Throttle Voltage Query

What would cause a thumb throttle to exhibit 4V over its' full range?
I have two which I've tried and they both present a similar result (one with 4.32V one with 4.54V).
I've made a 'break out' cable so that I can test each throttle in situ with the Bafang powered up. Using this cable I've recorded
4.83V (Red) and 4.93V (White) which looks to me like the Hall signal is higher than the power supply.

My situation is (from a previously unresolved post) that of a WOT response to any use of the throttle.
A bench 'off bike' test using a 5V supply and resistor legs as probes into the Higo connector verify my initial findings.

Should I replace the sensor FET in each throttle to see if I can achieve the required less than 1V reading which I should be getting when the throttle is not activated? I've pored over Tommy Cat's excellent throttle dissertation as well as a few others but have not found anything describing the fun voltage exhibited throughout the whole range such as in this case.

The ground is intact between the rear wheel sensor (breakout cable) and other parts of the Bafang motor wiring loom e.g. the ground of the e-brake connection. If it's not a throttle issue, then it looks like I'll have to consider the $200AUD Bafang controller route.

The PAS still works perfectly but my physical disability necessitates throttle use to gain sufficient momentum in my velomobile to get away from the cars.

Hi Reflector,

I would be very suspicious of your actual hall throttle wiring, And recommend opening it up, and verifying the correct wires to sensor pin-outs (5vdc,0vdc, and Output)as seen below...






NOTE: The wire COLORS referenced on the illustration are just typical colors used for the discussion in the thread. They can be ANY color.

Of course whilst your in there, verify that the magnet(s) and hall sensor are in there correct positions. Then bench test again with your own 5vdc (or close) voltage supply. Looking for that glorious ~1 to ~4vdc linear output.

If all's well with the throttle, verify the controller's 5vdc, 0vdc and output wiring.


Regards,
T.C.

Reflector said:

What would cause a thumb throttle to exhibit 4V over its' full range?

Click to expand...

Miswiring can do that, including damage to the hall sensor from a wiring problem.

Broken ground could do that.

Sometimes a magnet broken off inside and stuck in the wrong place can do that.

4.83V (Red) and 4.93V (White) which looks to me like the Hall signal is higher than the power supply.

Click to expand...

Since that's not possible, something else must be going on (perhaps a problem making connection to the Red with the meter). Just to be certain, is that measured while connected to the controller, or with the throttle completely disconnected from the bike?

A normal throttle signal doesn't go that high, and typically varies from around 1v to around 4v. In my expereince so far, as low as around 0.8v and as high as around 4.3v.

My situation is (from a previously unresolved post) that of a WOT response to any use of the throttle.

Click to expand...

Does this mean WOT when a throttle is connected at all?

Or only when the throttle is engaged at all, not in the no-rotation position?

FWIW, there are some Bafang systems (BBS02B, IIRC) that simpy won't respond to throttle normally, as the firmware was changed from the original so that throttle is just basically on/off. But when I looked at your most recent thread before this one,
https://endless-sphere.com/forums/viewtopic.php?f=2&t=106063
it says your problem is not just that you get a WOT response to any use of the throttle, but that it is stuck on and wont' disengage once moved past a certain point, which is a different problem, and different from your description in this thread as well.

Since it occurs with two different throttles, (which I would guess are mechanicaly different, and look different, as well), then it is unlikely to be caused by the throttle itself, as it would have to be something like a magnet broken off inside and moving with the throttle mechanism until it reaches some point, then sticks there when you reduce throttle. However, then it would not reset to zero and work again afterward, it would just be stuck at that point so next time you powered on the system, it would stay there, and most likely the controller would detect a stuck-on throttle, and either error out or disable the throttle, if it has any smarts to it at all.

But it is unlikely both of them would happen to be broken in exactly the same way. However, if it's verified that the throttle signal output, when not connected to the controller, is definitely stuck at a single voltage regardless of rotation of throttle, its' easy to check for, by opening one up. If nothing is physically moved or broken inside, it couldn't be a mechanical malfunction in the throttle causing it. It could be electrical, like a broken ground anywhere from the hall sensor itself all the way to the connector.

Should I replace the sensor FET

Click to expand...

