Page 1 of 1

PWM & QEi Signal - cable Length

Posted: Wed Jul 05, 2017 12:50 pm
by Docara

I'm looking to drive a DC motor (via a motor controller IC) with an encoder feedback, and I was wondering how long, practically speaking, could the (shielded) cable lengths from the processor be?
I need about 8-10m


Re: PWM & QEi Signal - cable Length

Posted: Wed Jul 05, 2017 1:26 pm
by Benj
Hi Matt,

8-10 meter should be possible but you will probably have to keep the signal frequency as low as you can get away with and also maybe look at transmission line theory, you might need to balance the line by adding terminating resistors and capacitors at the far end of the cable. I take it the driver will be at the motor side of the cable to avoid significant losses due to cable resistance.

Transmission line theory takes me back to hours in a college lecture hall being bored out of my mind and wonder where I would ever apply it. Wish I had listened a bit more closely now :wink:

Re: PWM & QEi Signal - cable Length

Posted: Wed Jul 05, 2017 2:46 pm
by Docara
Ha Ha Yeah reminds me off my tech drawing days at college - gawd I wished I listened.

Oh Yes!!! - reference motor driver

Whilst it is not strictly twisted pair my screened 1pair signal cable does have a comparable twist in it (cough cough) I was hoping the screening would be enough in an automotive environment, connected to a separate earth that I would get away with it, but slightly concerned about capacitance. Do you think unscreened CAT5e would be better.

"transmission line theory" - stop swearing I'm a customer :D :D - I was considering RS485 transmission line but don't know if I could bang PWM down it reliably.


Re: PWM & QEi Signal - cable Length

Posted: Fri Jul 07, 2017 12:20 pm
by petesmart
now I'm interested... being and old RF engineer... :wink:

so lets start with the basics...

whats the frequency of the PWM signal?

do you know what the output impedance of the PWM driver?

what is the required voltage level at the PWM (motor controller) input

what is the frequency of the Qei signal? do you know the impedance?

data sheets should provide this info

once you know this info you can calculate the length by cable type selection



Re: PWM & QEi Signal - cable Length

Posted: Fri Jul 07, 2017 2:29 pm
by Docara
Hi Pete,

Bugger! I've been humbled (and probably rumbled) :oops: All good points um um err Its 32DegC in London at the moment um um it might rain tomorrow um um

Seriously, yeah you're spot on requesting the info you have. I've not decided on the specific components yet and not fully committed to a servo, stepper, dc (+feedback). Depending on the size of motor I would be looking at the same drivers as I mentioned here ... 75A#p82469 located within 1-2m of the actual motor but the PWM signals from the uP would be around the 8-10m as previously mentioned.

From the datasheet - Taking the DRV8871 device as an example . . . . . .DC motor with 6.5 to 45 V and up to 3.6A peak current.......Two logic inputs control the H-bridge driver, which consists of four N-channel MOSFETs that have a typical Rds(on) of 565 mΩ (including one high-side and one low-side FET). Motor speed can be controlled with pulse-width modulation, at frequencies between 0 to 100 kHz./b]

I can see no value for input impedance only the output impedance above. The datasheet also mentions a maximum PWM frequency of 200KHz. Logic level inputs I think are CMOS TTL 0-5.5V and in the range of 0-35uA

So assuming a given output motor (12v) will be driven, and I use a DC motor with limit switches for simplicity. What do you come up with?

Now I've never used and encoder. and if I'm honest I was only entertaining the idea because FC has a macro built it. Are you able to give any real world insight into using these devices?