New Rad Rover 6+ - stutter, hesitate, jerk when applying throttle from stop

[tbrooks123]

Hi, I've searched and don't see any mention of what I am experiencing.  If there is a prior discussion, please point me to it.

When I apply the throttle from a stop, the bike/motor stutters a little bit before it gets rolling.  Tried to take a video of it, but it was to subtle to notice on video.  Not a lot of sound associated with it.  Wondering if this is normal or anyone else has experienced.  Otherwise it rides fine. My only observation is it accelerates slower (smoother)? than the Radmini and the 2018 Rad Rover I previously owned.

Thanks!

[JimL]

I have been working on a similar problem (with RAD's help).  If I am pedaling the bike, and try to add a little throttle, the bike starts bucking and jerking.  If I am in PAS 1 (pedaling) and try to add a small amount of throttle, the Watt meter starts jumping back and forth between 41 and 750.  I was able to send them a video of the condition.

It does not occur when riding PAS only but is a little jerky on "throttle only".

The bike is seriously down on hill climbing power compared to the four "Mini ST" I have bought over the last 2 years.  PAS 1 is so weak it is almost undetectable on level ground in a paved parking lot.

So far, we have tried a new throttle and then new controller.  Today I received a new wire harness which I will install tomorrow.  I will post the results when I get done.

Just a side note:  My son has a lot of experience with high performance permanent magnet motors in very high-level RC racing.  He has run into motors that were mechanically mis-built, where the "timing" of the components is actually clocked wrong.  He said the bucking is called "cogging" and it goes along with reduced power (in relation to the actual watts fed).

[JimL]

Wire harness installed this morning but no change in the condition.  I have let RAD know and will wait to see what comes next.

[JimL]

Rad sent me a new rear wheel, which completely cured the problem.  It is also noticeably stronger climbing the steep street to my house.  PAS 1 now actually has some assist (which was very faint with the problem motor).

The wheel is remarkably easy to change if you take the right side 18mm nut completely off, and pay attention to the lock washer on the left side which goes INSIDE the frame.  You can easily do the job from the sidestand if you have someone to hold the bike (rocked over) and be sure to have the rear gear in 7th position.

I unbolt the rear brake caliper and swing it out of the way, before removing or installing the wheel.  When you put it back in place, don't fully tighten the two mount bolts until you are holding the rear brake lever engaged.  This will make the caliper properly center when you finish tighten the two bolts.

[JimL]

Here it is July and the problem is back.  If I am riding at low speed, very light throttle, and then start pedaling....the bike starts switching back and forth between 750 watts and whatever light throttle wattage I am using.  It is very strong bucking, and actually spins the rear wheel on gravel (when it kicks to max power).  As I noted in May, this motor is definitely stronger than the one I removed.  Hmmmm....??

If I am riding PAS only (no throttle) the bike is fine until I try to add even a tiny bit of throttle.  Then it goes into the same crazy action.

The ONLY part that has not been changed is the PAS sensor and left crank bearing.  It looks like the sensor on Rover 6+ may be reading the ball bearings in the left end of the crank shaft.  Reading bearings has been a common ABS system function in many cars, for about 15 years or more, and so it is not an unproven system.

Because I have the original harness (changed during the long process), I am going to make a "pig tail" with access to all 3 wires.  I will operate the bike with a portable oscilloscope and see what the PAS sensor pattern looks like.  I think I can create a "drag" condition with the bike up on my bike holder (a Park unit).  I will be able to operate the pedal by hand, working against a rolling drag load against the rear tire (so as not to damage the brake).

I am suspicious that the "timing" of the "bucking" conditions is in sequence with a bad ball in that left side bearing.  The balls, in that type bearing, do not pass the sensor at the same speed as pedal crank rotation.  They would be "walking" around the inside of the outer race based on the speed they are driven by the pedal crank shaft (the inner race for those bearings).

If I can get an oscilloscope pattern (and store it) I should be able to calculate the frequency of PAS sensor signals.  Hopefully, I can see if there is a glitch and figure out the next step of repair.  This will take a few days, and I am sure it will be trickier than I want. 

I may try using my Fast Fourier Transform spectrum analyzer to see if it can "hear" a bad bearing and display a frequency.  Unfortunately, my current FFT does not have a magnetic accelerometer or clip-on microphone (which are the real "magic" for bearing diagnosis).

