[Jimdear2]

Hello QER and Tesseract.

I was wondering if your controller is able to be set to a "do not exceed" RPM but still able to maintain maximum current outputs available from the battery pack for maximum torque at the selected RPM. Please excuse me if the questions I'm asking have already been answered. I just may not have been able to understand the answer as presented or I may have missed it while in the hospital.

For the last 2 years I have been seriously applying EV tech to tractor pulling. One (of many) problems I have had is maintaining a fixed wheel speed irregardless of ground speed. Stock tractors generally use brute traction and torque to move the weight. Super stock and modified tractors use a combination of friction and traction, In many cases friction being the primary.

I just aquired a rebuilt 13 inch motor for the cost of components used in the rebuild. I plan on using the motor in a modified ultralight tractor with suitable modifications to run it at 120 + volts and 800 + amps. The total 1350 lbs battery, motor, driver and tractor weight will determine the maximums.

I was initially planing on a removable chopper type controller using low voltage for low speed driving and back up duties, with a separate set of contactors to apply direct battery increading voltage/amprage for pulling. BUT . . . Limiting RPM to keep the motor from exploding in this situation becomes a problem. One slippery spot or one bounce might destroy a motor. Backing off thre throttle will lower the voltage and rpm but it also will reduce tire speed which affects ground speed which effects inertia, etc. etc.

My current, self designed, RPM limiter just uses a tachometer shift light signal to shut down the main contactor and requires a reset to continue. The driver has to be on top of things to keep the shutdown from happening. It saves a motor but looses the pull. I tried setting up the system to come back up when the shift light went off but the cyclic jerking that produced was not good.

Is it possible using the software hardware in your controller to set a RPM limit that will adjust voltage to maintain up to a maximum RPM and still allow maximum current to develope maximum torque. Kind of a maximum power cruise control.

In 90% of the hooks you do, it's pedal to the metal, hang on and steer, maybe touch a steering brake. ICE tractors rely on soft rev limiters (randomly shut off a random cylinder) to maintain a safe rpm.

Ideally what I want is a reasonable assurance that I can get up to a fixed tire speed, I can then adjust tire pressure and ballast weight to control traction and ground speed. Then when the weight starts coming down I can know that maximum available power is going through my tire/weight set up.

If I wasn't clear in the above please ask any questions needed.

Have a great day,

[Tesseract]

Originally Posted by DIYguy

...
Secondly, regarding the water cooling.... Is there any recommendation on on cooling capacity? I read the 1 - 2 gpm... but obviously the method of cooling the fluid can change a lot.... ie; reservoir only vs small rad/fan combination. Perhaps Dimitri has some feedback on this? (I did read your recommendation on ambient temp considerations)...

There is no simple answer to that question, which is why I don't even try to cover it in the owner's manual. Fortunately, though, the Soliton1 has significant heat-dissipation capability as is, so liquid cooling is only necessary if any of the following apply: 1) high ambient temp (>35C); 2) long periods above 400A (ie - above 75-80mph on the interstate); 3) hilly conditions (similar to #2).

So my first suggestion is to try driving around without liquid cooling at all. Collect some log files and look at the 7th column to see what the temp gets up to. If you routinely exceed 60C, especially given the much lower ambient temps up there, then I would consider liquid cooling. If you routinely exceed 70C then I would say you need liquid cooling.

If you want to keep the controller cool regardless (because you have OCD, for example), then there are infinite possibilities. Way too many to cover in a post like this, and nothing is really all that critical, but since you asked personally prefer a sump (reservoir) type system compared to the closed loop that dimitri has. As little as a liter or two of coolant in the reservoir is more than enough; a larger reservoir may eliminate the need for a heat exchanger completely (relying on the specific heat capacity to act as a "heat sink"). I've seen one clever installation that simply routed copper tubing around the engine compartment and relied on incidental airflow for cooling - cost effective, and works, too. Then there is the tried and true auto tranny cooler, the hardcore computer gamer's cooling systems (Seb likes these for some reason), etc... Sky's the limit.

[Qer]

Originally Posted by Jimdear2

Is it possible using the software hardware in your controller to set a RPM limit that will adjust voltage to maintain up to a maximum RPM and still allow maximum current to develope maximum torque. Kind of a maximum power cruise control.

That is a very good question and one that I unfortunately must answer with that I seriously don't know. From the tests we've done so far it's been proven that handling RPM is tricky as hell simply because it's very hard to measure. Typically when you measure current or voltage it's there, instantly (well, almost), when you measure RPM you actually can't measure RPM but you either measure pulses per time or time between pulses. This means that no matter how you do it when you get the results it's too late and the data you get is what was, not what is. This brings havoc to your calculations, trust me.