For ease of locating a replacment if you need to, it's not listed as a FET, but a hall effect sensor. There are many variants, but a common replacement used is the Allegro A1301 or A1302;
https://www.jameco.com/z/A1302KUA-T-Allegro-Microsystems-5-Volt-10-mA-Hall-Effect-Sensor_2135881.html
there are others that would work too, like the Honeywell SS495
https://www.jameco.com/shop/ProductDisplay?catalogId=10001&langId=-1&storeId=10001&productId=1915923
. However, getting the throttle apart without breaking anything is often tough, and getting it back together with the spring just right is harder. If a magnet comes out, it has to be put back in exactly the same way it was originally, or you won't get the right voltage range out of the throttle and it wont' work on the controller. It's typically easier to replace the entire throttle (and often cheaper, depending on the cost of shipping to get the tiny hall sensors, which themselves are really cheap).

The ground is intact between the rear wheel sensor (breakout cable) and other parts of the Bafang motor wiring loom e.g. the ground of the e-brake connection.

Click to expand...

Not sure what this means. Is the throttle connected to one of those two connections? Or are you simply trying to express that the ground is intact at every single connector (including the throttle)? Since you don't mention that it is intact on the *throttle* connector, which is the one that's important for your problem, it is not clear what this statement is intended to convey with reference to your issue.

TommyCat said:

Hi Reflector,

I would be very suspicious of your actual hall throttle wiring, And recommend opening it up, and verifying the correct wires to sensor pin-outs (5vdc,0vdc, and Output)as seen below...

Click to expand...

As per the diagram above.

NOTE: The wire COLORS referenced on the illustration are just typical colors used for the discussion in the thread. They can be ANY color.

Of course whilst your in there, verify that the magnet(s) and hall sensor are in there correct positions. Then bench test again with your own 5vdc (or close) voltage supply. Looking for that glorious ~1 to ~4vdc linear output.

Click to expand...

Both magnets are in situ.


4.99V supply from surplus cigarette plug charger.
Using this method, voltage (Hall to Gnd) remains stable at more than 4V throughout the complete range of the throttle; both throttles tested in this manner.

If all's well with the throttle, verify the controller's 5vdc, 0vdc and output wiring.


Regards,
T.C.

Click to expand...

Thanks Amberwolf

amberwolf said:

Reflector said:

What would cause a thumb throttle to exhibit 4V over its' full range?

Click to expand...

Miswiring can do that, including damage to the hall sensor from a wiring problem.

Broken ground could do that.

Sometimes a magnet broken off inside and stuck in the wrong place can do that.

4.83V (Red) and 4.93V (White) which looks to me like the Hall signal is higher than the power supply.

Click to expand...

Since that's not possible, something else must be going on (perhaps a problem making connection to the Red with the meter). Just to be certain, is that measured while connected to the controller, or with the throttle completely disconnected from the bike?

A normal throttle signal doesn't go that high, and typically varies from around 1v to around 4v. In my expereince so far, as low as around 0.8v and as high as around 4.3v.

My situation is (from a previously unresolved post) that of a WOT response to any use of the throttle.

Click to expand...

Does this mean WOT when a throttle is connected at all?

Or only when the throttle is engaged at all, not in the no-rotation position?

Click to expand...

Throttle can be connected, system powered up etc. and all is well until I apply a minuscule movement of the throttle whereupon I am presented with WOT (only to be resolved by e-brake application). The throttle movement / vehicle speed required to induce this varies; you never know when it will hit WOT...but you know it will happen!

FWIW, there are some Bafang systems (BBS02B, IIRC) that simpy won't respond to throttle normally, as the firmware was changed from the original so that throttle is just basically on/off. But when I looked at your most recent thread before this one,
https://endless-sphere.com/forums/viewtopic.php?f=2&t=106063
it says your problem is not just that you get a WOT response to any use of the throttle, but that it is stuck on and wont' disengage once moved past a certain point, which is a different problem, and different from your description in this thread as well.

Click to expand...

Mine is a BBS02A and you are correct, the physical thumb throttle returns to its' start position but the WOT remains. The throttle (throttle no. 1) which I dismantled is the original crap plastic-on-plastic one issued with my motor. Throttle no.2 is the one which I bought to better fit on the 'bars and it has a much slicker action.

