Catalyst Circuit Questions

I've been trying to figure out the circuit diagrams provided for the extruder of profounddevices' Catalyst. The diagrams are a bit of a mess, and I'd like to get an opinion. Here is my take on it:





The originals are at:
version
version

Thanks
--Kris

Kris,

I was trying to make sense of his diagrams. I finally decided to throw out his ideas and go with more standard reprap design. James did not provide me with all the parts to do either of his diagrams. I bought all the parts from Digikey that I needed for a little over $5 including shipping. I got the heater and temp sensor working yesterday. I will post the info later today.

Buzz

Here's a more comprehensible version of the schematic. I did add one rather essential part though: a diode (parallel to the relay coil) which protects the transistor from flyback currents.

See attachment.

Success!

Thanks, that makes sense. I'll give it a try tonight. Clearly I didn't know how to wire the relay. I'm going to stick that diagram into the wiki.

--Kris

wouldn't it be a lot easier with a mosfet? I use the IRL3803 for my heater, much simpler than this- arduino pin -> gate, gnd -> source, heater -> drain.. no other parts needed

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

wouldn't it be a lot easier with a mosfet?

Depends on what you have laying around ;-)

The big advantage of using a mosfet is that you can use PWM for precise power control.

Can anybody confirm the schema of Reinout. I'v tried this and indeed it heats up.. but is so noisy..

so my question is

A: is that normal or what might be wrong. the relay is not clicking once or twice but at a rather high frequence!!!

B: How is the easiest heater circuit build up.. just using the arduino + a small handfull of common electronic parts?



I all heated but it didn't stop at the given temp.. !?!?!

Ouch...

this is going to be very hard to debug via email/messageboard.
Your best bet it to pair up with someone who has some measurement equipment (and knowledge) available in order to home in to the source of the problem quickly. Perhaps there is a HackerSpace near you, or a college where you can ask a teacher to hook you up with an electronics enthusiast.

There are various ways the symptoms you see could arise, for example
- incompatible software,
- using an AC power supply where DC is called for,
- a poor power supply (wher activating the relay affects the temperature sensor readings),
- a wiring error
and of course I could have made a mistake in the circuit diagram (but i don't see any right now).

Probing your circuit with an oscilloscope should get you out of the frustration zone the quickest I expect.



HTH,

R
-

hmm I took out the led and then the heater starts heating a lot faster and it is silenced!!! Just a slight hum...


The led was lit slightly before removing .. so there is some current going through pin 13.. on times not needed I guess

Edited 1 time(s). Last edit at 04/29/2010 05:55AM by mhensen.

Great!

does it stop heating when at temperature now?

Nope not yet .. I found that I made a mistake with the transistor.. it was the wrong way :-( I am ashamed.. but then it still didn't work I hear it ticking at reset..

So now I've written a small routine that puts pin 13 high and low.. nothing.. Then I replaced the transistor for another.. ( don;t know which one.) and now the relay is clicking like I want it to be..


Now I have to resolve the thermistor routines in the firmware and set it to the pin 13 high when temp reached.. Looks like that one is corrupt in my firmware..

But now I have one more question.. Can I replace the 10 ohm variation of James with a 6 ohm makerbot variant without having to replace the hardware? I am not that good in calculating the currents and flow in a electronic circuit...

Thanks to all so far!!

Great hacking! Brings back old memories...

The formula you need is called 'Ohms law'

   V=I*R  
(tension in Volts = current in Amps * resistance in Ohms)

So assuming a 12V power supply you would get:

Old situation: 12V = I * 10 Ohm ==> I = 1.2A
New situation: 12V = I * 6 Ohm  ==> I = 2A.

So the relay contacts should be able to handle 2 Amps, and your power supply should be able to handle 2 Amps plus whatever current other heaters/motors take from it.


The power of the heater can be calculated with:

   P=V*I 
(power in Watts = tension in Volts * current in Amps)

Old: 12V * 1.2A ==> 14.4 Watt
New: 12V * 2A   ==> 24 Watt

Success!

R
-

I love to hack around in this although I don't know too much about it..
but with some logics I manage to get by:-)

bTW what is missing in the diagram is the common ground..

Arduino is self powered en the rest of the circuit is powered by another source.. And they need a common ground :-) That was something I was missing too.

So for everyone thinking about to make this.. attach the 2 - ..

I hope this is allowed :-) but it seems to work :-)

thinking about the V=I*R and P=V*I .. the lower the R the faster it heats up.. but can the thermisistor keep up with that??

I think I'll go and search the values I need :-)

> bTW what is missing in the diagram is the common
> ground..
>
> So for everyone thinking about to make this..
> attach the 2 - ..

Yup, this is hardly ever drawn explicitly in circuit diagrams.
When you see ground symbols they are all connected. Ditto for power symbols, when you see multiple +5V power symbols they would be connected to each other too.


>
> I hope this is allowed :-) but it seems to work
> :-)

Keep them connected, it is essential for proper operation!