There are today two ways the controller handles RPM, idle-RPM and overrev. Idle-RPM tries to actively regulate the RPM and compensate for differences in load (mainly an AC shutting on and off etc) and it means that it's constantly trying to fine tune the current to keep the RPM. While this method works fairly well for medium loads I'm not sure it's that good at handling a tractor that's burning CURRENT and then suddenly slips. There are some safeguards that should catch it, but I would like to emphasis on should...

The other way, overrev, is probably better for your needs although I'm not sure it's a good way. What it does is simply to kill the current (very much like an ICE simply kills the ignition) and then the current ramps up with the current slew rate again so you actually get some degree of control there since you could experiment with different slew rates for optimum grip. Possibly. Maybe...

Our tests show that the overrev seems to work pretty well for saving the motor. You can see the vid here:

http://www.youtube.com/watch?v=aWJBEiNYqic

It might sound like Tesseract is only cranking the motor up and down several times, but the protection actually kicks in several times per peak to keep the RPM sane. If the controller didn't do that (or if the tach had failed in that test) the motor would've blown up. Somewhere I even have a log over it....

Ah, can't find it, but here's some data from my mockup to at least give you an idea how it behaves:



You could get the current to ramp much faster by increasing the slew rate, which you probably want to do anyway. The default slew rate targets a commuters needs rather than someone that wants to win a race.

[Jimdear2]

QER;

Thanks for the quick reply. I am not sure that I understand your information so let me reply with what I think you said.

Your RPM limit system from the graph shows current ramping up to 700 peak amps at about 272 seconds with the RPM reaching 5000 at about 273 seconds at which point the current drops to zero at 274 seconds the RPM drops to 3500 and the current has risen to about 350 amps and continues back to 700 amps and the RPM again rises to 5000 where thwee whole sequence starts again. That is equivaqlent to the driver flooring and letting off the throttle on a two second cycle. A sequence like that would produce broken drivetrain parts or total loss of momentum every time. That is why I went to a complete shutdown on my own system.

A far as the video went, do you mean that the periods between that violent application and reduction of throttle (motor jerks severly) that the controller had pulsed the motor severel times to control the RPM. Or was the several severe jerks the controller applying RPM control.

I envision a rpm limiter as allowing you to apply full throttle to the motor with no load and having the motor rise to a set rpm and maintain itself there

The first system you described sounds much more like what I am looking for. As you describe it. Idle-RPM tries to actively regulate the RPM and compensate for differences in load (mainly an AC shutting on and off etc) and it means that it's constantly trying to fine tune the current to keep the RPM. From what I have read it sounds like you may be trying to overcontrol. In an ICE I doubt that the range is tighter then +/- 100 rpm and may even be wider then that.

As I have said before I might not understand what is happening or the limits of your controller. But is it possible for your idle control system paremeters to be set at 5000 rpm,1000 amps and activated by the full throttle signal instead of the no throttle signal. If the system could maintain even +/- a couple of hundered RPM at that level it would be a succses.

Jim

Originally Posted by Qer

That is a very good question and one that I unfortunately must answer with that I seriously don't know. From the tests we've done so far it's been proven that handling RPM is tricky as hell simply because it's very hard to measure. Typically when you measure current or voltage it's there, instantly (well, almost), when you measure RPM you actually can't measure RPM but you either measure pulses per time or time between pulses. This means that no matter how you do it when you get the results it's too late and the data you get is what was, not what is. This brings havoc to your calculations, trust me.

There are today two ways the controller handles RPM, idle-RPM and overrev. Idle-RPM tries to actively regulate the RPM and compensate for differences in load (mainly an AC shutting on and off etc) and it means that it's constantly trying to fine tune the current to keep the RPM. While this method works fairly well for medium loads I'm not sure it's that good at handling a tractor that's burning CURRENT and then suddenly slips. There are some safeguards that should catch it, but I would like to emphasis on should...

The other way, overrev, is probably better for your needs although I'm not sure it's a good way. What it does is simply to kill the current (very much like an ICE simply kills the ignition) and then the current ramps up with the current slew rate again so you actually get some degree of control there since you could experiment with different slew rates for optimum grip. Possibly. Maybe...

Our tests show that the overrev seems to work pretty well for saving the motor. You can see the vid here:

http://www.youtube.com/watch?v=aWJBEiNYqic

It might sound like Tesseract is only cranking the motor up and down several times, but the protection actually kicks in several times per peak to keep the RPM sane. If the controller didn't do that (or if the tach had failed in that test) the motor would've blown up. Somewhere I even have a log over it....

Ah, can't find it, but here's some data from my mockup to at least give you an idea how it behaves:



You could get the current to ramp much faster by increasing the slew rate, which you probably want to do anyway. The default slew rate targets a commuters needs rather than someone that wants to win a race.