Since it occurs with two different throttles, (which I would guess are mechanicaly different, and look different, as well), then it is unlikely to be caused by the throttle itself, as it would have to be something like a magnet broken off inside and moving with the throttle mechanism until it reaches some point, then sticks there when you reduce throttle. However, then it would not reset to zero and work again afterward, it would just be stuck at that point so next time you powered on the system, it would stay there, and most likely the controller would detect a stuck-on throttle, and either error out or disable the throttle, if it has any smarts to it at all.

But it is unlikely both of them would happen to be broken in exactly the same way. However, if it's verified that the throttle signal output, when not connected to the controller, is definitely stuck at a single voltage regardless of rotation of throttle, its' easy to check for, by opening one up.

Click to expand...

Yep, I have already done a bench test as you suggested and verified that voltage stays at a constant value regardless of throttle rotation.

If nothing is physically moved or broken inside, it couldn't be a mechanical malfunction in the throttle causing it. It could be electrical, like a broken ground anywhere from the hall sensor itself all the way to the connector.

Click to expand...

It looks like that's where I'm at now; verifying ground continuity. Obviously I suppose I check from the controller back to the throttle and thence to the wheel sensor?
The red LED flashes as the rear wheel causes the spoke magnetic to pass the sensor.
Also, the hall effect sensor removed from throttle no. 1 gives the following readings;
signal / ground: 0.670V
power / ground: 0.214V
signal / power : 0.860V

However, getting the throttle apart without breaking anything is often tough, and getting it back together with the spring just right is harder. If a magnet comes out, it has to be put back in exactly the same way it was originally, or you won't get the right voltage range out of the throttle and it wont' work on the controller. It's typically easier to replace the entire throttle (and often cheaper, depending on the cost of shipping to get the tiny hall sensors, which themselves are really cheap).

Click to expand...

I can get the hall sensor for less than $10 which is less than half what a new throttle + postage would cost. If it is a sensor problem, I'll make up a 3 pronged drift out of plastic conduit so that I can get all 3 plastic tabs out simultaneously...or buy a 3rd throttle.

Is the throttle connected to one of those two connections? Or are you simply trying to express that the ground is intact at every single connector (including the throttle)? Since you don't mention that it is intact on the *throttle* connector, which is the one that's important for your problem, it is not clear what this statement is intended to convey with reference to your issue.

Click to expand...

The 'breakout cable' which I made fits between the controller and rear wheel pickup. I used it so I could check that the signal from the spoke sensor was working and so that I could verify ground continuity for that function. This being proved, what I haven't done yet is make a similar device so that I can test the throttle in the same way. It seems a waste to ruin a perfectly good spare Higo cable, but if you recommend it as a useful tool to enable in situ testing of components, I'll get into it.

TommyCat said:

Of course whilst your in there, verify that the magnet(s) and hall sensor are in there correct positions. Then bench test again with your own 5vdc (or close) voltage supply. Looking for that glorious ~1 to ~4vdc linear output.

Click to expand...

Sorry guys, I rediscovered a diagram which showed the throttle pins and on rereading it I noticed it was for the male pins from controller NOT the female ones of the throttle itself as I had it in the previous photo.

After reversing the probes on throttle no. 2 ...

Now I get a low reading of 0.83V and a high of 3.92V
Okay, so now we know that it isn't the throttle but an electrical connection (or controller)
Any suggestions on how deep to start this trace?

Reflector said:

Throttle can be connected, system powered up etc. and all is well until I apply a minuscule movement of the throttle whereupon I am presented with WOT (only to be resolved by e-brake application). The throttle movement / vehicle speed required to induce this varies; you never know when it will hit WOT...but you know it will happen!

Click to expand...

That's just plain wierd. I've just never run across a throttle that has a stuck voltage *except* when a magnet is physically broken loose inside, or the ground is broken. In the latter case...it's always WOT even without touching it, in the former case, it depends on where the magnet is for what happens.

But none of them ever got stuck the way yours both do, at the exact same voltage for the full movement range after just the tiniest turn-on movement, then not returning to "off" voltage after letting go of the throttle.

