Timing Adjustment Tool

SplinterOz said:
Arlo1
You have to remember that the PWM is turning on and off many times (100's I think) per electrical cycle of the motor. The position of the Halls (either mechanically or electrically) has no effect on how well the PWM turns on and off. The timing can only effect the trapezoidal change event timing out of the controller so I am with toolman2 here (literally and figuratively) I think that RPM would probably be more than 90% of the issue if not 100% of it.
Click to expand...


Bad form to reply to yourself but here goes...

I have re-read what you said and basically you are saying that due to the fact under load you are putting more energy into the stator the flux will build quicker therefore reducing the delay that the timing advance is meant to overcome.

Here is a good question then... is the delay a factor of the motor or the controller? When toolman2 and I were measuring it with my motor the delay appeared to be a static value of time. This of course translates to a linear increase in retardation in degrees. From memory this delay was even present and constant when we had the motor being driven mechanically and the controller (using the hall sensors from the turned motor) was driving a non inductive load (in our case 3 x 240v halogen down lights). Based on this testing I think that we are fixing a controller delay not a motor delay and that we are really tuning the controller in relation to the number of poles

toolman2 does my recollection sound right to you? Burtie, Jeremy does this sound right?

or should I just shut up :?
 
The load doesn't directly effect timing needs.

The reason throttle position should be involved in the timing is to optimize the performance vs economy balance. If you're looking to milk all the power possible from a motor, yet still cruise at the most efficient setting, then you run retarded timing at lower throttle positions to arrive at whatever it takes to be using the least amount of current to pull the RPM you desire. However, this position would be terrible for performance, and it reduces the RPM range of the motor as well (drops Kv). So, your throttle input signal is there to tell the tell the box, OK, I'm asking for 100% throttle, advance timing (which in-turn simulates lowered BEMF), and the motor draws more current and spins to a higher RPM than would otherwise be possible without the timing advance.

Normally you have to simply pick either a timing retard for economy, a timing advance for performance, or a neutral position that does neither well, but both OK.

With Burties magic-box, you get to instantly have both as you need them.

This is just insanely cool.

Burtie, you are the man. You saw and understood the need, you struggled to create it, you refined it, you made some protos and they do what we always dreamed of having the power to do. Huge respect to you my friend. You should throw a paypal address out there and the folks who have so many problems solved by your clever work can at least kick in a few bucks to buy you a nice after Christmas dinner out.
 
-that i agree with luke, i think youve nailed it.
we may need to purge most of the IC engine stuff from our heads here, BUT there is a good few extra rpm to be had with further advance (obviously with higher amp draw) so yes, theres a case for say "boosted mode" that cranks things up -and its not really efficient for constant use..
and the TA kinda allready has this option, so yes put up your paypal burtie so your fine work can be rewarded. :D
i might get back with some numbers on this kv increase thing.
 
Whilst I'm sure that rpm is the dominant factor controlling timing, I can't help but feel that load (or more specifically current) will affect it to. I'm reasonably sure that the motor effective inductance will change with current, particularly as the stator gets close to saturation flux, plus the winding resistance will change as the wire heats up from increased current. Both of these may act to change the motor time constant and this *may* have an effect on the timing.

There's also the issue of motor acceleration. The timing is most probably dependent on the instantaneous rate of change of the magnet pole position relative to the stator slot plus the torque demand. This may mean that the timing needs to be proportional not just to rpm, and possibly load, but also speed demand. For example, a high speed demand (i.e. a big difference between the current motor rpm and the demanded motor rpm) will mean a high phase current (with a possible motor time constant change) and also a timing requirement that maximises the torque available from the motor at it's maximum angular acceleration rate. This may well be a different timing than that required to run the motor at the same rpm but with no change in load demand.

I'm not at all certain about all this, but there's something nagging away telling me that the timing won't just be proportional to rpm.

Jeremy
 
We are all guessing and i realy hope no one see this as me trying to argue! I am just so excited for this its hard not to want to keep guessing at what we need. I was trying to refresh my memory with what castle creations did and found a little info like luke said and thinking about this now i for see the end product having a three plot system or more.
The thing i thing i just realized is throtle position is not the same as current! So then you will likely want three plots for the adjustment!
#1. Rpm most important
#2. Tps
#3. Load (current)
The throtle will tell how fast you want to go and try to select better timing for efficiency at low positions and max speed at hi positions.
The load will change with out the throtle so as the current goes up and down the timing will need to be adjusted to whats best for thats load vs rpm vs tps (how efficint the rider wants to be)
This is just what I predict we will have in our futures.
 