[JimInPT]

I may try using my Fast Fourier Transform spectrum analyzer to see if it can "hear" a bad bearing and display a frequency.  Unfortunately, my current FFT does not have a magnetic accelerometer or clip-on microphone (which are the real "magic" for bearing diagnosis).

I just love seeing overwhelming geekiness dumped onto the heads of problems, LOL.  A hundred years ago in a previous life, I used to strap accelerometers to the housing of gearboxes the size of a VW Beetle to listen to and identify the various bearings and watch for developing faults, recording on very expensive analog tape recorders for later playback in the lab.  Now that tech is far more advanced, fully-digital, and cheap enough to routinely and continuously monitor machinery to predict failure before something really prangs.  Nowadays, it's probably handled by people's phones.

[JimL]

I was able to build a "plug in" test wire harness for the pedal rotate sensor.  I haven't borrowed an o-scope yet, but my Fluke 87 is showing really dramatic voltage fluctuation off of that Hall sensor when I rotate the pedals slow and steady.  Any ideas.....anybody?

Maybe tomorrow I can learn a little more.  My iPad FFT is too crude to see anything cleanly.

I had thought that the pedal rotation sensing was strictly to identify the pedals rotating forward.  It doesn't seem logical that a fluctuating signal would be causing this strange problem, but who knows??

Hey "JiminPT", do you remember those adjustable tuning forks we could hang on the dash to identify the frequency of buzz or sounds....before the high tech tools came along!  And...then we had to do the math to figure out what component could be vibrating at that freq of RPM or driveshaft, or wheel, or whatever!  Luckily, we had those handy "slide rules" to do math quickly.

[JimInPT]

Hey "JiminPT", do you remember those adjustable tuning forks we could hang on the dash to identify the frequency of buzz or sounds....before the high tech tools came along!  And...then we had to do the math to figure out what component could be vibrating at that freq of RPM or driveshaft, or wheel, or whatever!  Luckily, we had those handy "slide rules" to do math quickly.

I believe I've heard of those but never saw one in person.  Resonant frequency rules!  My high school taught us slide rules and my Dad helped get us to the Moon with his.  I remember all sorts of primitive stuff!

[JimL]

Yup....my wife thinks I should have a custom hat made, to read:

"I write cursive and drive a stick shift" (which is true...stick shift Jeep Rubicon).

Latest news on my wifes bike is that the pedal sensor signal is clean and normal.  It really only switches about once or twice per pedal rotation, serving simply as a PAS "turn on" trigger.

The speedo display, however, does actually have some functions related to motor control.  It is the only part we have not tried, and so, RAD is sending me a unit to try.

When I get it installed, I will post the results.

P.S.  Those "dash mount tuning forks" were a GM item.  You simply adjusted the fork stopper up or down, and read the frequency where the fork vibrated best.  During my years in the field, I often travelled with an Ono Sokki CF-250 FFT.   It was a wonderful unit, and could perform raster display.  That function allowed me to "see" the frequency of things that come and go, while driving.  One of the most interesting cases involved a particular "oppressive" hum in the passenger compartment of a new "sport model' sedan. 

The condition only occurred on certain freeways (concrete joints), while driving at certain speeds.  It turned out to be the "pipe column resonance" of the smaller front tires, on that model.  A tire is nothing more than a tube, bent into a circle.  The length of the tube (just like in old church pipe organs) determined the frequency response, fed back into the unibody and using the roof as a big "speaker".  A small change in tire spec, and bonded "weight insulation" in the roof, effected the cure.

We have to give folks like RAD a lot of credit for dealing with all the complex new problems that go along with product development.  It is NOT simple and easy!

[JimL]

Display did not fix it.   We have now changed the complete wire harness, controller, throttle, display, and rear wheel (motor).  The PAS sensor signal is clean (confirmed on an oscilloscope).

If anyone has an idea, here is the symptom:  Ride about 5-6 mph throttle only on flat ground (very little throttle).  As soon as I start to rotate the pedals (but not actually loading the chain), the bike goes to full power.  It hits so hard that it spins the rear tire on gravel!  I have to stop pedaling and hit a brake.

