fechter said:
An Audrino or other MCU would certainly be a viable route also, but I'm not up for learning to code right now. I'd use an Atmel if I was going that route (cheap).
I'm using an Atmel atmega168: It's pretty standard with arduinos(The arduino is the not-so-cheap part for JUST a throttle signal converter, but for learning and being introduced to microcontroller programming and that realm in general, it seems so invaluable.). Without an arduino, it seems you'd have to use existing c libraries and possibly assembly and that involves some mental gymnastics, but the arduino has simplified the programming so much with its own "interface" and libraries that I'm sure you could easily get upto speed if you wanted. I mean, to write a HIGH voltage to an output pin, it's like....
setup(pin_number, OUTPUT); (only needed once in the program)
digitalWrite(pin_number, HIGH);
EASY.
To implement current limiting, there would need to be a current sensor of some sort. A shunt would be the cheapest, but I don't know where to get that resistance wire. I could also use an Allegro hall current sensor, but those are expensive. Using existing wiring works as a shunt, but is difficult to calibrate.
Anybody know where to get shunt wire?
I don't know about shunt wire. The shunts I've seen seemed pretty dang expensive for just being a metal, but I think that's because they had a pretty excellent temperature coefficient... Like the resistance only changed .01% per degree celsius.
I've seen Allegro sensors for around $5 on digikey. That isn't too expensive compared to the $10-30 shunt resistors, is it? Finding some cheap, but excellent shunt material seems to be beyond me besides internet searching but that doesn't always help in finding the BEST deal available(Price and product effectiveness combined) or some approximation. Even eBay didn't seem to provide anything more competitive for shunt wire.
To do overtemp limiting, you'd need to decide on a sensor. It would be easy with a LM335 or similar if you have room for a TO-92 package as the sensor. You would also need to decide on the maximum allowable temperature and the foldback profile.
Noted. I'm looking for some usable temperature sensors. Does the Lm335 stay accurate with use and does it output a simple, usable output voltage? The ATMEL microcontrollers have analog sampling pins that can read voltage values with ease. But, why is a sensor needed? I can imagine it'd be nice to make sure it stays precise and the current limit value known, but I can also imagine limiting the maximum pulse width based on a trimmed-voltage value would work to limit the current... you may not know how exactly much it's limited(unless the current is measured somehow else...), but it's limited and with testing, it can be "good enough". But, yeah, that's kind of a discomforting situation when not good enough could mean busted.
Also, does anyone know how picky RC controllers are about the servo PWM frequency? I think the standard is 20ms frame time, but what happens if that varies?
I found a source online that stated anywhere between 17-23 ms is fine, but going too slow or too fast may either reset it or "confuse it" and lose sync. I would think better controllers, like the Castle Creations ones, would have better tolerance.
Will the RC controller be able to calibrate to the throttle range or does the servo tester need to provide that adjsutment?
Seems like that differs from RC controller to RC controller. The Castle Phoenix auto-calibrates to the throttle by requiring WOT at startup for upwards of 4 seconds, and then dropping to no throttle. Assumably, cheaper controllers just assume 1 ms as 0% throttle and 2 ms at 100% throttle and do some kind of line/curve in-between(possibly linear, exponential, or etc.), but I'm getting clarification on that at the RCgroups forums.
So... how about a little box that has a connector for a standard hall effect throttle, a connector for a temp sensor, a current limit adjusting knob, output to the controller throttle and a built-in 5v regulator that can take over 100v in?
Sound like I might need to round up some more guinea pigs...
I'd be interested how this format would be received. Quick question, though, how do you think the current sensor would be connected to the battery line? It sounds like all these sensors, current and temp, would complicate the setup process beyond simple "plug and play"... well, perhaps the more advanced features can remain as "optional" in the software, so you could have it either way(and it'd be automatically detected - or these modes could be manually set by small switches.). Ultimately, though, at least a few connectors will have to be adapted somewhere as there doesn't really seem to be a universal standard with throttle, controller, battery connectors and temp sensors.
