New fork and MP3 HIL testing

[Sam.Vanratt]

Hi folks
I'll receive my new custom fork in a few days. If this fork will fit to the specification (200Nm constant torque in the dropout, deep dropout, 26mm surface for the nut) I'll publish it for EU drivers of the MP1/2/3 as a solution. Beforehand I have to test it (couldn't afford a company to do the tests) in a HIL (Hardware-In-the-Loop). This one will simulate start/stop, recuperation, climb and downhill scenarious for a few thousand times a day. Even when I measure current, voltage, temperature as well I'm much interested if the dropout and the fork crown will keep to it's spec not to open a 1/20 of a mm in the tests (as I planned it).
Now my scenario needs the test hardware which I'll programm with a simple relay board to deliver signal (on/off/short/open/x Volt)
What I need to know is: what does the MP3 need as a minimum to start/stop/recuperate and which pins I need to contact with what signal. I've the opportunity to disassemble the last cableset I have but I fear to damage it (which I need in the end to finish my second MP3 bike).
I guess the assemblers (like Gary) here knew best about such infos?!
Cheers
SAm

[Bikemad]

Hi Sam, the easiest way of operating the motor would be to connect the relay's switch contacts in place of the wire loop shown below:



Here's a diagram showing the rest of the wiring should you want to utilise alternative connections:



If you connect via the front harness, I suggest you double check the voltages on the throttle connector to ensure the connections are still as shown, as Gary once discovered that there were some slight variations on the throttle wiring on some models.

I suspect that constantly switching full throttle on and off could be detrimental to the health of the controller, so be warned.

Alan
 

[Sam.Vanratt]

Hello Alan
it's exactly what I'm planning: the MP3 should be started, stopped, recuperting, just like normal biking in a streetsimulation (two wheels in the ground with switchable load up to 8kW). The apparatus is normally used for high speed bikes.
When PAS is enabled didn't I have to simulate signales on the PAS as well? Afaik the throttle should be useless as long as there's no PAS signal, isn't it?

Cheers, merry christmas and a happy new year to you all!
Sam

[Sam.Vanratt]

Hello  Alan and all GM users
took some time but finished now:
the last two weeks I build the HIL constellation. I wired a minimum (brake for recup; throttle at +5V), simulate a PAS (12Hz) and applied the sensors (torsion, torque, time, mileage, current, voltage....) to the computer. When I applied the PSU (Goessen Metrawatt @27V 120A) it took two seconds until +5V is online, I triggered PAS (rectangle 12Hz) so the Full throttle could be active and
guess: nothing happened
Second I rotated the MP3 and expected to have recuperation (as it is active at 50%) so a big resistence (and a return current), but nothing happend as well.
Load is at a few 100mA, so no coil gets active.
Did I completely misunderstand the process or is the brand new (OK it took 9 month to get started) MP3 controller defective?

An unhappy
Sam

[Bikemad]

Hi Sam, it sounds as if your not having a lot of luck.

If you are using a 27V supply, and your controller has not been programmed for 24V operation, there's a good chance it won't work at that voltage level as it is likely to be below the Low Voltage Cut-off setting.

If the voltage setting is correct in the controller you can check whether the controller and motor are working by disconnecting everything apart from the power supply and then try the following throttle test:



Please note: The power must be turned on before the piece of wire is used to join the two sockets.

I am not sure whether operating the throttle with programmed 5V pulses on the PAS signal wire will actually work, as when I first tried to get my PAS to function during a bench test I found it very difficult to make it work at all.

I had the magnet ring mounted in an electric drill so that I could spin it at different speeds in order to see whether the PAS was proportional to the pedal speed or simply a fixed amount regardless of pedal speed.  I tried many things, including turning the magnet ring over, changing the direction of rotation and adjusting the gap between the magnets and the sensor.
I eventually found a combination that would make the motor spin correctly but it was not easy, and I was able to confirm that the motor speed did change in direct relation to the speed of the drill.

I therefore suggest that simply sending pulses along the signal wire may not make the motor run as you might expect it to.

I'm not sure that I understand correctly what you have done with the throttle, but if the signal wire is permanently set at +5V then the motor will not run due to a safety feature of the controller that prevents the motor from running should you accidentally turn on the power while the throttle was activated.

Regarding the Regenerative braking, grounding the brake signal wire should induce a noticeable braking effect if regen is set to 50% percent in the controller, but don't expect to see any regen current at slow wheel speeds as I get virtually nothing below 15mph.

Alan
 

[Sam.Vanratt]

Hello Alen
thanks for the fast reply. The MP3 ist of course programmed for 24V (so 27.0 is the max of LiFePO4 would give under load).
My luck.... well say it's maybe not my luck but low Q Standards at GM? One of my three work flawlessly since arrival (Plug&Play), one was bad quality control (leckage in the case and needed a complete turnover which I discribed) and this is my third one.
What I did NOT do was to apply the shortage (throttle-5V) AFTER PowerOn
As I had no experience in such MP3 testing I'm not aware how the controller works at all. So this is a place where I can start and then "back to the basics" (PAS and brake away)

PAS: well it had to work, and the simulation (done in Matlab) is exactly how the three PAS I own (6 pol and two different 12 pol versions with high active signals) shows on my Oszi when working on a bike. I was warned when buying that PAS only is a bit sturdy and only useful in 250W config. The config allows me a wide range (0.00001Hz up to 1.999MHz) and I tried a few, but as nothing reacted I did not try on that value too much. All three PAS I have give Pulses (one for every pol moving in the right direction) and needs about 2-3 pols until the direction is found (HALL) and the direction/interpretation is found.

Throttle: I did apply +0V or +5V and nothing in between [I just wanted full load and no load and recu braking to test (max for the fork dropout)], so that's another point I will work on.

That's a lot of info and points to work on. THANKS!
I'll report back
Sam

[Sam.Vanratt]

Hi Alen (and all others)
my luck (so far) run out today. All I tried did not show good health: the +5V varies from 4,6 to 3,7V (no load, on certain switch on situations it's 3,7 or 4.6V), one of the expnader cables have (Port 3) no Battery+ contact (which is just a defective cable), but no reaction to the +5V on the throttle Signal and no reaktion on anything on PAS.
OK, not so bad so far, lets reset the box and try the thumb controller=> native must work
No connect to the controller. Last week I did the last checkup and it worked (I even saved the config for the future). Now no connection possible....

Sound like a bad controller, doesn't it?

My luck did really run out when I checked today the company (HN Motor) in Germany which was partly taken back to Netherland, so no cheap sent it back and wait until someone repaired it.

Jeeze
Sam

[Bikemad]

Unfortunately, it does sound like you now have a faulty controller.

It appears that testing those forks is proving to be very difficult indeed.

Alan
 

[Sam.Vanratt]

Hi Alen
OK let's try my luck with GM [Please do not laugh!]. I heard a few good words about it. I'm not sure if it's an alternative to choose an external one. I really waited for the MP3 as the internal controller (and the preliminary tests of the dynamic PAS implementation) sounded tempting 18 month ago...
Do anybody use an MP3/2 (the 1 seems much different than the differ of 2 to 3) with an external good quality controller.
As mentioned PAS is important (less external sensors or gadgets) and solid design (good dimensioned FETs, Diodes, Caps, ....).

Any hints
Cheers (from RMA subject in waiting and fellow)
Sam