And having the identical problem happen to TWO completely different throttles seems just a little strange...it certainly could happen....but it seems like something else might be causing it to happen.

Only thing I can think off offhand is that if there is a battery voltage level wire running in the same cable as the throttle, like for a throttle-mounted voltmeter, perhaps it got wet and shorted to the throttle signal, damaging the hall in it. Or similarly, inside the controller at the PCB?

It looks like that's where I'm at now; verifying ground continuity. Obviously I suppose I check from the controller back to the throttle and thence to the wheel sensor?

Click to expand...

See the bottom of the post for my *possible* understand of what the wheel sensor has to do with this (because under normal circumstances, it doesn't).

The ground for the throttle only matters between the hall sensor inside the throttle itself, to the controller PCB inside the housing wherever that is located. If you're accessing that ground thru the wheel sensor or other connection, such that you are *also* testing thru the controller's internal grounds, then that's fine. As long as the test *includes* a connection to the controller PCB, via whatever route that happens. FWIW, the battery negative input of the controller would suffice for this, since you know the controller has power, it must have a good ground to that input wire.


However...the ground from the throttle cable connector on the controller to the controller itself is irrelevant to the bench test you've done on the throttle. Only the ground from the throttle connector to the hall sensor inside the throttle matters for that test.


Note that for a ground test on the bike itself, you would disconnect the battery, and use a resistance measurement at the lowest ohm setting on your multimeter. It should be in the single digit ohms, or less. Same for readings on a test of the throttle's own cable.

Also, the hall effect sensor removed from throttle no. 1 gives the following readings;
signal / ground: 0.670V
power / ground: 0.214V
signal / power : 0.860V

Click to expand...

Under what test conditions does it give those readings? If you are applying 5v and ground to it's power and ground inputs, then if you get any of the voltages you list, then neither of them is properly connected to the 5v and ground inputs.

If you're using some other test method, you must specify exactly what you are doing, which test lead goes where for each reading, and which mode your test equipment is set to. Otherwise, the readings you give indicate what seems like a complete wiring failure, which I doubt is the case.

Is the throttle connected to one of those two connections? Or are you simply trying to express that the ground is intact at every single connector (including the throttle)? Since you don't mention that it is intact on the *throttle* connector, which is the one that's important for your problem, it is not clear what this statement is intended to convey with reference to your issue.

Click to expand...

The 'breakout cable' which I made fits between the controller and rear wheel pickup. I used it so I could check that the signal from the spoke sensor was working and so that I could verify ground continuity for that function. This being proved, what I haven't done yet is make a similar device so that I can test the throttle in the same way. It seems a waste to ruin a perfectly good spare Higo cable, but if you recommend it as a useful tool to enable in situ testing of components, I'll get into it.

Click to expand...

I think I see what you're getting at--that the breakout lets you access the ground in the system without accessing the controller itself?

If so, then using that breakout ground connection to verify all the way up to the sensor inside the throttle should be sufficient testing of the throttle ground.

Reflector said:

Sorry guys, I rediscovered a diagram which showed the throttle pins and on rereading it I noticed it was for the male pins from controller NOT the female ones of the throttle itself as I had it in the previous photo.
Bafang Throttle Pins copy.jpg
After reversing the probes on throttle no. 2 ...
probes corrected.jpg
Now I get a low reading of 0.83V and a high of 3.92V
Okay, so now we know that it isn't the throttle but an electrical connection (or controller)
Any suggestions on how deep to start this trace?

Click to expand...

Ok, now THAT sounds like a fully functioning throttle.

If you get that on both of the throttles that show the WOT problem, then the ground test from controller-side throttle connector to the controller PCB ground (or battery negative), with battery disconnected, using lowest ohm setting on meter, is the next test I'd do.

If the ground is fine, then I'd find a place to access the throttle voltage *at the controller PCB*, or at the least at the internal connector / wiring to it, if it's potted and inaccessible. That's probably inside your BBS02, and I don't know where it is, but I think it's referenced in this thread:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=93977
What is possible is that if water has ever intruded in there, it could've caused corrosion on contacts preventing the throttle signal from correctly reaching the controller, or even caused a connection between the throttle signal and something else. (though I cna't imagine what would cuase the behavior you see).

