal. It should not change when you are off the throttle. That is a really strange problem and I haven't seen that one before.

Click to expand...

Thanks for the reply. Its only since getting the CA #pole dialed in correctly that it has become obvious that the no throttle reading is three times higher.