Series vs Parallel z-axis steppers

I'm struggling in the commissioning of my Mendel90-based printer with getting the x-axis consistantly parallel with the bed. Once I've manually turned each of the z-axis lead screws to get the hotend just kissing the bed, once I raise and then lower it a signifiant amount, say >50 mm, or home the z-axis, one side is off by a noticable amount. I'm guessing it's missing steps.

I'm using these steppers RAMPS 1.4 electronics, and pololu knockoffs that seem to be decently made. Initially I hooked them up in parallel as the RAMPS board has dual headers for the z-axis and it was the easiest...no rewiring of connectors needed. The pololu is set with a Vref of 1.2v (1.5 amps * 8 * 0.05 ohms * 2 steppers). Anything less than this and they just squeal and don't move. That results in my current situation of missing steps.

Seeing that it's recommended (both in nophead's build guide as well as other forum posts) to wire them in series to reduce the current requirements, I wired them up in series. Reducing the Vref down to .6v I can not get both of them to turn. Half the time they just squeal similar to when they were in parallel with a Vref < 1.2v. When they do turn, only one or the other turn but not both. If I give a slightly nudge to the couplers to help them turn, they'll turn maybe 1/4 or 1/2 turn then squeal some more. Adjusting the Vref +/- a few tenths fo a volt does not appear to help the situation.

What am I doing wrong? When in series they both turn the same way when they do decide to turn, and the fact that they both do move some would seem to indicate that they are wired correct (no wires crossed, mixed between coils, etc) right? I'd have to try again, but I think very small movements, like .1mm, worked, but that's less than the distance it would be just twitching while it squeals with the M6 lead screws.

Ultimately I don't care if they are ran parallel or series as long as they work reliablity and accurately.

I use motors that are similarly rated to those, on an over-sized MendelMax with 10mm rods, which means an X axis that is a fair bit heavier than most. I run my Z axis motors in parallel, with a Vref of 0.4V, giving about 1A total current, or 0.5A per motor. For most machines this should give plenty of power to drive the Z axis. Have you calculated the steps/mm for the Z axis? Have you set a reasonable max speed and acceleration for the Z axis?

Just cause they turn doesn't mean everything's wired right. I had a similar issue when I first started hooking up steppers and it wound up being a wire that wasn't making contact very well, and the motor did turn but it had absolutely no torque to it. I'd also make sure you try very slow speeds as well as hooking up the motors to a different stepper driver to rule that out as the problem.

Quote
NewPerfection
I run my Z axis motors in parallel, with a Vref of 0.4V, giving about 1A total current, or 0.5A per motor. For most machines this should give plenty of power to drive the Z axis.

I was basing my calculations on the "reference forumla" to get the full torque and holding power. As I mentioned above, if I go lower than the 1.2v (.6v/motor) both motors squeal and don't move. That's why I haven't tried a lower reference voltage.

Quote

Have you calculated the steps/mm for the Z axis? Have you set a reasonable max speed and acceleration for the Z axis?

Both settings are the default that nophead distributes as part of his Mendel and it's built to his specs with respect to axis movement. steps/mm = 200 * 16 for 1/16th microsteps. I do remember that one for sure. I'd have to double check once I get home but feed rate should be 4mm/sec for Z and acceleration of 150.

What kind of pololu that is and how hot it runs, if you put your finger on it. Perhaps instead of 0.05 the sense resistors might be 0.10 and then its like 1.5a total, which feels more reasonable than 3.0a/pololu, and needing just double vref to get same setting makes sense- ockham's razor.

Doh. I feel really kinda foolish now. I'm not sure where I got the .05ohms for the sense resistor, but I finally got some free time to double checked things and it wasn't .05 ohms. It was .3 ohms. That put my current at 1/6 what I thought it was. Computing with the correct value, 1.5 amps * 8 * .3 ohms = 3.6v, well over twice what I could get Vref to be. I had another pair of driver boards from a previous project that have .1ohm sense resistors that I used for the z-axis. With those, I was able to successfully get to run the two steppers in series but only at 230mm/min. If I pushed it to 240, the default with nophead's modified firmware, they stall out. I'll probably back them down to 220 or 200 as they sound better (less whine) at the slightly slower speeds.

Oh, 0.3 rsense is rather big, i think there are like two caveats you should watch for, one the voltage across rsense has to be less than 0.5v, and another is vref(max) should be less than 4v when microstepping. Not sure of the last one, probably should check datasheet. With 1.5a setting both these limits seem pretty close. I think this makes 1.5a sort of the max usable setting. You could change rsense resistors, especially if they seem to heat too much, but even if so, this chip probably wont give any significant setting increase. Hopefully the 0.3 resistors are at least 1w, perhaps better if are 2+w (++bigger package than typical 1206 smd).