Fechter's Throttle Interface for RC controllers

[fechter]

If you are using two motors and two controllers to get to 200A, you could always use two throttle boards, fed from the same throttle.

Actually, I think you could use one throttle board and split the servo connection to two controllers. As long as the motors and controllers are the same, it should share evenly.

 

[northernmike]

Richard, will a bigger load produce different performance with the overcurrent throttle reduction?

ie, when YOU dial it in with a load, it might be different than the load I am going to present to it - so might the "reaction speed" adjustment have to be a user adjustable parameter? I was thinking of this oscillation behaviour you're speaking of... how will you smooth it out?

+1 on the user adustable current limit also!!

I would not be surprised if my setup draws over 90A on startup... how will the board behave it it does see a 100A draw for, let's say, 5 seconds?

 

[fechter]

Once the limiter response is dialed in, it shouldn't matter much what the actual load is. That's how it works on my other limiter circuits. If the limit is set for 90 amps, and you try to pull 100, it will just top out at 90. If the response is too slow, there is a bit of overshoot in the current before it settles down. This is undesirable. I'm pretty sure I can get it 'just right'. Response may vary somewhat with different models of controllers, but I suspect it will be very minor.

Making the response speed adjustable is tough since it is done by changing the value of a capacitor. I'm planning to leave an extra set of empty holes in parallel with the feedback cap to allow 'trimming' the response if necessary.

 

[northernmike]

That's great. I'm beginning to wrap my head around this after the 5th re-read!

So - the Magura throttle will just plug in, yes?

There is the option of having a pot mounted outboard of the circuit (ie on the 'dashboard' or somewhere accessible) for adjusting the current limit, yes?

Also - could a switch be incorporated to turn off the current limiting function?

 

[fechter]

Yes to all.
I don't think you would want to disable the limiting, this might blow the controller, but you could use a switch to go between high and low settings.

 

[etard]

I like the adjustable current switch. It would be good for those of us with friends that want to "try it out", read : break it outright. (what happened to the crying emoticon?)

I am gonna ask this question one more time and then I promise you won't hear from me until I order my board:

Can this board be used in any way, shape or form to convey amps, volts, or ah used? Can this board at least have an led light up when it hits current limit, or LVC? Anybody....

 

[nomad85]

(what happened to the crying emoticon?)

its still there, just looks different.

 

[deecanio]

all the emoticons have been abducted and returned to us with peculiarly long faces - alien indeed, must have been when knightmb did the upgrade, weird tho that only some are odd shaped like these ones - :lol: 8)
seems they all have a touch of the "sarah jessica parkers".

Anyway thanks for the update Richard, seems like we are moving along quite nicely.
Give us a shout when we get anything i can physically get my hands on and i'll take 2!!!!!!

Cheers for the great work!!!

D

 

[liveforphysics]

Could an update be added to the first post in this thread that covers the features of this unit?

Is this a cell balanceing system, LVC, and throttle now? How many channels is the ballancer?

 

[recumpence]

This is a PWM RC throttle with adjustable current limiter. (Old style round emoticon)

Matt

 

[fechter]

I am gonna ask this question one more time and then I promise you won't hear from me until I order my board:

Can this board be used in any way, shape or form to convey amps, volts, or ah used? Can this board at least have an led light up when it hits current limit, or LVC? Anybody....

I wouldn't be too hard to add an LED that lights up when you hit the limit. With my other circuits, I did the opposite, the LED is on until you hit the limit, then it goes out. You're rarely looking at the circuit when riding, so the LED is more of a pilot light indicating the circuit has power. I usually watch my ammeter, and I can tell from that when I'm at the limit.

As far as displaying current, the output of the current sensor could be used to drive a meter of some sort, but it is V/2 referenced, so it would take some additional circuitry to make a useful display from it. I suppose I could use the unused half of the op-amp as a reference so you could have a 0-2v analog signal that corresponds to 0-100% of sensor rating. This would be fairly easy to hook up to an analog meter. Downside is if you accidently short out a meter wire, you blow the op amp or the current sensor. More resistors could be added to protect the lines.

I've been trying to keep the circuit down to the bare minimum number of parts to minimize power consumption, size, and cost. Any of the above features will add cost to the unit, but probably only a few dollars. I'll take a look at how hard it would be to fit the parts into the layout.

A microprocessor based system could use the current sensor like a CA and display amp-hours, etc. but that is not really possible in an analog system like mine.

