Can I use a mosfet driver as a high current logic voltage shifter

[drakejest]

So i have been scratching my head on my problem of driving a 800Khz signal through a 10 meter twisted pair cable (signal+ground) with an abysmal worse case scenario of 1nf wire capacitance.

What i know works, is that the arduino uno's gpio has the drive capability to do that, i have physically tested this myself.

The arduino unos gpio has a push/sink capability of 40mA, above the rating of what most common voltage shifters. In my circuit there are power mosfets that are being drivern by a push-pull mosfet driver IX4427MTR. I was thinking of using this mosfet driver as a voltage shifter, the capacitance of the long cable should prove no problem for these IC,

I could even use the faster mcp14a witch has better propagation delay.

am i correct to assume i can use the mosfet drivers this way?

[capt bullshot]

It'll work.

Anyway, using RS485 transceivers and differential signalling through a properly terminated twisted pair (plus GND, it won't work without GND) would be the way to go if you need best possible signal integrity.

Next bet would be using the MOSFET driver and properly terminate the cable, either back terminated (series resistance at the "start") or AC coupled termination at the "end".

If you use the cable and driver without termination, you'll experience the signal ringing (due to the reflected waves) that might or might not interfere with your devices operation. In this case, using a driver with slower edges (limited slew rate), as many gate drivers do, is a good idea since this reduces the signal distortion due to reflections. Choose the slew rate as slow as possible to bring your desired signalling rate through.

[Ian.M]

I believe this is related to your other two recent topics:
https://www.eevblog.com/forum/projects/how-to-make-a-800-khz-digital-signal-reach-10-meters/
https://www.eevblog.com/forum/projects/making-my-own-buffer-circuit/

A MOSFET driver is likely to cause more problems than it solves, as driving high current fast edges into your 10m twisted pair, which is a transmission line with characteristic impedance in the 100 to 150 ohm range, *WILL* cause severe ringing if it isn't properly terminated, and as ringing can easily double the voltage at the open end of a transmission line, there is a significant risk of blowing the data input of the first SK6812 at the far end.  Also MOSFET drivers tend to pull very large supply current spikes when transitioning so are a PITA to provide adequate decoupling for.

As you are adverse to using a line receiver IC between your 10m twisted pair you don't have many good options.   One option that might work would be to drive the line (twisted pair) with a 10V level signal source (series) terminated with a matched resistor, and another matched resistor shunt terminating the line at the LED strip, which will attenuate the signal to 5V levels, with a 470 ohm resistor between the terminated line and the strip data input to limit input currents from any residual ringing taking the input outside the rails.   However that's probably over-kill, as it should be possible to drive the line with a slew rate limited 5V signal and avoid most of the ringing, possibly using a dual Schottky clamp at the LED strip end to clean up and residual ringing.  AC termination, with something like a 150pF capacitor and a matched resistor may also be needed at the LED end.

Ground bounce may be problematic if the strip is fed power via the same 10m cable, and the LED supply and line driver share a common ground at the driver end.

[drakejest]

I believe this is related to your other two recent topics:
https://www.eevblog.com/forum/projects/how-to-make-a-800-khz-digital-signal-reach-10-meters/
https://www.eevblog.com/forum/projects/making-my-own-buffer-circuit/

A MOSFET driver is likely to cause more problems than it solves, as driving high current fast edges into your 10m twisted pair, which is a transmission line with characteristic impedance in the 100 to 150 ohm range, *WILL* cause severe ringing if it isn't properly terminated, and as ringing can easily double the voltage at the open end of a transmission line, there is a significant risk of blowing the data input of the first SK6812 at the far end.  Also MOSFET drivers tend to pull very large supply current spikes when transitioning so are a PITA to provide adequate decoupling for.

As you are adverse to using a line receiver IC between your 10m twisted pair you don't have many good options.   One option that might work would be to drive the line (twisted pair) with a 10V level signal source (series) terminated with a matched resistor, and another matched resistor shunt terminating the line at the LED strip, which will attenuate the signal to 5V levels, with a 470 ohm resistor between the terminated line and the strip data input to limit input currents from any residual ringing taking the input outside the rails.   However that's probably over-kill, as it should be possible to drive the line with a slew rate limited 5V signal and avoid most of the ringing, possibly using a dual Schottky clamp at the LED strip end to clean up and residual ringing.  AC termination, with something like a 150pF capacitor and a matched resistor may also be needed at the LED end.

