No major difference seen with 24 VDC > ramps

I modified my ramps to accept 24VDC, powering the arduino from USB.

Overally, performance was identical. Subjectively, there may be a very slight increase in torque, but overall the motors and system on a printer were the same. I did not modify the firmware to allow for faster speeds.

So, is there anyway to reap the benefits of 24VDC, such as higher torque if possible??

Faster step rate. But also a more reliable faster step rate. Lastly is the heat up time of the hotend is way faster (at least at 15v it is).

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I think you are right, but given that I can hit 350 cm/min, I'm not sure the extra speed is needed. I also see that the speed vs. torque issue is only a trade off at higher speeds. i.e., at normal print rates like 50 cm/min, you aren't going to get any benefit from 24V, would that seem fair? Overally it seems like no benefit at all...

On the hot end, good point. It could be remedied with a low resistance heater...just requires a rebuild. I never liked the use of only a 2.5 amp heater that takes forever to heat and can barely resist cold drafts.

Now try turning up your acceleration to default 3000 and max 8000 and see how that extra voltage helps with the sudden increase in steps.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Can you outline the steps you took to get to 24v?

Did you change the heater resistor?
Did you modify your HPB?

I have read about this, but was uncertain what other mods were required.

Yes.
-Remove up the X and Y stepsticks on ramps
-Desolder or clip off the D1 diode (be careful for surface mount components & monitor heat level)
-Reattach drivers
-Apply 24VDC in place of 12 VDC (i'm using 24 VDC from [www.circuitspecialists.com])
-Use PC to power USB
That's basically it

is HPB heated printed board? I imagine that would work A LOT better with 24VDC, may overload it even.
As for hot end, I make them myself. Mine are 2 ohm, while standard reprap nozzles are above 6 ohms, I get a lot more power in there. I'm trying to goto 1 ohm now with thick nichrome wiring or stainless. 310C + is the goal for the "Augr", my next custom printer.

Interesting. Others have been more cautious with the heater resistance... I guess you are not using a heated print bed?

Can you explain this "cautious" in what way/ for what purpose /end goal? What are they afraid of breaking?

I am not using a heated bed because of other reasons, but I could if I wanted - how'd you know?

As far as cautious, they are clearly using higher resistance values to prevent higher currents through the heater. What they are protecting, I am not so sure. Thinking it through it could either be:

• Heater Resistor / Heated Print Bed burnout
• Mosfet overload
• Thermal fuses tripping

I am not sure.

How did I know that you were not using a HPB?

Quote

I imagine that would work A LOT better with 24VDC

It was your previous post that made it probable.. Don't worry I am not spying on you

6.8ohm @ 12v was chosen to prevent fires and burn out of the resistor. 21 watts being enough to keep the hotend hot enough for PLA and ABS without being able to reach their combustion temperatures or that of the surrounding parts.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Justblair Wrote:
-------------------------------------------------------
> As far as cautious, they are clearly using higher
> resistance values to prevent higher currents
> through the heater. What they are protecting, I
> am not so sure. Thinking it through it could
> either be:
>
> [*] Heater Resistor / Heated Print Bed burnout
> [*] Mosfet overload
> [*] Thermal fuses tripping

- What thermal fuses?
- How does a heater overheat?
- Mosfet overload is a real possibility, but unlikely at these power drains. RAMPS was designed for 11Amp/5amp draw at least, and at 27V potentially (33?, 36?).

I think the best reason given is that the extruder is barely capably of exceeding 250 C, that makes it inherently safe for the user.

Simba Wrote:
-------------------------------------------------------
> - What thermal fuses?
> - How does a heater overheat?
> - Mosfet overload is a real possibility, but
> unlikely at these power drains. RAMPS was
> designed for 11Amp/5amp draw at least, and at 27V
> potentially (33?, 36?).

By Burnout I mean that the tracks or the resistor is damaged by the current passing through.

The fuses are on the ramps board on the power lines.

I too reckon Sublimes answer sounds like the best answer.

I am wondering though about trying the 24v with the current equipment that I have and just make sure that the machine is not left unattended. The feedback from the thermistors will under normal circumstances prevent overload of the heater, the HPB I may just gamble with. They are not too much money to replace and I have a Mk1 as is..

Just if anyone is curious, I did the math that shows heat evolution (Based on maximum continuous feedrate). It comes out to about 7 W of 18 watts used. Adding insulation layers around the nozzle bumped it up and unknown but significant amount (perhaps 25% or more).

From my notes>>>(.25 mm) * (2.5 mm) * (25 mm) = 15.62500 mm^3 (i.e.6.5 mm filament/second). This is max extrusion rate.
For PLA this is close to 15.6 mg. For heat capacity, 1.670 J/g-K. with 175K temp rise, and 15.6 mg, this is 4.5591 J needed. Thus for 18 W only 25% of the nozzle heat is being used. Assume latent heat of fusion around 160 j/g, then anther 2.5 watts goes into phase change. Thus the system is using 7 W of the total 18 for heating the plastic.

At present I think this is the biggest limitation that needs to be overcome for reprap. We don't just want small gears. We also need large parts. A 7W nozzle is fundamentally limited.

Say you want to print a 5 x 2.5 x 2.5 (in^3) part. The total amount (solid) would be 512,000 mm^3. Lets assume 10% solid infill...now it is 51,200 mm^3.
At max feed rate, this is approximately 3280 seconds, not including print moves. With print moves, thats more like double or triple, and with retraction maybe up to 10 times slower. For this part, about 3 hours to print is what it would actually take. That's no good if say, you are trying to make a dozen tee fittings. Also, you run out of filament. A filament roll is 216,000 mm^2. So you could print only 4 of these relatively small 10% infilled parts with 2.2kg filament.

This is all the proof I need that reprap prints so very slow and that we need a better solution for rapid feed rates. To this end I'm working on a granule fed, dual extrusion (3mm and .2mm) for rapid infill, with 200 W heater nozzle. With this, such a part could only take 20 minutes to print, and would be stronger and much cheaper. The only question I have is if reprappers are happy printing super-low-density models or is there a real need to print larger dense parts like I want.

Edited 1 time(s). Last edit at 12/17/2012 06:16PM by Simba.

I'd love to eventually see a granule extruder, similar to the one built by Dirk Vanderkooij:

Creating an endless chair

I'm sure there are some economies of scale that are helping him but I'm still impressed.

I'm not concerned about constructing dense models. To me, the biggest advantage to FDM type fabrication is being able to control the interior structure of an object. Typically I like to define the infill type orientation and density depending on the requirements of the part being produced. But, as always, faster is better.

Yes, my favorite guy who believes its great to take new technology that should be making things cheaper for everyone and charging $1000 for recycled chairs.
More power to you if you can get away with it, I'm just jealous and in awe, or apparently clueless about market dynamics.

I'm glad you are interested in the granule extruder. I think it will be done, and I know at least 5 people working on it. Generally, there are a huge number of challenges to overcome, and though they won't tell you, most people failed for one of three reasons:

1. terrible inconsistency
2. failure to eject plastic with enough pressure to make it through a small nozzle diameter
3. inability to be able to handle cool-downs and warm-ups without gunking the system or cracks forming due to the thermal expansion changes.

If anyone tells you they've solved it (i.e., by selling an unproven design out of their notebook page for $150) make sure they've addressed these three things before believing them.