Originally shared by James Kasper - CBWP how many of you are familiar with

Originally shared by James Kasper - CBWP

how many of you are familiar with the ultimaker?

I have been comparing the schematics to the ramps 1.4 and there are a lot of pull up resistors not used in the ultimaker. Are they using the internal pull ups of the mega? enable pins on the Pololus, the limit switches…
They use the same mosfets for the heaters but on the ramp1.4 schematic, there is a 10r resistor inline between the digital pin and the gate. is that to absorb a surge? what is 10r? misprint on the schematic? it wasn’t there on accident. On the ultimaker schematic there are pull up resistors for the 3 multi-stepping jumper selections they are not on the ramps… the value on the print is 4k7 Is that supposed to be 4.7k?

Is the software defaulted for internal pull up resistors? (Marlin) or is there a config setting for it?

10R = 10 Ohms. The R to make sure you didn’t leave off a K and to provide a decimal point. Compare 4R7 and 4K7.

(And yes, 4K7 is shorthand for 4.7k)

thanks, I figured, but
I didn’t want to fall in the assUme trap like usual

any other answers?

Sorry, the other ones I can’t help with.

np thanks.

@Vik_Olliver :wink:

Dunno about the Ultimaker specifically, and I’ve only just laid hands on my first real RAMPS1.4 board - embarrassingly, it’s not going yet. But 4K7’s I recognise as usually what you pull up a thermistor with. The 10R is a horrible, horrible abuse of the CPU’s 40mA output limit, allowing the chip to drive the MOSFET gate at well over 100mA and I’m surprised more people don’t have problems.

So should it be there or not? What do you use? It id there for the ramps, but not the other. Which boards do you use?

Oh yeah, it should be there. It limits the current draw to a mere 140mA in combination with the internal Arduino safety electronics. I use TIP122’s and a 1K resistor instead for my heaters. Just keep the PWM rate low and they work awesomely. If you have a metal frame, use a copper “saddle clamp” from the hardware store to attach them to it as a heatsink. Put teflon tape in as an insulator tho. Live frame bad idea.

I have http://www.adafruit.com/products/355 IRLB8721

@Vik_Olliver Folks don’t have more problems because the ST55 MOSFETs won’t draw 140mA through the gate, more like 100nA. I agree that the 10R is inadequate to actually current limit the outputs (not much better than a shunt–your 1K0 would cost the same and limit to a non-damaging 5mA) but it would be protecting from something other danger than the MOSFET gate.
(edit: Oops. that’s all at steady-state: 120R would limit the transient current to 40ma for ~2.5uS rise-time)

The 10R is to control ringing.

Too low a resistance and the output current required to switch the gate capacitance goes up, and you can get ringing from the gate capacitance and parasitic inductance.

If you have too high a resistance the turn on/off times go up and the switching losses increase.

You occasionally see a diode in series or parallel, to control the turn on vs off times separately.

So what is the recommendation for the mosfet?
@Adafruit_Industries did not have anything inline to the gate in their tutorials. (Also note vik is driving darlington transistors, so I could imagine a 1k in there to reduce base current.)

I just never personally designed a mosfet circuit. They seem better than a transistor for switching loads, but I seem to remember the darlington being real fast like the schmitt trigger. So it is probably made to be more of a power switcher than the audio amplifiers and basic 2222 npns that I used to play with.

@Qitian_Dasheng_Sun_W what kind of frequency does that start appearing? Must be pretty fast. What is a standard pwm rate for an extruder heater for example? The pid (i havent looked it up yet) shouldnt allow it to oscillate too quickly should it?

You can get ringing after any single step change, in practice it should only be a problem with sloppy wiring or long track runs. Marlin switches the heater around 7Hz, IIRC.

Disclaimer: I’m not an expert. If there’re any EE’s around, speak up.

Rough rule of thumb is the RC time constant should be 3-5 times the sum of the turn-on & turn-off times. For that MOSFET it would be
3 * (93ns * 17ns) / 1077pF = 306R.
Personally I would err on the low side as you don’t want the turn-off time too long, I’d try 270R and check it with a scope.

The lowest PWM the ATMega’s drivers do is about 60Hz-ish.

Yeah, my 1k is to reduce base current. I like me old, reliable Darlingtons :slight_smile: