adjustable Servo speed circuit [ramp control]

[olaf-lampe]

The idea behind the ramp control is to avoid suddenly increased throttle signals to the ESC.
I fully understand and agree with CC that longer periods of half throttle can kill the ESC. I don't see, how to avoid them in daily commuting, but that's another story...
I just want to regulate the ramp from 0 to 100% throttle to be around 0.5 seconds or 1 second.
Does that fall into the above mentioned scenario? I don't think so.

Question: Would the same 25A ESC survive sudden full throttle attacks from a 45A motor? EG using a pushbutton instead of a potmeter.
My 2 cents: No it doesn't, because impulse current can become much higher this way.
-Olaf
BTW: My CA120 is rewired to star and has a comfortable inner resistance of 15mOhm. Almost the same as the 80-100 130kV which has proven to work with a 160A CC ESC. I 'll use a 240A aquastar ESC and a centrifugal clutch.
The combination of all this makes it possible to use an RC-ESC for a HoBo motor in a chain driven landvehicle. ( propeller and airscrews are a much easier task as we know)

 

[Jeremy Harris]

My view is that the best solution to this throttle response problem isn't to faff about with slugging the throttle response down, it's to switch from speed control to torque control.

If, instead of allowing the throttle to control the voltage applied to the motor (by directly controlling the PWM duty cycle), we let the throttle control the current that the motor is drawing, then we get a throttle that will feel very close to the conventional response we're more used to.

I'm currently (no pun intended.......) working on this, but it's not as easy as it seems, as ideally what's needed is a PID controller with a fairly fast response. I think I have a way of doing this for a conventional voltage controlled controller with shunt current sensing (after a whole day of playing about) but I need to do some testing to see how it feels. Hopefully I should have something that works, and that can be replicated by others, over the next few days.

Jeremy

 

[HumboldtRc]

While I was taking a picture with three ramp-controls a big turnswitch, the turnigy servotester and a 3.7V_to_5V voltagebooster, I came up with a much simpler idea:

Just use one ( instead of three ) ramp-control and replace/rewire the relevant smd-potmeter with a standard-potmeter. Mount it on the 'dashboard' where the switch should have been located and we've got a stepless ramp-control

I still could make some marks on the dashboard with useful ramp-speeds for dry/wet/snow/ice conditions
The critical part is, that the useful range of ramp for my application is pretty small. I don't need a 10s ramp from 0 to 100%.
The replacement potmeter would therefor have a much smaller value ( eg 1kOhm instead of the original 10k) but there will be a resitor in series to spread the useful ramp-range to a full 270° turn. ( I hope I'm not confusing anyone )

-Olaf

Anyway here is the picture how it's not gonna be

What pot were you going to use? I was reading the output on pot, of the servo speed-regulator, and it had a weird output... O kohm-3.8kohm- in about the middle, then back to 3.3kohms at full right... i want to put a bigger one on it and mount it to the handle bars like you were going to.

I don't really care what other people are saying that it is bad for the esc's. I want it to try and cure the sync-loss issue, by smoothing out the pwm change be having a little delay.

Also on all my RC's (Trucks, buggies, crawlers, boats, planes, and heli's and my mountainboard) i only use full throttle probably less than 50% of the run time, and the esc's handle it fine.

 

[olaf-lampe]

Oops, I almost forgot that thread
I haven't thought about measuring the original potmeter. I was sure it'd be a common 5k pot?!

@ Jeremy
Are you sure the infineons are only speed controllers? I thought at least they would provide some kind of torque control. Would it be possible to adjust recuperation with a speed controller?

Torque control can be a real neck bracker when you try to code it yourself. Do you try to do it purely analog with only a pid chip? This way you wouldn't have to bother with slow or faulty AD conversion, but setting parameters is a bit tricky then.

 

[HumboldtRc]

I haven't thought about measuring the original potmeter. I was sure it'd be a common 5k pot?!

Yeah i thought it would have been 5k pot, also... But it's output was not linear... Any one with a clue of what pot to replace in with??

