Thursday 26 November 2015

New laptop, old Melzi

First time trying with my new laptop, so of course half an hour went by finding out that the wifi password isn't on the wiki anymore. But now I'm connected.

First order of business: Get the firmware back up and running. I tried with the old version from the Mac, but without luck. Instead I grabbed a Mendel90-specific Marlin and uploaded that, after moving an old clashing dir out of the way it worked.

Cloned Printrun and ran it, all axes work nicely. Temperature sensors still read 40C. I measured them before putting power on, and they were at 11.5K ohm. Now they read - (!) 90K. There's probably a reason they change when there's power, I'm not going to worry right now.

This post claims it's because some Melzi boards still have 10K instead of 4.7K pullup resistors for the thermistors. And indeed, it does. So even though 3dnielsen.dk calls this a 2.0 Melzi board, it has old resistors. Bother. The old one does have 4.7Ks.

Manually holding another 10K resistor changes the temperature reading a little bit, but not down to 20C. I might just not be able to hold them appropriately.

Three options now:

1: Buy a new board. Expensive, and I'll have to rewire, and it may have other flaws.
2: Resolder the SMD resistors, taking the ones from the old board which are correct. Haven't tried soldering SMD before, but these are relatively large:



3: Recalculate thermistor tables. There are some posts about this, but they all assume you have a way to measure the temperature. I tried measuring my hotend earlier with an infrared meter, but it never read anywhere close to what it should.

Just for shits and giggles, I wired up the old board with just power (easy now with the name), ETemp, EHeat, and direct connection to the extruder motor. Heat it up, and whaddyaknow, it extrudes! Well, runs the extruder motor at least. So it worked all along, maybe just with too little power and definitely with the wiring problem. So now I have another option:

4: Switch back to the old board again. Try not to cry.

At least I didn't start soldering before I tested this. Could easily have destroyed both boards for nothing.

I think I'll spend the next evening soldering the tips of all the wires going into the Melzi, so I don't have further issues with wires going off on adventures.

If need be, I can have separate wires circumventing whatever it is that's broken with the normal extruder motor wires.

In other news: FabLab now has no less than three (3) laser cutters, including one 150W oxygen-laser 1,5 m2 cutter. It slices metal sheets. Sweet.

Sunday 15 November 2015

Wiring up

In our last exciting episode, I found that the new Melzi board works on the extruder port, as does the new motor, and there is some problem between the board and the extruder motor when mounted. Let's see what today brings.

First trying if the old extruder motor actually does work. It would be nice to not have to take stuff apart to get it out. Setting the extruder temperature to 185, but noticed that it started out say 43, which is definitely too high. Shouldn't be a problem for now, though, since I have no filament in there.

Yay! The extruder motor works! Now I'll let the print head cool off while I mount the other things.

Wired up the stops and the bed temp. The X, Y, and Z endstops work. How do I know when I don't have motors wired yet? I turned on communication debugging, made each axis home, then pressed the endstop and saw it triggered in the debug window.

The bed thinks it's at 38 C. That both temperature sensors are off by about the same amount maybe hints at a wrong setting.

Next test the bed heater. When I tried starting it, the power supply stopped. Turning it off for a while then on again made the power supply happy, but something is bad. I don't expect that the bed would draw more power than the extruder heater, but it looks like the fuse is blowing, even when I don't have the extruder heater on. Short circuit is most likely then. On close inspection, there's a loose wire:



Rewired more carefully, now it works.

Wired the motors, X, Y, and Z work beautifully. Since all parts of the Melzi work (except the temperature adjustment part), I mounted it back on the frame. Now try to get some more info on the thermistor problem... Aaargh! Nothing works! Yelp! Oh, it's just the yellow power that popped out when mounting it. I have to say, the Mendel90 doesn't leave much room for all the wiring. If I were to redesign something, it would be putting the Melzi out at the rear edge and the resistors in the corner. Might be too far away for some power supply cables, though.

I got tired of how the power connectors are awkwardly pressed in. Instead, I put a nail on the black wire and a larger wire on the yellow, soldered them well and wrapped them in insulating tape:


After this, Y and Z works fine, but the X motor has a ticking sound when doing nothing and is very rumbly when moving. Apparently, that's what happens when the red wire gets pulled out. Reattached, but boy is that a bother. Upgrade idea: Attach jacks on all wires.

Everything works now except the extruder connection and the temperature. Looking at the connections first. With the extruder motor disengaged, I measured the resistance between the connectors. Between K, B, and G, there is connection, but not between any of those and R. Even when the print board at the extruder is disconnected.

Compared to the other stepper motors. They have connection between R and B and between G and K, but not between those groups. According to the schematics, they are all the same. So something is wrong. Again:( According to an electrical engineer who came by at open house, I should understand what the variable resistors do and how to safely adjust them before powering up. That might have been a good idea.

At this point, my laptop ran out of power and I had no charger with me.

Regarding the temperature measurements, one page mentioned that a 50K thermistor would show up as 40 C. That's what would happen if there's a short-circuit that puts the two thermistors in parallel somehow.

Sunday 8 November 2015

New Melzi, new motor, new hope.

I got a new Melzi and Nema17 from 3dnielsen.dk. The motor didn't budge when hooked up to the old Melzi, so it does seem to be the Melzi (at least).

I carefully marked the existing wires using masking tape and unmounted the old Melzi. Admittedly, it was a rather crappy cheap make, not silk screen or anything.

Wired up E Temp, Power, and E Motor, and tested. Didn't work, of course, without Hot End wired. Added then, then tried again. It responds to input (it has an actual Blinkenlicht), but doesn't heat. Oh, right, haven't uploaded the right firmware. Do that, adjust baud rate, trying again, and now it heats! Ah, toasty!

That's not enough to satisfy the extruder motor, though, it still refuses to budge. Try wiring it straight into the board, and that works! So there is definitely a problem in the wiring. Wonderful. Enough for now, though.

It seems to heat faster, so I have that going for me, which is nice.