24V PSU for printer

Quote
anvoice
The IRFP4368 will thus be getting about 10V and, being an N-channel mosfet, will be on when the heating signal (now the triggering signal) from the RAMPS is on.

Wrong, it will get 12V (if you are powering the RAMPS from 12V) when the RAMPS heater output is off, and 0V when it is on. The RAMPS switches the negative side of the heater - just like your mosfet does.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

Quote
anvoice
The IRFP4368 will thus be getting about 10V and, being an N-channel mosfet, will be on when the heating signal (now the triggering signal) from the RAMPS is on.

Wrong, it will get 12V (if you are powering the RAMPS from 12V) when the RAMPS heater output is off, and 0V when it is on. The RAMPS switches the negative side of the heater - just like your mosfet does.

Not wrong, because as I said multiple times and have in my circuit diagram, I'm powering my RAMPS with a 24V PSU. I don't know what you mean by "switching the negative side of the heater", but the voltage on the D8 positive pin will be about 24V above ground when the RAMPS attempts to heat, and 0 when it's off, so it will correctly switch the NFET on and off. For this same reason the relay worked and heated the bed when I tested it with my 12V power supply, just that it got way too hot even then and burned out at the 24V because it was defective.

Edited 2 time(s). Last edit at 04/24/2015 08:01PM by anvoice.

The voltage on the +ve pin of output D8 will be +24V all the time, because it is connected to the +ve heater power input pin via the 11A fuse. The voltage on the -ve pin of D8 will be Ov when the heater is on, and the pin will be essentially.open circuit when the heater is off. The heated bed mosfet connects this pin to ground to turn the heater on.

Also be aware that most RAMPS boards won't take 24V input except on the heated bed power input, although there are a few that do. Same for the Arduino.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Right, I see what you mean. That I did not know, and it looks like I'll need to rethink my circuit... Probably will have to add a P-channel transistor to be triggered from D8- and in turn trigger the power FET. Would have gotten a P-channel power FET if I knew that ahead of time.

The RAMPS board does take 24V btw, assuming the capacitors and the small 5A polyfuse have a 30V rating. Just need to replace the 11A polyfuse (16V) and remove D1 which feeds the arduino (unless you have a Taurino power) so 24V don't blow it. Perhaps it's not as safe and something could overheat, but it does work.

Edit: the modified circuit would probably look like this:

I could replace the NPN and N-channel FET with a P-channel FET if I can find one. I see a problem with this configuration though: if PSU1 fails, the D8- pin will go low, which means the power FET will remain on and heat the bed indefinitely. I may have to power the heat bed input on the RAMPS from PSU2 to avoid that.

Edited 1 time(s). Last edit at 04/25/2015 06:39AM by anvoice.

In that circuit the emitter of the npn transistor needs to go to the negative side of the power supply, not the positive side. You also need a resistor from the collector of the transistor to the positive output of psu2 (2.2k 0.5W will do), and another from the base of the transistor to the positive side of psu2 (e.g. 10k). With those in place, failure of psu1 should not result in the heater turning on.

Although a p channel mosfet would in principle be simpler, they have higher Rds(on) then their n channel counterparts so are less suitable for high current loads.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks for pointing out the circuit errors, I'll fix those.

Is there a way to reprogram the arduino to invert the logic of the D8 ve- pin? That would be the perfect solution with no extra parts.

Yes you could change the firmware to invert the PWM, but then the heater would run at full power if PSU 1 fails, or if the Arduino fails to start up for any reason.

Is there a reason you are using two PSUs instead of just one, given that they are both 24V?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

In general, I need 2 because I can't find a single PSU (reasonably priced) with enough wattage to power the electronics + heated bed (needs something like 700W total). However, I'm now thinking that since D8 is powered by its personal power input on the RAMPS I should be able to connect PSU2 to that input and thus eliminate PSU1 from the circuit completely, since the extra power drain on PSU2 should be minimal due to the D8 signal. Still doesn't solve the heater staying on if the arduino doesn't work properly though.

