Adding or acquiring controller withsoft start functionality?

[neptronix]

Hey yall.
I know that soft start controllers have mainly gone the way of the dodo, but i really want one for my geared motor to save those poor lil' gears as i am planning on upping the amps again.

So, is there a mod for an infineon ( 9fet ) controller to add soft start?
Or are there such controllers available?

 

[amberwolf]

One way you could do it is to add a resistor/capacitor network on the throttle signal, to force the throttle itself to ramp up slowly (of course, it'll also ramp *down* slower, too). That's the easiest way I can think of to do it.

Resistor goes in series with throttle signal, then capacitor from controller end of resistor to ground. Calculate values based on the time you want it to take.

Might even be a programming option for this, too.

 

[spinningmagnets]

AW, will the resistor/capacitor mod you describe work with RC throttles (a servo-tester) or just standard e-bike controllers?

Either way, I plan to have a standard controller soon (driving a geared hub mounted as a non-hub), and I also want a soft-start. If I send you some money, would you solder one together and take pics of the process? I won't need you to mail it to me, I can make one for myself after I see how you do it, and the parts are cheap.

I plan to have a geared hub drive a Nexus 3-speed, and the sprockets will be ratio'd for torque instead of speed (20-MPH in 3rd gear), and even though the Nexus has a reputation for being one of the strongest IGHs, I fear a sudden throttle twist might snap a gear-tooth...

If you are snowed under right now, even just a drawing would be a great help to me.

 

[amberwolf]

AW, will the resistor/capacitor mod you describe work with RC throttles (a servo-tester) or just standard e-bike controllers?

Because the RC throttle input at the controller is a square wave signal, you'd have to use any kind of res/cap filter at the throttle pot itself, input into the servo tester.

Either way, I plan to have a standard controller soon (driving a geared hub mounted as a non-hub), and I also want a soft-start. If I send you some money, would you solder one together and take pics of the process? I won't need you to mail it to me, I can make one for myself after I see how you do it, and the parts are cheap.

I can post a diagram and some pics if needed, no need for money (but I'll take any that anybody wants to throw at me :lol

I plan to have a geared hub drive a Nexus 3-speed, and it will be sprockets will be ratio'd for torque instead of speed (20-MPH in 3rd gear), and even though the Nexus has a reputation for being one of the strongest IGHs, I fear a sudden throttle twist might snap a gear-tooth...

You can also use a flex-mounted sprocket input into the hub; apparently this is done on some motorcycles and stuff. Liveforphysics and Thud and others can probably tell you more about that, but there have been some posts about it in various hi-power build threads.

It's kinda what I am considering for hte input into my Sachs or SA 3speed.

 

[amberwolf]

Below is the basic diagram for the most simple of the soft-start circuits.

It's disadvantage is that the capacitor takes time to discharge on it's own, based on whatever the resistance from signal to ground is inside the controller. That means that in addition to soft-start, it also kind of has a soft-stop.

In theory you can place a small-value resistor in parallel with the capacitor to discharge it faster, but at that point you wind up with a current drain thru the throttle, the first resistor, and the second one, and you have to ensure that isnt' beyond the ratings of the hall sensor in the throttle to handle. I don't know what the typical ratings are for those things, so I'd hate to hazard a guess as the the second resistor value might be safely lowered to, but it depends on the first resistor's value.


For the first resistor, and capacitor, something high for resistance, like 100Kohm, and low for capacitance, like 50nF, should be viable for a delay that would ramp from 0-full on (assuming you slam on the throttle from nothing) in more than half a second (probably closer to a full second), with the response being faster for the first part and slowly getting closer to full after that.

I have been trying to remember exactly how to calculate out the time delay created, but my brain is not working even as well as it usually does, so not only can't I remember the formula, I also can't figure out how to apply it. It's here:
http://en.wikipedia.org/wiki/RC_circuit#Series_circuit
with the first formula being the one that matters to us here.

There is a simple equation I used to know for figuring out what capacitor to use for what amount of "delay", based on knowing the resistor and the starting and final voltages. But I can't remember it nor find it on the web. I'm sure someone has a calculator out there on a page, too, but I can't find that either.

Some days (or even weeks) I'm just not much use.


EDIT: A much better way to do this is using an op-amp, but again, I can't remember enough right now to put anything useful up.

 

[anpaza]

You might place a diode parallel to the resistor (anode to controller, cathode to Hall sensor output) to get "quick stop" and "slow start". Some low-power diode like 1N4148 will do just fine. A Shottky diode could be better (will discharge the capacitor to lower voltages), but I'm not sure it is required.

 

[Arlo1]

you can also just program the phase current down lower to limit the power off the start!

 

[JohnC]

.

 

[number1cruncher]

I think Arlo has the easiest answer. That's what I did for my front wheel drive bike. Set both the battery and phase amps real low, 15A Battery, 25A Phase.

Takes off slow, has slow acceleration, doesn't climb hills so well without assist, but does make it up to full speed (~40 mph) eventually.