The motor, although bought new in 2014 has less than 50km on it (all under cover of the velomobile shell and never in the rain) so there is not a water problem.
I get 4.94V on the 5V power from the controller to the throttle.
The throttle is either ON or OFF, no gradual increase at all which doesn't tally with the bench test result. After the throttle sticks wide open and I then use the e-brake to kill the motor, sometimes the throttle refuses to work again until after I reboot the display.

I just had a flashback of my experience with a 2 stroke RD350 Yamaha back in the 1970s. On very cold mornings both carburettor slides would go WOT at the merest movement of the twist grip. Easily fixed by screwing weights into the tops of both slides to counteract the low pressure which was sucking the slides up. In that case of course the twist grip didn't return to the starting position as my current one does.
Could there be a momentum or friction issue due to the chains (2), tensioners (2) idler pulley (1) and chain run length be causing this problem (i.e. 'motor run-on')?

I'd try to get an accurate throttle sensor output reading as close to the controller as possible to eliminate ether a shorting of power to the output wire, ( +5vdc) or an intermittent ground. Both causes of spiking the output. Perhaps this will help...




And wanted to say what an awesome looking nosecone! But I don't think the propeller is near big enough... :lol:

Thanks for your advice TC.
So I need to read across the blue throttle sensor cable and ground (at the controller) as the thumb throttle was moving through its' range?
Ah, the propellor...
This is what I wanted
This is what I made

Yes, but I did think of perhaps an easier test... What happens when you pull the plug of the throttle connector when it sticks in high?



What, and I thought that was store bought! Nice. Lol, was also reminding me of a Van's RV aluminum type aircraft. What type is that?

I'll try disconnecting the thumb throttle when it goes WOT. I hadn't thought of that.
Van's RV-7 yep, that's the one. I used Everbrite 'Protecta clear' to try and maintain the finish (just for safety's sake :wink: ) I have no idea what they use for their mirror finish. The semi-monocoque body on my machine is only 0.6mm thick so I won't be polishing it much.
The machine? i bought it as a flat-pack Alleweder A4 produced by Alligt in The Netherlands from the mid 80s and ceased production in December 2019; now Alligt produce composite models.
I learned a lot about hand bending aluminium tubes in more than one plane and also discovered Clecos (the fly boys call them 'skin pins') and how much effort is expended in applying 1,500 rivets by hand.

I'm running two dolphin batteries using an NKK S823D to switch between them. The accessories (LED lights, hazards, thermostat, fan etc. all run off a separate 12V system. I've done the temperature sensor install and the thermostat (that you see at top of picture) works as it should. My current throttle testing is being conducted with the rear wheel on rollers and the front wheels strapped to ramps. Due to the width of the BBS02 I had to install a freewheel jackshaft and fit a GrinTech PAS sensor to it. I might get around to creating a build log when I find out where to submit it on this site.

Is it normal for the BBSXX to not have graduated throttle power (i.e. only a choice of 'OFF' or 'ON')?

Looking forward to your build log.

Are your different voltage supply grounds kept separate and not using a common ground or frame?

I applaud your testing set up.


"Is it normal for the BBSXX to not have graduated throttle power (i.e. only a choice of 'OFF' or 'ON')?"

I have seen this mentioned in several threads, especially after Bafang messes with their firmware. With some resolved by setting the parameters correctly in the programming. But yours seems different with the fact of it "sticking" on with release of your throttle...

Polishing to a mirror finish... I Like it.
https://vansairforce.net/community/showthread.php?t=181607&highlight=polishing

That's what a pneumatic riveter's for... :wink:

Try using the throttle without pedaling at all and see if it behaves smoothly. This will tell you if it's just the crappy Bafang software issue.

fechter said:

Try using the throttle without pedaling at all and see if it behaves smoothly. This will tell you if it's just the crappy Bafang software issue.

Click to expand...

Hi Fechter,
I'm managing the throttle testing without pedalling. The velomobile is on ramps (front wheels) and rollers (rear wheel) while I crouch beside it. This way I remove any PAS involvement and it gives me room to plug and unplug cables etc.

TommyCat said:

Yes, but I did think of perhaps an easier test... What happens when you pull the plug of the throttle connector when it sticks in high?