What about defining bed heating on another pin and using a positive signal from that to drive the circuit? Sounds like that would be simpler if doable, just need to figure out which of the pin files I need to change (can't find a RAMPS 1.4 pin header file in Marlin, only some RAMPS_old one and a RAMPS_13 one). Basically the root of the problem seems to be that the originally positive logic from the arduino D8 pin is inverted by RAMPS to produce the D8 ve- signal, if I bypass that problem solved.

Edited 2 time(s). Last edit at 04/25/2015 04:31PM by anvoice.

An alternative is to tap in to the pin that drives the bed mosfet gate on the RAMPS, and drive your own N-channel mosfet from that as well (or instead). But then you will need to use a mosfet that works reliably with 5V gate drive.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Right, or I could drive a NPN transistor driven by the 5V and have it switch the 24V from D8+ (reduced to 10V by Zener) to drive my current N-channel FET. Seems like that would take care of all the problems, including the safety issue of bed staying on.

Quote
anvoice
Right, or I could drive a NPN transistor driven by the 5V and have it switch the 24V from D8+ (reduced to 10V by Zener) to drive my current N-channel FET. Seems like that would take care of all the problems, including the safety issue of bed staying on.

No, you would get an extra inversion of the signal if you do that.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

Quote
anvoice
Right, or I could drive a NPN transistor driven by the 5V and have it switch the 24V from D8+ (reduced to 10V by Zener) to drive my current N-channel FET. Seems like that would take care of all the problems, including the safety issue of bed staying on.

No, you would get an extra inversion of the signal if you do that.

How so? A positive 5V (heat on) will switch the NPN on, which will in turn switch on the FET. 0V will do the opposite to both transistors, so no reversal. The NPN just acts as an amplifier in this case.

To have the npn transistor switch 24V from a 5V input with common ground, you have to use the common emitter configuration, which inverts the signal.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

As much as I look at this it seems that a positive 5V signal will switch everything on, and 0V will switch everything off. Otherwise I'm quite confused.

Edited 2 time(s). Last edit at 04/26/2015 05:34AM by anvoice.

Common newbie mistake. That circuit will only feed about 4V to the gate of the mosfet.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

So the PNP transistor would eat about 21V? Could you explain why that would be, goes against what I've learned (granted it's little) so far. What if I use something like another smaller N-channel power FET instead of the PNP?

Might not be exactly what you're going for... but... my printer(that i'm currently hacking to Duet+x4) originally had/has a 400w 110AC chamber heater.

110v input
Schaffner line filter all power for the machine runs through it(the people that built it used an aluminum plate for common/ground fyi)

2 leads are connected to that ^ (+110v & -110v)
+ lead runs through a board with a relay and...

2 leads direct from + on a 12-24v power supply(i haven't measured)

On the relay board (WATCH OUT FOR TECHNICAL TERMS!!!)
there are two hot dogs(resistors?) one with colored stripes, one is black with silver stripes, one round brown/orange disks and the relay

2 wires run from the relay board to the old controller. (low voltage... 22g'ish wire)

just one solution... If you want more info or measurements let me know...

I have a similar setup, except that I use a SSR-10DA solid state relay to control the 230V 350W bed heater from the Duet, and no line filter.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
Preferably, fit an over temperature cutout

That's sound advice worth repeating. Unfortunately it's a lot easier to find thermal fuses with current ratings that would suit a mains powered heat bed than it is to find them for a large low voltage bed.

I know about [ ] this much about electronics, but this doesn't really sound like a great idea... driving a huge bed from DC 12/24v. seems very inefficient.

Why not use a silicone heater the way they use in Tshirt presses. They are readily available and so are the controls. Unlimited size too

You might say one is safer than the other (ac/dc), but both will burn your house down.

Edited 1 time(s). Last edit at 06/13/2015 09:26PM by thetazzbot.