Hi guys,

Surely, as far as the motor is concerned, load demand and speed demand are all part of the same thing.

There is the power generated and there is the power the load consumes. If they are not equal then the rpm will change. If the load inertia is larger than the motor's then its just Newton's 2nd law. The motor doesn't know whether its being fed power in order to turn the load or in order to speed up; it's all the same thing.

I think it's only when you introduce possible speed or current control loops inside the controller that there's a difference.

Back to the ICE analogy. Diesel and petrol engines used to have very different feels to them, partly because the throttle on a petrol engine controlled power and the throttle on a diesel was an input to a mechanical governor, so it was a speed control. I think the electronic management systems on modern diesel cars have changed that and made them much more like petrol engines.

Nick
 
Tiberius said:
Hi guys,

Surely, as far as the motor is concerned, load demand and speed demand are all part of the same thing.

There is the power generated and there is the power the load consumes. If they are not equal then the rpm will change. If the load inertia is larger than the motor's then its just Newton's 2nd law. The motor doesn't know whether its being fed power in order to turn the load or in order to speed up; it's all the same thing.

I think it's only when you introduce possible speed or current control loops inside the controller that there's a difference.

Back to the ICE analogy. Diesel and petrol engines used to have very different feels to them, partly because the throttle on a petrol engine controlled power and the throttle on a diesel was an input to a mechanical governor, so it was a speed control. I think the electronic management systems on modern diesel cars have changed that and made them much more like petrol engines.

Nick
Click to expand...

The motor can easily have a high load demand (i.e. high throttle setting) yet be running at low rpm, so there isn't necessarily a tight link between the two. The PWM duty cycle will be high at the low rpm, high throttle demand, phase currents will be high, the motor inductance may be lowered (because of the high current) and motor resistance may be increased as a consequence of high I²R losses and subsequent heating. The torque demand will be high, which may demand a timing that maximises torque even though the motor rpm might not normally require that degree of advance/retard.

Jeremy
 
Yup. Lets word it like this you want to go 20km/h so you twist the throtle to 50% because thats what usaly gives you 20km/h then as your cruising at 20 k on the flat you start to go up hill you still want 20 km/h but something has to change for the increased load on the total system. And i feel 50% throtle should still mean you want to go 20 km/h
 
Jeremy, Arlo,

I agree with what you say, but I still think, as far as the motor is concerned, demand is demand. Maybe we should call it something else to avoid confusion with control loops.

So there are only two parameters, not three, to consider.
One is rpm, or speed.
The other is what we refer in these parts as "wallop".

Wallop could be measured from throttle position, torque or phase current, but its all the same. Take Jeremy's example of low rpm and high throttle opening. The motor doesn't know if it's being given the wallop in order to maintain rpm with a high load, or to accelerate with a low load. At least to a first approximation, it's all the same.

Arlo's example sounds like a controller with a constant speed function - ie, TPS translates to speed. In that case the controller varies the wallop according to the load. The answer here is to remove TPS from the list of parameters and just use rpm and wallop.

Nick
 
The idea with the timing is to get the magnetic flux to reach its peak right when it lines up with the right spot on the rotor.

The motor phase current (motor torque) definitely affects how fast the flux builds up. I think it will peak sooner with higher load, so the timing should be less advanced as current increases.

I also think there may be a way to look at the slope of the current signal (dI/dt) to tell where the sweet spot is.

Here's something I dug up that sort of eplains it, along with their very high tech solution:
http://www.toshiba-components.com/microcontroller/M370_FOC.html
 
Tiberius said:
Jeremy, Arlo,

I agree with what you say, but I still think, as far as the motor is concerned, demand is demand. Maybe we should call it something else to avoid confusion with control loops.

So there are only two parameters, not three, to consider.
One is rpm, or speed.
The other is what we refer in these parts as "wallop".

Wallop could be measured from throttle position, torque or phase current, but its all the same. Take Jeremy's example of low rpm and high throttle opening. The motor doesn't know if it's being given the wallop in order to maintain rpm with a high load, or to accelerate with a low load. At least to a first approximation, it's all the same.

