FREE - RC Heli Optical Tachometer DIY

[TheBum]

Originally Posted by HELINHAWAII

I'm having trouble finding a 2.5K Ohm precision pot. Will any 10 turn pot do?

It should. I'd make sure to mount it sideways or drill an access hole so that it can be adjusted without removing anything from the case.

[DominicD]

Panel mount pot is what you really want.

[jared_of_atlanta]

It would be sweet if you would post links to all the parts your finding.

[DominicD]

Mike You might want to be clear about the pot taper... linear ?

[Finless]

I would think linear taper as an audio tapered pot would suck in this application I would think.

Bob

[DominicD]

I would think so to but iam not the designer so its not my call.

[HELINHAWAII]

Thanks man! I'm kinda dumb regarding this stuff... I'll post a parts listing from DigiKey or something when I get it ordered.

[DominicD]

Mike if you don't mind can i make a few suggestions. Get rid of the pots and resistors. Drive the motor off the a PWM from the PIC. Let the PIC automatically sweep across the whole range of RPM. When the blades become stationary in the eyeslit the user hits the button which records the RPM.

[jared_of_atlanta]

I'd rather use the pot--even if it costs more. Using a button while it's cycling would be like trying to stop a stop watch on x:00:00 (a fun game but not useful).

[HELINHAWAII]

Originally Posted by HFG

Mike if you don't mind can i make a few suggestions. Get rid of the pots and resistors. Drive the motor off the a PWM from the PIC. Let the PIC automatically sweep across the whole range of RPM. When the blades become stationary in the eyeslit the user hits the button which records the RPM.

Oooohhhh, that'd be cool!

[Finless]

Hmmm you will still want to allow users to control the sweep rate as well as direction. Even using my skytach I usually wind up going past the blade stop and have to back up a little.

Bob

[DominicD]

The problem with tuning the pot is the delay caused by the rotating mass of the shutter. Your constantly overshooting the target RPM (aka severely under damped). When you sweep and hit the button it will record the RPM at that time which is the exact target RPM. Let it take 20-30sec to sweep from 1000-3500 RPM, will be more than accurate enough.

[jared_of_atlanta]

I would think the delay is the 1.5 second counting period.

[DominicD]

Thats a long time, if you have to wait a full second every time you adjust the knob your gunna need a new lipo or refuel the tank before you get a reading

[fireup]

Originally Posted by HFG

Mike if you don't mind can i make a few suggestions. Get rid of the pots and resistors. Drive the motor off the a PWM from the PIC. Let the PIC automatically sweep across the whole range of RPM. When the blades become stationary in the eyeslit the user hits the button which records the RPM.

The problem with using the PWM in open loop is there is not very much resolution.

[TheBum]

Originally Posted by HFG

Panel mount pot is what you really want.

You're right. I wasn't looking at the documentation at the time and was thinking HELINHAWAII was referring to a trimpot. My brain's refresh clock stops occasionally.

[DominicD]

Fireup who runs anything in open loop ?!?!?!

[mikef]

You can use any pot. I have a higher precision pot (turns 25 times or so) and it allows you to adjust the RPM easier. A cheap 1 turn pot makes it real hard to dial in the exact rpm ( I tried this).

You do not have to wait 1.5 seconds to adjust the speed. That is just how long the CPU counting time is and it sounds much worse than it really is. I have not had any trouble with waiting or dialing in the RPM. You do not “need” the memory button and I rarely use it since the tach is very stable and does not overshoot much due to the rate the wheel gains speed. The RPM gets up to speed much faster than the Min. Aircraft tach does so there is no worry about waiting for anything to catch up. I basically use the tach like a camera….you just look through the eye piece and adjust in realtime until the blades stop just like any other tach. When you get it locked you just look down at the screen and it will be rock solid on your RPM. It’s fast guys so don’t worry about any delay or the 1.5 second counting speed. The memory button hooks to RB.0 which is pin # 6. Just connect the switch to the +5V so that when you press the button, the +5V will go to the #6 PIN on the PIC Chip. I am willing to bet that the memory button will get used very little. I’ll update the schematic.

HFG – Your PWM idea sounds good but to be honest I have little experience with that part of the PIC chip. I need to play with that some day or you can send me some info on it. I have played with servo control a little bit but controlling this motor is a little different isn’t it?

Mike

[DominicD]

Yes controlling the motor is different than a servo. Basically the best way to do this is with a closed loop controller. Connect your little IR sensor into the Schmitt trigger then thats feed into the PIC, BUTTT you have to use an interrupt NOT polling the input. The interrupt will be used to count the frequency. You can use anything as simple as a P, PI or PID controller to control the closed loop characteristics. The PID controller will then control the duty cycle of the PWM output. Most of the PICs have a built in PWM controller so its just a matter of setting up values in the proper register.

Its VERY easy to do a basic PID controller in C, you gotta stop writting code in ASM (I use to always write stuff in ASM but just recently ive started to learn C for microcontrollers and i HIGHLY recommend it)

Here is the pseudo code i ripped off of wiki

previous_error = 0
start:
error = setpoint - actual_position
P = Kp * error
I = I + Ki * error * dt
D = (Kd / dt) * (error - previous_error)
output = P + I + D
previous_error = error
wait(dt)
goto start

[mikef]

oh man Assembly??? I have no clue how to program in Assembly. I program in PICBASIC and then cross compile the Assembly and HEX file. I originally played with the Basic Stamp in college and then moved on to the PIC chip because they are so cheap. The program “PIC Basic” was so close in language to the basic stamp language that I easily migrated over to that. I have to keep it simple for my simple mind.