Motor speed control via back-EMF detection

Phil Pemberton · 2008-08-31T23:35:11+00:00

Hi guys, I've been tearing apart a Brother PT-1000 label printer with the eventualgoal being to connect it to a desktop PC to print labels for my variouscomponent storage boxes. I've figured out how the print head communicates withthe controller board, which just leaves the motor drive circuitry. Naturally, being a low cost device (£15), the PT-1000 doesn't use anythingremotely stepper-motor-like for the label feed. Instead, it uses a cheapMabuchi DC pancake motor (an RF-300C-11440, for which I have yet to find adatasheet) and a speed reduction gearbox. The speed control is performed by aROHM BA6220 chip. Ideally I'd like to eliminate the ROHM chip (seeing as it's basicallyunobtainium) and use a PIC of some description (probably a 12F675) to do PWMspeed control of the motor. But first I'd like to understand how the ROHM chipmanages to do what it does. I found a datasheet for the '6220 here: .. but like most ROHM datasheets, it doesn't say much about the chip, otherthan that it uses back-EMF sensing and how to determine one of the twoexternal resistor values. What I don't get is that there's no obvious way for the chip to senseback-EMF. Everything I've been able to find about bEMF sensing suggests thatit's normally used with PWM control -- the motor is powered up for a shortperiod of time, then in the off period the voltage across the motor is sampledand used to (roughly) determine the motor speed. Unless it's sensing current,but if it is, the "application circuit" (BA6220 datasheet, Fig. 2, page 2)doesn't look like any current sensing circuit I've ever seen. In fact, itlooks like a voltage comparator, but the polarity of the voltage referencedoesn't look quite right... I've hooked the scope up to the driver IC's pins (and the motor itself) anddidn't see anything that suggested the driver IC was switching the power onand off. In fact, the voltage remained more or less constant, excepting the~50mV sine wave (plus one ~200mV spike per cycle)...

Aug 31, 2008 21 Replies

Rate this thread:

Philip Pemberton 17 years ago

LOL!

The only reason I'm using a PIC micro and a FET is because both of those items happen to be sitting in my junkbox gathering dust. The 8-pin PIC12F615 is 50p in one-off for DIL, or 45p if you prefer the SOIC package. Total parts cost including a small FET is probably about 75p (or US $1.50)... hardly worth worrying about. The PIC has an internal oscillator, PWM, A/D and a 20mA output stage -- if you disable MCLR, you don't even need the external reset pullup resistor.

But yes, it's overkill... Not on a Heath Robinson or Rube Goldberg scale, but still overkill.

Thanks,

Phil. usenet08@philpem.me.uk http://www.philpem.me.uk/ If mail bounces, replace "08" with the last two digits of the current year.

Vote

Joerg 17 years ago

I wonder if TI ever reaches down into those price categories. They'll have to, since Atmel does it as well.

Still, I'd have done it sans micro. Just for kicks.

Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.

Vote

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required