Dyno'ing the VX-1

Hi guys

Figured we could do with some real numbers. So i am going to have a chat with the guys at my work (harley dealership) who do motorcycle dyno'ing and pull some favours and get my vx-1 dyno'd.

I actually don't know if they can get torque figures because if i recall correctly the torque calculations are dependant upon the ignition pulses (and of course there are none on a vx-1) but i have literally just emailed the dyno manufacturers to find out if its possible.

Everyone here will know as soon as i know.

Should be interesting which-ever-way

Cheers

Chris

I like the way you think!

The ususal two-wheeler dynos use ignition pulses to get engine RPM info. However, these dynos also record drum rpm and thus rear wheel speed. That can easily be converted to Vectrix motor RPM when you know it's planetary gear ratio. On the other hand rear wheel RPM would suffice if you want to know what gets transferred to the road.
The only E-Scooter dyno run documentation I have seen as far was of the ZEV7100, and that was recorded as power over speed (km/h or mph), from which hub motor torque over motor RPM can easily be clalculated. This was done on a Dynojet dyno and analysed via Winpep7. Apparently there was some mode switching done with the ZEV's "gears", because there are two dips in the power curve, one at 30 and one at 55km/h. Normally our direct drive electric rides should produce a continuous power/torque curve without such dropouts.
I am also fixed on dynoing my Thunder, but not in winter. The Honda dealer I talked to was quite apprehensive about putting an E-Scooter on his dyno, and apparently it cannot do all I would require, so I will probably team up with my dealer and plan on a dyno round at Stuttgart University who are into E-Mobility anyway. He with his stock "Fury", I with my tuned version:-)

When you do this,

please use a cycle analyst to see whats being drawn from the battery
It will give you an opportunity to measure the efficiency of the system :D

I have been meaning to do this myself, now that a few people have done a battery change as I have, the information may be more useful

Cheers,
Matt

i'll have to buy one first and fit it ! Are they expensive ?

Ah fair enough,

they can be had here:
http://www.ebikes.ca/store/store_accessories.php

you need the large screen high current model

you will also need this shunt:
http://www.ebay.com.au/itm/Shunt-Resistor-DC-100A-75mV-Current-Meter-Ammeter-/300632511737?pt=LH_DefaultDomain_0&hash=item45ff18...

you have the can bus cable right?
you could do a memory dump while doing the dyno (that may actually be more useful)

Does the dyno output a table to values aswell as a graph?

Also,
any chance you could do a run with both the old firmware and new firmware (it'd be good to put a number to the performance difference)

Matt

Recommend when you do the dyno test since the scooter itn't moving to use a house fan on high blowing on the motor area so you don't overheat it.

Over what will the dyno record power then? It does need something for the horizontal axis, or ist really just power? But that in itself will require an rpm reading, else it cannot calculate power???

Chris,

If you've ever worked with arduino or similar programmable microprocessors, you might find an easy fix in it.
It could easily spot a white dot on a rim/wheel with a simple optical sensor (even a lightsensitive diode) and output any kind of solid signal to virtually any device.
google it up and you'll be surprised how cheap and versatile these little toys are.

turok

24hp = 17.8967kw ....hmm I'm missing a couple of kw and thats important to me !.

hp is hp regardless of load (the dyno usually accounts for inertia change)

Given your battery still has marginal cells, you may be bouncing off the low voltage limit

up to 50kmh, the controller is operating at full field strengthening
50kmh to 75kmh, the field strengthening is backing off, PF is improving (more motor current is going to acceleration than field strengthening).
past 75kmh, field weakening is being used to reduce motor voltage

Matt

So yes, it is speed as the horizontal axis, just like I thought. Let me do a little math now to come up with a graph of your rear wheel torque :-) 18kW is quite something!!!
But there appears to be a very long startup limiting slope, though that could be misleading as this is only the power graph. What you feel in the behind is the result of torque.

OK, so here is the wheel torque from that dyno run:

I don't really see any field strengthening activity here, just a near-constant wheel torque of around 266Nm possibly from standstill. Despite the hefty weight of the VX1 that should rather feel like some serious acceleration!
Approximately from 50km/h onwards back-EMF sets in and begins to limit motor current and thus it's torque output.
If field weakening is applied by the controller (which again is not necessarily so) it would probably be from around 70 to 75km/h onwards.
At slightly above 95km/h the controller cuts off current and thus torque and power quite abruptly. This means the scooter will likely reach not much more than 95km/h on a level road, and accelerating up to that speed may feel like bumping into an invisible barrier.