 

[deecanio]

lol.

Richard i'll take the most basic board that works :lol:
too much messin' takes a lot more testin'

D

 

[etard]

Thanks Richard, that was what I was looking for. I didn't realize this was analog, I thought it was all digital and already processing this information, so why not display it? As for the LED, I would probably run a wire up to my throttle for that. But then again, I will be able to feel it hit the current limit anyway I would think.

Please keep it simple for now, and maybe in the future other features can be added. 8)
We appreciate all that you are doing for this endeavor.

 

[fechter]

OK, finally got enough time to reprogram the controller and wire things up.
My scooter has a mess of wires.

I reconfigured the batteries to give me 36v, since a full 48v pack would be too high right off the charger.

The first test sucked, as the limiter failed to work all together. A bit of troubleshooting reveled that the adjustment pot I got out of the junk box was, well, junk. It was open on one side.
Another dig in the 'pot box' found me one that's good. More testing tomorrow.

 

[etard]

I used to have a "pot box" but it looked a little different than that :lol: :wink:

 

[northernmike]

What a tease.

 

[fechter]

A bit more testing in..........
Seems there is still a considerable oscillation in the current when I change the load suddenly.
I was afraid something like this was going to be an issue.

When I increased the capacitance in the feedback loop to around 1uf, the oscillation was about 1hz and would smooth out after a few seconds.

The problem stems from a slight delay in the controller between a change in the input and the change in the output. If I slow the response of the limiter enough, it will dampen the oscillation, but makes the current regulation rather poor. I want good regulation without oscillation.

One approach I might use is to make the limter respond very quickly to over limit , but respond very slowly to undercurrent. I think I'm going to need to reconfigure something in the circuit.

I tried programming the fixed throttle limits, and that seems to work perfectly.

I tried changing the throttle response setting to high, and that helped a bit, but there is still a significant delay.

I tried setting the startup current to high and the thing has really impressive starting torque for a sensorless controller.

 

[lawsonuw]

I tried changing the throttle response setting to high, and that helped a bit, but there is still a significant delay.

Ah, sounds like the throttle ramping in the controller is biting you in the butt. I'd add something to limit how fast the output pulse of the current controller increases. Your could also try upping the rate the controller sends out pulses to the esc. I have one of the higher end Futaba radios and it puts out pulses at ~73Hz. This will provide a bit more bandwidth in the servo loop and allow higher gains.

Lawson

 

[deecanio]

Hi Richard,

mmmmm cannot help you, small brain see.
Do you think given the issues your having that the cable to program the esc will be an absolute must for us to be able to tweak our settings to get best performance??
I had thought i'd get one anyway but i ask as lots of folk may need one too?
Before i lob some money at Castle does anyone on the forum sell the cables????


Cheers,

D

 

[magudaman]

Its a commercially available product for about 21 bucks.

http://www.horizonhobby.com/Products/Default.aspx?ProdId=CSEPHXL&utm_source=froogle

 

[deecanio]

thanks magudaman.

i'll go the commercial route (hopefully UK) but i always like to keep money in the family if poss

Cheers,

D

 

[MitchJi]

Hi D,

i'll go the commercial route (hopefully UK) but i always like to keep money in the family if poss

Cheers,

D

I bet John would be happy to sell you one.

 

[northernmike]

Richard, what do you mean, "change the load suddenly"?

Suddenly, as in, starting off from a stop suddenly? Or ...

 

[swbluto]

A bit more testing in..........
Seems there is still a considerable oscillation in the current when I change the load suddenly.
I was afraid something like this was going to be an issue.

When I increased the capacitance in the feedback loop to around 1uf, the oscillation was about 1hz and would smooth out after a few seconds.

The problem stems from a slight delay in the controller between a change in the input and the change in the output. If I slow the response of the limiter enough, it will dampen the oscillation, but makes the current regulation rather poor. I want good regulation without oscillation.

One approach I might use is to make the limter respond very quickly to over limit , but respond very slowly to undercurrent. I think I'm going to need to reconfigure something in the circuit.

I tried programming the fixed throttle limits, and that seems to work perfectly.

I tried changing the throttle response setting to high, and that helped a bit, but there is still a significant delay.

I tried setting the startup current to high and the thing has really impressive starting torque for a sensorless controller.

You could use an uC. :wink:

 

[northernmike]

what, an ultra Capacitor?