> thinking about the V=I*R and P=V*I .. the lower
> the R the faster it heats up.. but can the
> thermisistor keep up with that??

You mean in terms of speed? No problem.
In terms of temperature thermistors should not be taken above 300 C.


Enjoy the ride!

Glass bead thermistors have a time constant of several seconds. If your temperature rises more than a few degrees in that time you have to use proportional control. e.g. PID, rather than bang-bang. That might be noisy with a relay!

---

[www.hydraraptor.blogspot.com]

@Reinout:

Triffid_Hunter mentioned a solution with a mosfet..

I ordered the IRLxxx mosfet as I have a bit of a problem with the current solution with the relay..

It works.. but the temp is fluxing a lot +20C. And I guess this won't help stability in printing..

But does the firmware support the pwm solution Triffid suggests? Or do I need to write software to do this..

Or what other solution is there if I already use an arduino board for the controlling of the 4 axis board..I could setup another arduino board that controls the temp but if this is already in the current firmware why should I :-)


For anybody having an answer, thanks in advance!!!

Michael

dunno about official firmware, but mine not only has PWM for the heater, but its driven by a PID implementation for even more precise temperature control

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

> Glass bead thermistors have a time constant of
> several seconds.

Only when taken between extreme temperatures, when I calibrated my thermistor table the delay was not discernible. Furthermore I think the major time delay is temperature propagation from the heater to the thermistor, not so much in the thermistor itself. But then my thermistor is burried deeply in the heater barrel (excellent thermal contact), so your mileage may vary.


> If your temperature rises more
> than a few degrees in that time you have to use
> proportional control. e.g. PID, rather than
> bang-bang. That might be noisy with a relay!

Which illustrates how knee-jerk design like the 'use PID' I hear all over the place creates overly complex solutions. A proper model-based controller should be able to keep temperature well within one degree while using a relay. (the better room thermostats have been doing this for several decades now). Unfortunately I haven't programmed such a controller yet so it is not very helpful right here and now.

Edited 1 time(s). Last edit at 05/03/2010 04:11AM by Reinout Heeck.

I can't help you with the official software since it does not fit on my device. I'm running an older version of the firmware that I hacked a bit to keep it compatible with the current host softwares. So I'm still using the three-level bang-bang control which works well enough on my machine although it clearly gives rather course results (but not the +20C you mention, more like 5C over here).

Quote

> Glass bead thermistors have a time constant of
> several seconds.

Only when taken between extreme temperatures, when I calibrated my thermistor table the delay was not discernible. Furthermore I think the major time delay is temperature propagation from the heater to the thermistor, not so much in the thermistor itself. But then my thermistor is burried deeply in the heater barrel (excellent thermal contact), so your mileage may vary.

A time constant is exactly what is says. It is constant, regardless of the temperature swing. It is determined by the specific heat capacity of the thermistor element and the thermal resistance of the glass bead. It is stated by the manufacturer, so I tend to believe it. No matter what the temperature change is it will approach the new reading exponentially.

When the heater design moved to using Kapton the warm up time went over 100 seconds to tens of seconds. That is too fast for the thermistor to respond and caused the temperature swing with band-bang to go from a few degrees to a few tens of degrees so PID was implemented.

The simplest solution is to use a heater with enough thermal mass to keep the temperature swing to a few degrees with bang-bang, which is what I do. There is not a lot of point in making the extruder warm up very quickly because you have to wait for the plastic to melt afterwards anyway.

---

[www.hydraraptor.blogspot.com]

> A time constant is exactly what is says. It is
> constant, regardless of the temperature swing.

Yep, that's why I used the term delay, admittedly a bit vague/sloppy.

> When the heater design moved to using Kapton the
> warm up time went over 100 seconds to tens of
> seconds.

Good to know (I still 'bake' heaters with fire cement), saw no reason to move away from doing that.

> There is not a lot of point in making the
> extruder warm up very quickly because you have to
> wait for the plastic to melt afterwards anyway.

I have trouble seeing what you mean here, the thermal resistance of the plastic is too high when starting an extrusion, but once extruding it is no longer a problem?
Should I read this as an indication that the size of the melt chamber might need to be increased?

R
-

I have seen Makerbot users report that if they try to extrude too soon the extruder jams. My warmup times are over 100 seconds, rather than about 12 seconds, so I don't have that problem.

No the melt chamber is actually far too big in my opinion. The shorter it is the faster the extruder responds. There must be a limit, but the shortest I have been able to make it is about 10mm and that has no problem extruding at 40mm/s.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
> I have seen Makerbot users report that if they try
> to extrude too soon the extruder jams.

Ah, good to know, I lived under the impression they mainly had problems after idling the extruder a while at operating temperature.

> No the melt chamber is actually far too big in my
> opinion. The shorter it is the faster the extruder
> responds. There must be a limit, but the shortest
> I have been able to make it is about 10mm and that
> has no problem extruding at 40mm/s.

Ok, I'm about to make some new barrels (with metal temp barrier), I'll try to aim for a short and tight melt chamber.

Thanks!