[Woodsmith]

Hi Jim,

A lot of this controller stuff is way over my head, and I may be speaking out of the other end of my anatomy, but from what I gather, shunt motors will produce a steady speed regardless of load and just draw more and more amps to do so until it cooks itself. Would that be more useful for what you need when pulling?

[Jimdear2]

Originally Posted by Woodsmith

Hi Jim,

A lot of this controller stuff is way over my head, and I may be speaking out of the other end of my anatomy, but from what I gather, shunt motors will produce a steady speed regardless of load and just draw more and more amps to do so until it cooks itself. Would that be more useful for what you need when pulling?

Woodsmith,

I had also thought of shunt motors but I don't know enough yet.

Plus when somebody offers me a completly rebuilt 13 inch series motor for what amounts to lunch money and pickup within driving distance, I grab it and run with it.

Right now I have in the shed an unfinished rolling chassis, a short powerglide with a 1.83 low gear a narrowed Ford 9 inch with a 5.14 gear and a large number of 80 to 90% 16 amp hour Hawkers. As far as I know there is no controller and shunt motor available to me that would produce the kind of power I think this set up would make. I'm already out in the cold unknown with this.

Drag racing stuff really doesn't apply, the oppisites apply. I need maximum power (torque) at the end of the pull, RPM above a point that keeps the tires turning is irrelivant. I need maximum speed at the beginning to build up inertia. Thats why I'm hoping that EVNetics can help. I saved enough on the motor that I can start looking at better controllers. When I read about the idle control that they were working out I was hoping their RPM limiter worked similar.

This is probably more of an answer then you expected, actually responses like this help me clear up the muddy waters. Thanks for commenting and listening.

[Qer]

Originally Posted by Jimdear2

Your RPM limit system from the graph shows current ramping up to 700 peak amps at about 272 seconds with the RPM reaching 5000 at about 273 seconds at which point the current drops to zero at 274 seconds the RPM drops to 3500 and the current has risen to about 350 amps and continues back to 700 amps and the RPM again rises to 5000 where thwee whole sequence starts again.

Yes, but this is with a current slewrate of 500 or 600 A/s, you can change that setting to as high as 25000A/s. For a commuter such a high slew rate doesn't make sense (not to mention that the controller managed to break the tranny on a buggy that way ) but for you perhaps an extremely high slew rate actually makes sense?

Originally Posted by Jimdear2

A far as the video went, do you mean that the periods between that violent application and reduction of throttle (motor jerks severly) that the controller had pulsed the motor severel times to control the RPM. Or was the several severe jerks the controller applying RPM control.

It's Tesseract cranking up and down the throttle as the motor jumps around, but it's the overrev protection that makes sure the motor doesn't fling itself to pieces when it hits max rpm.

Originally Posted by Jimdear2

The first system you described sounds much more like what I am looking for.

That depends on how much the load change. What I worry about is that a slip and then suddenly grip again might make the idle algorithm start oscillating or it will force it outside it's controlling range which, on the other hand, will simply mean that the race is over.

[Woodsmith]

Sorry to be OT.

Hi Jim,

Maybe a twin motor set up would work. A shunt motor twinned with a series motor?
The controller will get the series motor to give you the speed at the start and part way through you power up the shunt motor as well to keep the wheels turning at maximum torque.

I'm just imagining a big knife switch, like Dr Frankinstein uses, to connect the shunt motor.

[280z1975]

Originally Posted by Qer

snip ...

Our tests show that the overrev seems to work pretty well for saving the motor. You can see the vid here:

http://www.youtube.com/watch?v=aWJBEiNYqic

snip ...

Ummm ... why is the motor just laying on the ground? I know it has little inertia to it, but still jumps a bit ... all I can think about is a motor overspeeding and blowing up with peices flying everywhere ... any reason why it's not mounted for these tests?

[dimitri]

Originally Posted by DIYguy

Secondly, regarding the water cooling.... Is there any recommendation on on cooling capacity? I read the 1 - 2 gpm... but obviously the method of cooling the fluid can change a lot.... ie; reservoir only vs small rad/fan combination. Perhaps Dimitri has some feedback on this? (I did read your recommendation on ambient temp considerations)

Thanks,
Gary

Gary, if anything I would recommend against doing the setup I have done. I wanted to route water thru my dash heater core, which drove my decision to buy automotive grade pump, which required multiple bronze fittings to step up to the hose diameter. When it was all done, I found that I can't even feel a slightest increase in cabin heat from controller, which makes the whole setup too complicated and not worth the trouble.

If I was to do it again, I would probably use small PC cooling pump, small ATF radiator (or maybe just a copper tube loop) and small expansion tank ( useful to drain the air from the system ), basically same thing Tesseract recommended.

I also second the notion of trying just the air cooling first, you may not need water cooling at all. I had mine done because I was testing first beta unit of Soliton1 which had different power ratings and more restrictive temp controls. Production unit is much more forgiving in this sense.

Hope this helps.