Arlo's example sounds like a controller with a constant speed function - ie, TPS translates to speed. In that case the controller varies the wallop according to the load. The answer here is to remove TPS from the list of parameters and just use rpm and wallop.

Nick
Click to expand...
I dont think you understand what I am getting at. The term "wallop" you use is more a final calculation of the two (tps and current.) What I am trying to say is the tps is what people twist to get a desired acceleration to a desired speed and stay that speed but when the terain changes or there is more or less wind or a up or down hill doesnt meen the user wants to go faster or slower so they will not want to move the throtle and the load will change indepentantly from the throttle.
 
arlo,luke thank you for the explanation and understand more now how this device work i am very dumb with électronics
hey burnie !! your the man !! can you tell if your device can work better then my 120% trottle setting? ,
i am ready to buy one and test it on my infineon controller + my 2807 motor for the science :lol: :lol:

ok exemple here of what i want if is it possible with your device:

ok right now i can ride my bike with 120% with my 12s lipo pack, i can obtain a top speed of 47 kph but i suck i little bit more current

what gonna happen with your device with a performance timing and a economy timing with the same same set up ?

if i understand the economy mode will be saving current and keeping my top speed

performance mode is a brute amount of current goes trougth the motor faster then normal time is take to reach the motor ? so less loss of time in the acceleration

sorry guys i just try to learn :D
 
lifepo4ever said:
arlo,luke thank you for the explanation and understand more now how this device work i am very dumb with électronics
hey burnie !! your the man !! can you tell if your device can work better then my 120% trottle setting? ,
i am ready to buy one and test it on my infineon controller + my 2807 motor for the science :lol: :lol:

ok exemple here of what i want if is it possible with your device:

ok right now i can ride my bike with 120% with my 12s lipo pack, i can obtain a top speed of 47 kph but i suck i little bit more current

what gonna happen with your device with a performance timing and a economy timing with the same same set up ?

if i understand the economy mode will be saving current and keeping my top speed

performance mode is a brute amount of current goes trougth the motor faster then normal time is take to reach the motor ? so less loss of time in the acceleration

sorry guys i just try to learn :D
Click to expand...
Its almost certian this will be better then the 120% setting there is a very good chance you are not at the optamum timing for your motor so having a chance to try and move the timing around will let you find out and save engine and bettery life once you do!
 
Happy new year folks,
I appreciate all the kind words, encouragement and suggestions from everybody -Many thanks :)

I am busy redesigning the hardware for the TA, so it uses a ucontroller chip that is slightly better optimised to the job. Using a new PCB which reduces the overall size a little bit too.
The parts are all on order and the PCB design is almost ready to go out for manufacture, just waiting for me to try out the new design with a hand wired one first.

If anybody wants to try one without waiting, I have a couple of the prototypes that I would sell (@ 49 GBP each) drop me a PM if you are interested.

I will start a proper thread in the for sale section when the new boards are available.
I must also try and write down some instructions of how to use the thing too.

@lifepo4ever
No one has tried one of these devices with a hub motor yet. It should work, but I cant tell what degree of benefit you are likely to get with your DD hub.
I would expect the benefits may show more with the geared hub motors (BMC, MAC etc) because the electrical rpm of these motors can be very high, so any delays that we can tune out should have a big effect :wink: :?:

Burtie
 
Burtie said:
I must also try and write down some instructions of how to use the thing too.
Click to expand...
If you haven't done this before you might check-out Goodrum's website, Click Here & Scroll Down for Instructions and browse through a few of his instructions for his battery related electronics kits. Having some good build reference photos and a parts layout helps too. Even a short video of using your software to tune a motor would be very (extremely! :shock: :mrgreen: ) helpful too. 8)

Anyway, it's fantastic to see so many ES members get excited about this; the chum is in the water & the sharks are frenzied for the bait. :twisted: :lol:

Good to know the feed-parts are about to hit the water... :wink: :twisted: :lol:
 
thank you for your reply burtie i will wait a little until you made some test with 9c hub or if someone try out , I am satisfied with my 120% for now but i use more current that why am asking if this tool can be useful for a better tuning .
 
I had been curious to see how well the TA would work with a hub motor for some time now.