266Nm is lower than I would expect from a 65Nm capable motor behind a 6:1 gearbox...interesting

If comparing values between bikes, I would use either power, or applied force (~1075N measured on this dyno run), as it removes any changes in gearing or wheel size from the equation...
Or even power/mass and accleration (N/kg)

Re top speed:
power does not drop to 0 once top speed has been reached.
It drops to the minimum needed to hold 100kmh real (105kmh indicated)

A PID control loop is used, updated many times a second:
speed exceeds 100kmh = reduce power
speed still increasing = reduce power more
speed now falling = keep power same

speed falls below 100kmh with throttle still high = increase power
speed still falling = increase power more
speed increasing = keep power same

You are right in that it feels like an invisible barrier, but the top speed is the same irrelevant of prevailing conditions (up hill, down hill, head wind, etc), assuming the battery has enough charge to put out the power required

Most dynos have a specific speed ramp, needed to allow the influence of momentum to be taken into account
The speed change per second is constant, even if the dyno has to accelerate the motor, instead of the other way round.
thats why power falls to 0 at what looks like 95kmh

New software and a good battery allow 108kmh top speed real (though ~14kw continuous is needed at times to hold that)

Matt

MEroller
You are close! The phase advance (field weakening) begins at 3500RPMs (70kph). It increases with RPM up to a max of 28 electrical degrees at 5000RPM. It may surprise you , but the controller uses proportional control only (P) not (PID).

MEroller
The phase advance (field weakening) begins at 3500RPMs (70kph). It increases with RPM up to a max of 28 electrical degrees at 5000RPM

On my bike, the speed the field weakening starts feels like it changes based upon actual pack voltage

For instance, when I was running 44 cells, start of the constant power region would not begin until above 80kmh (pack voltage ~134V @ 220A)

Also, I could have sworn there was phase retardation (field strengthening) occurring below 45kmh

If that's not the case, than a bit more firmware hacking could improve bottom end acceleration :D

Matt

266Nm is lower than I would expect from a 65Nm capable motor behind a 6:1 gearbox...interesting

That would indeed be a great loss if theoretical 65 x 6 = 390Nm would only show up as 266Nm on the wheel. However, I strongly suspect that the motor is only capable of 65Nm when supplied with the full battery pack voltage and would probably burn to death within a few minutes of operation at full blast. Remember that the natural torque over RPM "curve" of a synchronous electric motor is a straight line from a maximum torque at standstill right down to 0 torque at the max rpm where back-EMF is just about equal to supply voltage, leaving only a tiny current to overcome friction. But as can be seen from the torque curve from this dyno test the lower rpm region is strongly limited by the controller up to 50km/h. If you took the gradient from over 50km/h and extended that to 0 km/h you would actually come very close to the theoretical 390Nm at standstill. Interesting indeed!

Most dynos have a specific speed ramp, needed to allow the influence of momentum to be taken into account
The speed change per second is constant, even if the dyno has to accelerate the motor, instead of the other way round.
thats why power falls to 0 at what looks like 95kmh

Well, these Dynojet dynamometers are purely inertia-based, so nothing is constant except for the dyno's mass moment of inertia. This means that 0Nm at 100km/h is just that - 0 Nm. No torque left to turn the wheel nor the dyno any faster. These dynos cannot actively drive the wheel, only passively brake or accelerate via the inertia. So this power plot indicates that the controller software is still the older 95km/h type. But it is true, this kind of limiting results in a very stable top speed with little speed changes despite varying driving conditions such as headwinds or uphill slopes :-)

Matt
The phase advance does not depend on voltage at all. It is applied based only on RPM. The gain of the current loop also increases as the RPMs increase. Perhaps this is what you are sensing. The phase current is limited to about 375A in order to protect the IGBT. Off the line with full throttle, this limit is reached immediately, then begins to back off as the RPMs increase. The older MC SW allowed a little more phase current but not much. It was decreased to 375A to give the IGBT a little more heat margin. Unlimited, you could probably smoke the tires, but only once.