Riding PAS only, the bike works normal and fine UNLESS I turn in a little throttle.  Then it goes weird, again. 

This is not a wear problem.  This bike has less than 16 actual miles on it, in the 2 months we've owned it.  It did, in fact, work right the first time I rode it with the new rear wheel, but the next time it was bad again.

Does anyone know if there is a ground path used through the rear axle/frame contact point?

[JimInPT]

Display did not fix it.   We have now changed the complete wire harness, controller, throttle, display, and rear wheel (motor).

Given all those swaps you've already done, you've done more than your part to attempt a fix.  You ought to push the big red (yellow?) "Lemon" button and demand an entirely new bike or 100% refund, with no hassle about reboxing it or anything.  This bike needs to be dumped on Rad's doorstep by UPS and replaced.  While I'm almost always loath to call the cops, there are "lemon laws" in many jurisdictions to support your case.

How much more hassle would Rad expect you to endure?  You're almost rebuilt the bike already.

[JimL]

To their credit, RAD has already said they can replace the bike, if we can't fix it.  It is just such a strange thing, and I spent my life career fixing difficult problems. 

It seems that we (and me) are missing something simple, here.  There is some link between the PAS sensor circuit and the throttle.  When I made the breakout cable to look at the PAS signal with the o-scope, I discovered that if any of the PAS sensor wires (3 wires) are open circuit, the throttle will NOT work.

That relationship between throttle and PAS has got to be the source of the problem.  However, if the throttle is disconnected, the PAS system still works just fine.  I am going to try learning how to test that Hall sensor (for the PAS).  Maybe something is going on that disturbs the 5 volt circuit that feeds the throttle.  In all this work, the PAS sensor has not been changed (nor have any lights been changed, and I haven't changed the air in tires  .)

[JimInPT]

To their credit, RAD has already said they can replace the bike, if we can't fix it.  It is just such a strange thing, and I spent my life career fixing difficult problems. 

Yeah, the recovering-engineer still buried deep inside can appreciate that.  If you want to take it on as a challenge, hard to fault that!  Good luck; it'll be interesting to hear if you ever find the bug.

[Ddaybc]

Just an opinion here as I have no experience with this kind of issue.
I know you changed the throttle, however I suspect a broken wire inside the insulation. Maybe it's an assembly issue and it showed up twice for you? It looks fine and while it is sitting still or being used in a specific way the connection is 100% but as soon as a certain type of motion is used with the throttle then maybe the throttle cuts in and out? We had this kind of issue in some of the saw mills I used to work at. Very specific floor vibrations caused the wiring to separate inside the insulation. Once that part of the mill was shut down they always tested fine. It drove the millwrights nuts!
I don't know, just guessing. Maybe a capacitor or resistor issue? They're built in batches so sometimes whole batches are faulty?

[JimL]

Thanks for your note.  Rad did send me a complete new wire harness, display, controller, motor (wheel assembly) and nothing changed.

I have confirmed that riding about 5 mph "throttle only", with PAS set to "0", the bike still surges full power as soon as I rotate the pedals.  If I disconnect the PAS sensor, I can use throttle while pedaling and there is no unexpected result.

Of course, pedaling with no throttle and no PAS is pedaling a very heavy bike.

The only piece of wiring that has not been changed is the short PAS sensor section (which I have tested with an oscilloscope while pedaling on my test stand), and the push button controller on the left side of the handlebars.  I already tried disconnecting the lights and saw no improvement.

RAD is looking at the next step and I will post if I learn more.

P.S.  I did ask them about the new paint finish (possible frame ground problem?), which probably sounds dumb...but about the paint;  Over the years I had several cases of our high perfomance turbo model where customers blew them up and needed a new "short block" assembly.  After installation, the cars had repeated "check engine light" problems indicating the Knock Sensor was not working.  It turned out that when the factory makes "replacement short block" assemblies, they double paint them to prevent corrosion in possible years of storage in a parts warehouse.  The extra paint, in the thread hole for the knock sensor, was insulating the knock sensor from responding to the 8.8 kHz frequency that occurs in that particular bore diameter during a "knock" event.

So....Paint too thick on the new Rover 6+?  Probably not the issue, but my personal history makes me think about it.   At this point, be glad you are not one of the nice folks at RAD who have to work with someone like me.