 

[itselectric]

Hey yall.
I know that soft start controllers have mainly gone the way of the dodo, but i really want one for my geared motor to save those poor lil' gears as i am planning on upping the amps again.

So, is there a mod for an infineon ( 9fet ) controller to add soft start?
Or are there such controllers available?

My tech friend who help me to develop this inline softstart throttle device. It is very small, so it can just plug into the throttle line. Initially we made for infineon controller, it works perfectly. When I tried on Crystalyte controller, it didn't work the way it was expected due to differ resistor value, and my tech friend developed the 2nd prototype just for Crystlayte controller. We were think to create a single device with dip switch to between infineon and Crystlayte controller.

Anyway, i am providing the schematic here, if anyone can improve it, please comment here how you would make it better. If you use, just give us the credit for it.

Ken
View attachment 1

http://endless-sphere.com/forums/download/file.php?mode=view&id=60326

 

[spinningmagnets]

Thanks to everyone who has added to this!! I will try very hard to make one of these this summer and to test it out. I don't know much about electronics, but I believe I can also add a bypass loop with a simple on/off switch. With electricity taking the path of least resistance, bypass:ff/open equals soft-start?

This begs the question why can't we use the same electronic speed controllers the R/C community uses?

John, there was a long thread about this, with the definitive answers coming from Jeremy Harris (Thanks, JH!). Adding halls to an RC motor and using a conventional E-bike controller works (such as the common 9-FET), but the other way around...

I don't recall the details, but using an RC-ESC on a full-sized hub just doesn't work, and some people much smarter than me tried.

 

[amberwolf]

I don't know much about electronics, but I believe I can also add a bypass loop with a simple on/off switch. With electricity taking the path of least resistance, bypass:ff/open equals soft-start?

Yes. Just wire the switch from trhottle signal output wire to controller signal input wire, but with a warning--if the cap is large enough value, it's inrush current will last long enough to potentially damage the hall in the throttle, without that inline resistance, if the hall is only spec'd for tiny currents. I don't know what the limits are for the typical throttle hall.

So you might need a switch that is DPDT. Wire the NO contact of one pole to throttle signal output, and Common of that pole to controller signal input. Wire the NC contact of the other pole to the point where capacitor would connect to resistor, and the Common of that pole to the top of the cap. Then if you switch it one way, it disconnects the short and connects the cap, for slow start. Switch the other way and it disconnects the cap and connects the short, for bypass.

 

[nicobie]

I had a couple infineon controllers that were supposed to be soft start and it worked like SHIT with my bmc v2. It was mostly on or off...

If you ask Ilia he will agree that the only controller that really works right is the Clyte analog 35 - 72v one. And it will fry sooner or later over 1600w.

 

[neptronix]

you can also just program the phase current down lower to limit the power off the start!

Done that, already reduced my amps some, now my hill climbing ability suffers as a result.
I would like to run higher amps. It's the hard jolt to the clutch that prevents me from doing so.

It'selectric / amberwolf: THANK YOU for the schematic. I will learn how to read electrical schematics and build both of those.

Nicobie: my controller from cellman has some kind of modification in it for either different timing or higher RPM. Operation is extremely smooth at various voltages and amp levels, i've done 2600w constant ( 43 amps ) with no issues other than a warm motor and a 9FET that's crying for mercy.
The only downside is that partial throttle results in the internal clutch chattering loudly once the target speed is reached. Was that the problem that you had?

If i could get my hands on an analog crystalyte, i would give it a shot.

 

[Arlo1]

you can also just program the phase current down lower to limit the power off the start!

Done that, already reduced my amps some, now my hill climbing ability suffers as a result.
I would like to run higher amps. It's the hard jolt to the clutch that prevents me from doing so.

It'selectric / amberwolf: THANK YOU for the schematic. I will learn how to read electrical schematics and build both of those.

Nicobie: my controller from cellman has some kind of modification in it for either different timing or higher RPM. Operation is extremely smooth at various voltages and amp levels, i've done 2600w constant ( 43 amps ) with no issues other than a warm motor and a 9FET that's crying for mercy.
The only downside is that partial throttle results in the internal clutch chattering loudly once the target speed is reached. Was that the problem that you had?

If i could get my hands on an analog crystalyte, i would give it a shot.

Try messing with the phase amps only and the battery amps only (one at a time) I found that you can lower one and it will take off slow then kick in and accelerate harder at a certain rpm!

 

[neptronix]

Done it.. no real effect.. the max amps from a start and from a hill are the same.. torque kick is the same too.
It's very exaggerated at 15S ( 62v max ).

This controller is a also a bit weird in that, adjusting block time seems to do absolutely nothing, from 0-10, I don't get some kind of temporary bump in amps that other people describe.

Fechter's amp control throttle would make this motor buttery smooth across the board.
I hear that some crystalyte controllers can modulate their amps based on the throttle input. That may explain why the analog controller is such a winner... just a theory.. wish i had one to test.