Here is a video of a test I did using the TA with a MAC geared hubmotor.....

http://www.youtube.com/watch?v=bopz1KcE3bk
[youtube]bopz1KcE3bk[/youtube]


Summary of the figures:
MAC Magnet count = 32 (16 pole pairs), Planetary reduction motor to rim is 5:1

0 deg advance per 2000rpm, 3.3 amps, 2240rpm
5 deg advance per 2000rpm, 4 amps, 2372rpm
10 deg advance per 2000rpm, 4.2 amps, 2463rpm
15 deg advance per 2000rpm, 5 amps, 2673rpm
20 deg advance per 2000rpm, 18amps, 4783+ rpm :shock: (not even WOT)

The effect on the Kv of this hub motor by advancing the timing is amazing. At nearly 5000 rpm we are applying about 50 electrical degrees advance (which is a lot) and the speed of the motor has more than doubled!

I dont quite understand why the effect is so massive, I will try and get more testing done soon with a proper load on the motor.

:D
Burtie
 
OMG!

Burtie!

First, I want to tell you that you're a superhero.


Second, that video was intense!!

I have a hunch the 18amp current was so high at least in-part due to the system not being at equilibrium, and still putting power into making the wheel accelerate. Possible that once it reached speed it would drop down to something closer to reasonable.

I would love to see a test where you retard timing, and observe no-load currents.


This is cutting edge stuff.
 
Were the lower advance tests done at WOT? If not, there might be some problem with the throttle setting somehow getting buggered when the 20 degree advance is cranked in.

At 20 deg/2000 rpm, you are potentially creeping up on the same effect as rotating the hall phase wires one step (depending upon if the motor uses 60 or 120 degree halls)... Hmmm are the motor halls wired wrong and by getting the advance up so far, it gets into the realm of proper operation?
 
This reminds me of variable ignition timing on gas engines.

Funny to think of it this way, but i see an electric motor as sort of the same. It's ultimately deciding when to fire off the 'ignition' based on the RPM. You fire off too late and the power is wasted. Fire too early and you can spin the motor quicker but you waste some power. Fire just at the right time and you have max efficiency.

I wonder if the controllers we use have any kind of 'timing map' built into them at all? If not, i think something like this would help broaden out the efficiency curve massively.. and add some power at the top and lower end of the rpm range.
 
texaspyro said:
Were the lower advance tests done at WOT? If not, there might be some problem with the throttle setting somehow getting buggered when the 20 degree advance is cranked in.

At 20 deg/2000 rpm, you are potentially creeping up on the same effect as rotating the hall phase wires one step (depending upon if the motor uses 60 or 120 degree halls)... Hmmm are the motor halls wired wrong and by getting the advance up so far, it gets into the realm of proper operation?
Click to expand...


If I'm not mistaken my friend, the "degree's" advance are actually in e-degree's, not physical rotor body degree's, so when he say's 20, it's just the equilivant of a few mm actual hall position.
 
Timing advance should be much like field weakening. You could get a similar effect by weakening the magnets (but changing the timing is much easier). If you take this too far, the efficiency will drop off. At very light loads, you could advance the timing to extend the power band. At high loads, the optimum timing will be much closer to neutral.

I've done similar experiments with the timing on brushed motors by shifting the position of the brushes.
 
Hi texaspyro,
All tests were conducted at WOT except for the final one, where I was concerned the wheel speed was becoming dangerously high and stopped before WOT was reached.

I agree the advance of 50 e degrees is close to overlapping the next commutation state, however the hall sequence on this motor is fine, it has been ridden for nearly a year with expected performance. The terminal speed at 44v on the road was close to 30mph, which is normal for this MAC hub.



In the first test on the video, wih motor rpm of 2240 = wheel rpm of 448, the rim speed is about 35mph --> Normal behaviour.

In the last test with motor RPM = 4783, the rim speed is in excess of 75mph (120kph)------------------> Definitely not normal behaviour.


I like Fetchers explanation about field weakening coming into play as the timing is advanced, so there will probably be an associated loss of available torque.

Seeing this happen makes me think that there may be a good case to manage this effect to our advantage by monitoring and reacting to the current draw of the controller, as others have already suggested.

Trouble is, we would need a dyno to enable us to setup and test this feature........ Maybe I will have another chat with Jeremy Harris :wink:


Luke, -good idea, when I get time, I will try some testing with retarded timing and see what happens to the current draw.
 
You must log in or register to reply here.
Back
Top