Click to expand...

Just disconnected the thumb throttle connection while WOT...motor kept churning away.
I'm dreading the fact that it is a controller issue; but everything seems to be pointing that way doesn't it.

TommyCat said:

Looking forward to your build log.

Click to expand...

Here's a picture to go on with (oh how I love the beauty of CNC'd alloy).
The swing arm has two enclosed suspension springs (see top left of picture). Also a braced aperture for the U lock to go through.

Are your different voltage supply grounds kept separate and not using a common ground or frame?

Click to expand...

Yes, the 12V and 48V systems share nothing in common.
I designed the 12V system so all functions were doubled wired (each carried their own ground; all grounds terminating at the 12V battery terminal). Both systems are isolated from the body / chassis. I presume this is how ships are wired. I used the 'Western Union Splice' / 'Lineman splice' where connectors weren't a viable option as I thought if it was good enough to get a man on The Moon...

I applaud your testing set up.

Click to expand...

Thanks.

That's what a pneumatic riveter's for... :wink:

Click to expand...

No compressor available but I agree with you.

Reflector said:

TommyCat said:

Yes, but I did think of perhaps an easier test... What happens when you pull the plug of the throttle connector when it sticks in high?

Click to expand...

Just disconnected the thumb throttle connection while WOT...motor kept churning away.
I'm dreading the fact that it is a controller issue; but everything seems to be pointing that way doesn't it.

Click to expand...

If the pedals were moving even though you weren't pushing on them, it could make it stay on. Pedaling backward should make it stop immediately.

I suppose it's possible there is a short in the main cable harness that goes from the throttle to the controller.

Reflector said:

Just disconnected the thumb throttle connection while WOT...motor kept churning away.
I'm dreading the fact that it is a controller issue; but everything seems to be pointing that way doesn't it.

Click to expand...

Not unless some one has seen this issue caused by software...

Last ditch effort would be to slice open the 8 wire cable off the controller, (gets replaced anyway) cutting and tapping into the Blue wire throttle signal wire to see if voltage is being back-fed from an internal short in the controller. Eliminating a short possibility farther down the harness toward the throttle.

(EDIT: After first removing or isolating any extra converters and/or electrical add ons.)



Awesome swing arm, and the remove before flight tag is a nice touch... :thumb:

fechter said:

If the pedals were moving even though you weren't pushing on them, it could make it stay on. Pedaling backward should make it stop immediately.

I suppose it's possible there is a short in the main cable harness that goes from the throttle to the controller.

Click to expand...

First Prize goes to Fechter!
These be my preliminary findings during 10mins of testing at dusk.
As soon as I held one of the pedals, (thereby stopping the sympathetic rotation of the freewheel) I was able to hold the thumb throttle wide open then release my thumb and the motor stopped as it should. I found that if I then released my grip on the pedal, stuck WOT reappeared until I again stopped the pedal.
As Prof. Julius Sumner Miller said "Why is it so?"
Is this some sort of 'cruise control' which happens if you pedal while activating the throttle? Conversely, does it also mean that the throttle should only be used without pedalling for this not to occur?

Thankyou gentlemen for your brainstorming on this.

For some reason Bafang decided the PAS should always have priority over the throttle. Anytime the pedals are moving, the throttle will behave "all or nothing" which is very dangerous on a steep mountain trail. As soon as the pedals stop, the throttle behaves like it should. Some earlier versions of firmware didn't seem to have this issue. When riding, you can pedal backward to immediately stop the motor.

Here is the hardware fix:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=93977
this will give the throttle priority over the PAS. As soon as you hit the throttle, the PAS is disabled and the throttle behaves like a throttle is supposed to.

fechter said:

Here is the hardware fix:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=93977
this will give the throttle priority over the PAS. As soon as you hit the throttle, the PAS is disabled and the throttle behaves like a throttle is supposed to.

Click to expand...

Brilliant! Just what I was looking for. Throttle having priority over PAS means not having throttle operation interrupted by any unintentional pedal movement; just what I need. Using 'Karl's special sauce' settings I can achieve a throttle which can operate gradually. On my next trip to the nearest city I'll be investing in some transistors for your fix.
Thanks again.