my current throttle, will it work?

[dirty_d]

after looking at fechters current throttle i decided to learn how it works and make my own just a little different, i think this should work but i know from fechters experience i will need some capacitors to prevent oscillations, but the basic design is good right? theres no current limit just a zero adjust, i have the resistor values so the throttle range of 1-4V can control 0 to 150 amps. and i think i might need to increase the positive supply to the op-amp because its a lm324 and ive heard it might only be able to output +3.5V

 

[dirty_d]

well i have a problem now, i was going to use the hall effect throttles ground as the ground in my circuit and the +5v for the hall throttle for my +5v source as well, and have shunt signal come from a wire attached to the motor negative wire about a foot down towards the motor, this isnt going to work because the battery negative that is the controllers ground doesnt go directly through to the motor, the battery positive goes directly to the motor and the battery negative is switched so i cant use a shunt without chaninging this a lot, so i guess ill have to get one of those hall effect current sensors and put that inline with one of the motor wires and supply it with the halls ground and +5V

 

[fechter]

You could install it so it measures between the battery and the controller on the negative side. The capacitors in the controller will filter the switching spikes. If you really want to measure motor current, then you need to measure on the positive side as you mention, or use the hall current sensor.

One other problem is; as I recall, a 324 output can only pull down, not up. the output swing limitation you mention may be problematic also.

You might consider another common amp, like a TL082 (or TL081, or 84). Radio Shack used to sell these.

On the inputs, using a 100:1 divider gives you maximum sensitivity, but some amps aren't happy with going all the way to the negative rail.
If using a TL082, change the divider to 5:1, so the inputs to the amp never go below about 1v. This changes the sensitivity, but the amp will have plenty of gain.

Another nice feature of a TL082 is it's so slow that you might not need any input filtering. I used caps to ground on the inputs, but you might also just try a single cap from the output to the negative input on the amp to slow the response. Most controllers also have ramp up speed limiting, so a sudden change in throttle input won't cause a sudden change in output.

 

[dirty_d]

what do you mean by the 324 can pull down but not up, and i had to use those resistor values so that at the no throttle no current condition both inputs are 49mV and at the full throttle full current condition both inputs are 198mV.

 

[fechter]

No, on the output. I think the 324 can't drive the controller input up. You would need to use a resistor to the +5v to pull it up. The 324 would then pull the input down when it was in the low state. One problem with this is if the 324 decides to lose power, the throttle will be stuck high (runaway bike).

I'll have to dig up a datasheet to verify this.

 

[dirty_d]

i checked at radioshack.com they do have the TL082, but im still not sure what you mean when you say the 324 cant drive the controller input up, what exactly do you mean? as for the hall effect current sensor, ill probably use a 50A model, and set teh voltage dividers so the 1-3.5V range of the throttle is matched to the output of the sensor and at above 50A then the sensor voltage doesnt rise and i have the throttle at 4V the output of the op-amp is +5V for ful throttle to the controller.

 

[fechter]

I think the hall sensor will simplify many things. Most importantly you don't have to guess what the resistance of a piece of wire is, so you can predict the current range when designing.

As I recall (I still haven't pulled up a datasheet yet), the 324 was kind of peculiar in that the output stage of the op-amp was an open collector. Inside the amp, the output pin is connected to a transistor's collector and the emitter of that transistor goes to ground. When the transistor is on, it conducts and pulls the output down. When the transistor is off, well, it's just off and there's no current to drive the output high. If you want the output to go high, it must be pulled up with a resistor to the + power.

 

[dirty_d]

ok il use the TL082 then, is it rail to rain and can it go from below 1V to almost 5V?

 

[fechter]

No, it's not r-r. Since it's being used with a hall throttle, the output swing is enough to go from zero to full throttle. The controller typically responds over a range of 1.2. to 4v.

The input is not r-r either. You need to design the circuit so the inputs stay between 1-4v. If you go with the hall current sensor, it's output range is ideal for this.

 

[fechter]

Here's a quick drawing.

There might be some bugs in this as I was in a hurry. Check it out.

The throttle has roughly the same output range as the current sensor, so by choosing the appropriate sensor, you can get the desired range.

If you want a different range, you'd have to scale one of the inputs with a divider network.

The LED makes adjusting the zero possible without having the output engaged. The bypass switch is highly recommended for testing. If something screws up, you can switch back to normal. It's also great for demonstrating the difference in throttle response.

 

[dirty_d]

that stinks, the input has to be 1-4V too? whats a good op-amp where the output can go from Vss to gnd or pretty close as well as the inputs? i wouldnt mind spending a little more, not that i really need it to be that good for this with the hall throttle and current sensor but it would be nice. also should i put a low-pass filter of about 1KHz on the output of the sensor to make sure i dont get much noise from the PWM but not so low as to filter out human input?

 

[fechter]

Sure, they make amps with rail-rail inputs and outputs. If you wanted to make a circuit for a controller that uses a potentiometer throttle, you would need the r-r outputs.

In the "old days" these didn't exist, so designers had to work around the input and output limitations.

Some possible choices are:
ICL7612
OPA2344
MAX4162
LT1490N8

The .1uf cap in my diagram above is supposed to be the filter. I didn't actually calculate a frequency or anything. That's just a SWAG. Based on experience, you can filter the current signal heavily and still get good response. The controller usually limits the upward throttle response anyway (ramp up control).

 

[xyster]

At some point soon, I really need to delve into this stuff so I can follow these conversations...

 

[dirty_d]

i didnt know much about electronics until after i built my bike(a few weeks ago) ive just been reading stuff on the web, i found a really good free e-book, it covers a ton of stuff and explains stuff really good, http://www.ibiblio.org/obp/electricCircuits/

 

[xyster]

Thanks, dirty_d, I'll scope it out.

 

[Eric G]

Hey...Thanks a bunch for the link to the e-book on electric circuits.It sure has a lot more info there than my old high school grade 9 electronics textbook (circa 1970) which I've been using for reference to all things electric...

Thanks Again

Eric