Harris 92693 ic

You mighty be able to find datasheet over here in one of these catalogs :


What do you mean there is a flyback ?

Design wise the sine frequency to PWM DAC for the meter not a big issue to design. But not a footprint replacement.

Regards, Dana.

sorry, l mean speed closed loop

I suspect this may be a custom / ASIC. The number is unlike our in-the-day part numbering scheme for standard products (I started my career there and left after this part's evident vintage). And in those days we were doing a lot of analog ASIC business that included European.

Not to say that searching "Harris 92693" wouldn't turn up something. Amid the sponsored and irrelevant.

A micro would handle this. You would replace board with a dev board (since this is a one off design).

Some design work would have to be done to manage the speed control, especially safety issues. The
meter part of the design is a fairly simple solution. Just a comparator onboard processor preferred
to square up sine.

Hall sensor I2C or V output ? Either way processor with A/D or I2C interface would handle the interface
to sensor.

Is DC motor a servo or stepper ? How is speed control done in truck, fuel intake ?


Regards, Dana.

danadakk said:

A micro would handle this. You would replace board with a dev board (since this is a one off design).

Some design work would have to be done to manage the speed control, especially safety issues. The
meter part of the design is a fairly simple solution. Just a comparator onboard processor preferred
to square up sine.

Hall sensor I2C or V output ? Either way processor with A/D or I2C interface would handle the interface
to sensor.

Is DC motor a servo or stepper ? How is speed control done in truck, fuel intake ?


Regards, Dana.

Click to expand...

thanks for Harris databooks link but as mentioned above this is probably custom number and there's no 92693 number in databooks, not sure about hall sensor - sensor output voltage is 8.5 volts, the reason could be damaged 92693, this is DC brushed motor, this is military truck, no speed control.

danadakk said:

comparator onboard processor preferred
to square up sine.

Click to expand...

l had some experience with replacing sine wave speed sensor with square wave on onother truck manufacturer and it worked, l mean signal form can be not an issue for micro or IC

The below would be a simple way of getting a solution. Single chip. Basically a freq cntr
driving meter thru DAC.

The Output can be a current or a V DAC, user choice.

Inpuit is signal conditioned comparator with a small amount of hysteresis.

LCD optional if you needed it. LCD itself is off chip, but its control onchip.

Family to use would be a PSOC 5LP, ~$15 board would be CY8CKIT-059.





Small amount of code to convert the freq measured by the FreqCntr to the drive V or I to meter.

Note onchip wave generator to gen a sine for testing. Also note most chip resources remain for other stuff
in case you wanted to do additional features.

The IDE (PSOC Creator) and compiler free.


Regards, Dana.

I managed to capture a signal from speed sensor and it is definitely not sine, in this article was described these are sine wave sensors http://www.dinoplex.org/instruments/vegliaspeedosensor/index.html

If this control scheme is chopper frequency based (like something with a PLL in it, or just a crude f-to- v and an op amp) it may work better with square than sine input. Figure there may be squaring at the front end inside.

RF CMOS PLLs tend to have a cap blocked auto-centered input that can work with either, though faster edges make less PN / jitter.

The processor can handle the averaging to remove what appears to be jitter > 10%,
quite significant.

By the way, a vehicle has load dump issue that has to be handled by any solution
you use. TVS solutions good way to use.




Regards, Dana.

Let's think simple.

If you have a trigger edge and you have a fixed width output pulse, then you have duty cycle modulation, just at varying frequency. This is PFM, an alternative to PWM and common in low cost power supplies and resonant types both.

If the electromechanical movement is DC current operated then you might just square up that sensor to bang-bang (CD4009B could handle 18V but not much more, filter & protect) - and custom ICs are no more reliable than their designers made them, with many newbies put to the task in my days as an ASIC "godfather" - maybe this little piggy has a lot of brothers already bacon....

Hex inverter (one for auto-bias, use the single stage type), one-shot (CD4098B?), L-C-R filter and there's your frequency dependent DC current for a torque type movement.

Maybe the clutter is part of the problem. Can you deduce or find the movement's electrical interface attributes to work from, cut out all the middlemen, go from sensor pulse to needle swung in 2 cheap ICs of a size human hands can solder (SOIC-14 or DIP, where does this cobble have to fit?).