 

[electraflyit]

Jeremy

Iv'e been tinkering with Picaxe micros at the moment and am trying to work out how to insert one inline with the throttle to make a 'current throttle'.
Iv'e been slowly learning the basics of PID control (and Picaxe basic) that I would need to get this to work on a picaxe.
The 20X series chip can work from 8mhz to 64mhz, so can zip along quite well.
My take on the controls follows;

Set Point=Voltage from throttle (1 to 4v)
Input=Voltage from current shunt (scaled)
Output=Voltage to controller (1 to 4v)
proportional constant Kp, derivative constant Kd, and integral constant Ki
E(error) = Set point -input
Ed = E - Eprev
Ei = is the sum of all previous errors

Output = output + (E*Kp)+(Ed*Kd)+(Ei*Ki)

I think I can 'pwm' this value to give varying volts out. (The only way I can see to go from D to A)
Working on what next to do.
Any help would be great

Eddie

 

[olaf-lampe]

Electraflyit,
do you want to use the speed adjuster as filter/signalsmoothing or something? :?

I don't see how this is related to the thread?
-Olaf

 

[Jeremy Harris]

Jeremy

Iv'e been tinkering with Picaxe micros at the moment and am trying to work out how to insert one inline with the throttle to make a 'current throttle'.
Iv'e been slowly learning the basics of PID control (and Picaxe basic) that I would need to get this to work on a picaxe.
The 20X series chip can work from 8mhz to 64mhz, so can zip along quite well.
My take on the controls follows;

Set Point=Voltage from throttle (1 to 4v)
Input=Voltage from current shunt (scaled)
Output=Voltage to controller (1 to 4v)
proportional constant Kp, derivative constant Kd, and integral constant Ki
E(error) = Set point -input
Ed = E - Eprev
Ei = is the sum of all previous errors

Output = output + (E*Kp)+(Ed*Kd)+(Ei*Ki)

I think I can 'pwm' this value to give varying volts out. (The only way I can see to go from D to A)
Working on what next to do.
Any help would be great

Eddie

Olaf is right, this isn't really related to this thread, but this looks like a sound approach to getting power control, rather than speed control, Eddie. I've used the PWM function to get a D/A, it works well and is plenty fast enough for throttle control, the only issue with it is the slightly quirly way you need to set the PWM frequency and duty cycle in Picaxe basic.

Sorry for taking the thread off-track a bit, Olaf.

Jeremy

 

[olaf-lampe]

It probably sounds wierd , but if we took a infineon controller; configure the original throttle input to 'constant full throttle' and instead stimulate the current-feedback input ( shunt/R12?) with a modified magura throttle, wouldn't we get something like a current/torque control?
You see, I'm a real hardware man. I don't like coding/programming too much :lol:
-Olaf

 

[Jeremy Harris]

It probably sounds wierd , but if we took a infineon controller; configure the original throttle input to 'constant full throttle' and instead stimulate the current-feedback input ( shunt/R12?) with a modified magura throttle, wouldn't we get something like a current/torque control?
You see, I'm a real hardware man. I don't like coding/programming too much :lol:
-Olaf

Go try it!

It sounds like it would work, albeit in the reverse sense (a higher voltage from the former current sense connection would give a lower output power). You might be able to make a simple mixer, using an op amp, that takes teh current sense voltage and the throttle command voltage and provides the correct sense signal for current control. That way, you'd retain current limiting too, and be able to adjust the current limit easily by means of a trimmer.

Jeremy

 

[olaf-lampe]

Go try it!

Jeremy

I probably would, if only I knew what the advantages are compared to the normal speed control mode with current limit?

The other point of concern is, that the infineons current input seems to have multiple purposes. At least the R12 mod seems to influence more than one function...
-Olaf

 

[Jeremy Harris]

The R12 mod is the voltage sense for LVC, isn't it? Maybe the resistor numbers are different on different variants, but R12 on the 6 FET is the lower leg of the battery voltage sense potential divider.

The current sense connection from the shunt goes to two separate circuits; R43 and C20 that is the input to the peak current shut-down circuit and R44 which is the input to the current sense A/D on the controller chip. AFAIK the current sense input to the controller chip only has one function, to limit the battery current by reducing the PWM duty cycle. I'm not aware that it does anything else. Lifting R44 and inserting a circuit to modify the voltage that goes to pin 8 of the controller chip should be fairly safe, as long as the current limiting function can be preserved. It might be possible to just over-ride the current sense signal by injecting a voltage directly to the junction of R44 and C21, but some experimentation would be needed to see how this voltage varies when the controller is functioning normally.

Jeremy