Ground bounce may be problematic if the strip is fed power via the same 10m cable, and the LED supply and line driver share a common ground at the driver end.

Yes my main problem is that driving a signal through 10meters of cable, im exploring options right now


Wont the ringing be solved by using series resistors? just like how you suppress ringinng on the gate of a mosfet im willing to compromise and place a single resistor on the beginning of the led strip.


right now i think this is the closest i have been to solving the problem, and since i already have these componets right here i should atleast test it, Only problem is having no oscilloscope i could not really see the ringing, which might cause damage to the led pixel

[Ian.M]

Without a scope its going to be difficult to check for ringing, so you are going to need to be *very* pessimistic with your LED end line termination.  A BAT54S dual Schottky diode clamping the line to 5V and GND at the LED strip, and 470R between the line and Din to limit outside-the-rails input current would be advisable.  That may not be sufficient to permit reliable comms but will at least prevent damage provided there is adequate local decoupling for the power at the LED strip.

Does your twisted pair cable have a well specified characteristic impedance?

If so, you are in with a good chance, using matched series source termination and matched shunt AC termination at the far end + the previously suggested clamp and 470R resistor.

[drakejest]

 

Without a scope its going to be difficult to check for ringing, so you are going to need to be *very* pessimistic with your LED end line termination.  A BAT54S dual Schottky diode clamping the line to 5V and GND at the LED strip, and 470R between the line and Din to limit outside-the-rails input current would be advisable.  That may not be sufficient to permit reliable comms but will at least prevent damage provided there is adequate local decoupling for the power at the LED strip.

Does your twisted pair cable have a well specified characteristic impedance?

If so, you are in with a good chance, using matched series source termination and matched shunt AC termination at the far end + the previously suggested clamp and 470R resistor.

Yes, twisted pair came from a cat 6 UTP etheret cable, if  iremember correctly from the box it was 47pf/meter.

What if i just ditch the twisted pair and go ahead and use a single 14 AWG wire? will i have worse results than the twisted pai?

[Ian.M]

A single core in a cable will have unpredictable characteristic impedance depending on its proximity to the power and gnd wires and the geometry.  Odds are it will be more troublesome than twisted pair.

The spec for Cat6 UTP calls for a characteristic impedance of 100 ohms +/- 15%.   A 5V CMOS input can tolerate 0.3V of ringing without the parasitic input diode conducting significantly, so unclamped, you need a VSWR of 1.06 or better.   

Matching the characteristic impedance of your cable to better than 6% without a scope or a VNA is going to be quite difficult ...

[David Hess]

Can I use a mosfet driver as a high current logic voltage shifter?

You can, but there are likely better ways.

If you use double termination and single ended signalling over twisted pair, then gate has to drive about 200 ohms which is easy.  The problem then becomes the signal at the end which will only be half the voltage so below the logic threshold and a comparator or line receiver is needed.  There is a clever alternative.

Drive the transmission with *only* series termination.  The 100 ohm series resistor adds to the transmission line impedance yielding 200 ohms again, so easy to drive with a gate, but leave the end of the transmission line unterminated.  Now when the half level signal reaches the open end of the transmission line, it doubles restoring the original level, reflects back, and the reflection is absorbed in the series termination.  This only works when the receiver is on the end of the unterminated line because any taps along the line will see a half level reflection.

[drakejest]

Can I use a mosfet driver as a high current logic voltage shifter?

You can, but there are likely better ways.

If you use double termination and single ended signalling over twisted pair, then gate has to drive about 200 ohms which is easy.  The problem then becomes the signal at the end which will only be half the voltage so below the logic threshold and a comparator or line receiver is needed.  There is a clever alternative.

Drive the transmission with *only* series termination.  The 100 ohm series resistor adds to the transmission line impedance yielding 200 ohms again, so easy to drive with a gate, but leave the end of the transmission line unterminated.  Now when the half level signal reaches the open end of the transmission line, it doubles restoring the original level, reflects back, and the reflection is absorbed in the series termination.  This only works when the receiver is on the end of the unterminated line because any taps along the line will see a half level reflection.

Yes this is actually what i am doing, im waiting for components to arrive to test things out.