I have a bag of QHV5160 drivers and a triplet of BTT Octopus-pro’s and 48 volt psu’s to match. Building a tronxy-400-pro up for faster stuff. I have adapters to drive CL42 stepper/servo’s for XY. So I think I have that covered, using klipper since Marlin never worked on this printer.
Does anyone have a printer.cfg paragraph for those 60 volt, 3 amp drivers? I’d like to use them for z and extruder
One remark:
IMO such drivers and their matching stepper motor do not make any sense for the Z axis and even less so if this axis is using lead-screw, which naturally limits its accel and velocity.
Also for the extruder it is questionable if there is any real life benefit.
But in any case have fun and as @EddyMI3D pointed out: It is just the standard TMC5160 config.
Which, unfortunately, I do not have a copy of, and its not grep-able in
the klipper/configs directory. So where can I get a copy of it? But I
did find it following a discourse thread.
Thank you
Cheers, Gene Heskett.
Not to be snarky or sarcastic, but why wouldn’t you try Google? Searching on “TMC5160 Klipper config” will lead you straight to:
Which I didn’t ask about, I asked about the QVH5160, which might turn
out to be a totally different critter.
Which the latest debian 11.6 version of firefox won’t print, stopping
about 1/4" above the “firmware” header and what I need is on the next,
can’t be printed page down from where the printout stops. I’ve
bookmarked it, but will have to hand copy it if I can’t copy/paste it on
that machine.
Earlier I note on that link that it says I need to put that driver
socket in the spi mode, so now I’m trying to poster-ise the one board
pix that actually shows where stuff plugs into a BTT-Octopus-pro, but
poster crashes my machine trying to load that. Apparently 32gigs is not
enough memory.
BIGTREETECH-Octopus-Pro-V1.0-Color-PIN-V3.0.pdf
which is an odd sized image. Libreoffice-draw can load it but can’t blow
it up. I’d need it about letter high & 3 letters wide, the text is very
dense. Or 2 landscape sheets would do too.
To top that, I’ve reached the age (88) where macular degeneration has
started to affect my central vision.
Also, I usually use DDG, google knows way too much about us already.
There’s probably more, but short term memory is also a problem.
Anyway, if you know how to blow that pdf up at least 2x, I’d appreciate
it. I do have a printer that can do it.
Take care & stay well, Myke.
Cheers, Gene Heskett.
I’m not sure where the problem is - the QHV5160 uses TMC5160s along with high current MOSFETs which is what the TMC5160 is designed for as referenced in the datasheet:
https://www.trinamic.com/fileadmin/assets/Products/ICs_Documents/TMC5160A_Datasheet_Rev1.14.pdf
The module schematic is nothing surprising:
The QVH5160 wiki page (which @EddyMI3D linked to) has the same printer.cfg information as the Klipper documentation for the TMC5160 (which I linked to) which makes sense because the QHV5160s are based on the TMC5160:
In your original post, you asked about where you can get a sample Klipper printer.cfg information for the part and it’s simply the TMC5160. I’m not sure what the issue is.
If you’re having problems figuring out how to set the jumpers for SPI mode on the octoPro, then I suggest that you look at the manual:
And go to Page 15 as it has a nice marked up photograph of the board showing exactly how to do it:
that I have on dead tree. I worded the question poorly, it appears these jumpers can be configured on a per driver socket basis, is this correct and there is no interaction between sockets?
Plans are to tap the first 2 sockets with adapters similar to whats supplied with the ill-fated 42c motors, to obtain the step/dir signals which will then drive a pair of CL42 stepper/servo controllers. The motors I have, have optical encoders of the A/B quadrature style, avoiding the hall based devices used in the 42C, and which result in un-acceptable delays in the feedback from the a/d conversions causing the 42c motors to move in fairly gross steps=piss poor performance at speed. I have been using the 3 phase version of this tech on my cnc machinery since it came out a couple years ago. Nyquist phase shift in the feedback path simply does not exist in that encoder tech. The controller knows precisely where the motor is 4000 times a revolution. So I should be able to move a nema-17, .8nm motor at pretty high jerk and accels. I’ll use a separate 48 volt supply for those two motors, and could go to 90 volts ack the labels on the drives.
Thanks Myke, take care & stay well.
When I look at the schematics, there’s no reason why you can’t mix and match driver modules with different communications approaches (ie UART vs SPI) on the Octopus Pro board. I don’t know if the same can be said for Klipper (but I suspect it can). I would not recommend mixing addressed UARTs with any other types of modules.
I haven’t worked with SERVO42C boards so I don’t know how the serial communications are wired. It looks like they don’t work well with Klipper:
I have no idea how you will wire CL42s into your system.
What is the “Nyquist phase shift”? I can’t find a definition for it.
I have doubts here.
Good luck, hcet14
they take conventional step&dir signals via opto-coupling. But they have an advantage over conventional 2 phase steppers as they will issue a shut down if they lose home, stopping linuxcnc in its tracks in under a millisecond. They will get to where they/re commanded to go, on a step by step basis or stop the whole show.
It may take some finagling to transfer that instant stop to klipper but thats a bridge I’ll cross when I’ve got it moving motors, heating things and such normal for 3d printers stuff. The 3 phase version of these drivers need a powerdown reset to restore operation, but Hanpose, the makers of these, hasn’t yet shared that info.
I also have an ender 5 plus that is getting the same treatment except its not corexy, With a 20x20 alu x bar and direct ejection, it gets heavier and the y motor looses home at under 100mm speeds. So the twin 8mm shafts coupling the y belts is now a 25mm cf tube, driven by a 1nm 3 phase stepper/servo And the x motor is just a hair longer than the oem, but also has the optical encoder, and it too will be driven by a Hanpose CL42 Both of those motors will be powered at 48 volts. The Y shaft is driven by a 1/1 belt from that 1nm nema 23, so armature inertia may enter into how much jerk and accel it can tolerate, That 20x20 alu x bar is now a 20x20 cf tube carrying a 9mm linear rail, A bit flimsy perhaps, time will tell that story. I designed and printed the heads, using creality’s Spider V3 with a 70watt heater, which has a 4 bolt mounting on top, and printed ductwork. Latest version sucks its intake air much higher off the bed as the previous version like the creality base head sucked in some overflow filament trash, jammed the fan and melted it all down in the middle of the 3rd day of a 6 day run making woodworkers bench vise housings 6 up. Close to $60 worth of PETG+CF I tossed. 
I was using diamonback nozzles for the cf stuff, but it seems the cheaper carbide nozzle melts plastic faster, and its plenty hard enough for cf petg.
Thanks Myke, take care & stay well.
Nyquist is a term use in feedback systems in hifi audio since in the
1950’s and is normally expressed as the phase error between perfect
feedback, and that obtained in real world practice. A Nyquist phasing
of over 90 degrees at any frequency means its going to oscillate because
for that time period, the lag turns the negative feedback into positive
feedback, making the circuit unstable at best.
In the case of the poor 42C motor’s the phase shift is caused by the
nominally 5 milliseconds it takes to get a new A/D conversion from the
hall device measuring the motors shaft position. So if the motor is
moving at a usable speed, the feedback is always 5ms stale, and the
controller will jerk then stop, wash rinse repeat leading to the shingle
effect that has been quoted on the net several times.
The spelling may not be precise either.
Cheers, Gene Heskett.
