TMC5160 substitute?


eugenecisneros
 

I am trying to order some TMC5160 drivers and the supplier is telling me that there is a new version.  The "new" version has a different pin configuration from what I can see.  Hopefully, someone can advise as to whether or not these will work in the MaxSTM 3.6 board.  Thanks in advance..


Dave Schwartz
 

Looking at the MaxSTM 3.6 board, it only has room for drivers with 8 pins on each side (and no PCB overhang beyond the pins). The longer one won't physically fit.


On May 7, 2022 4:56:37 p.m. EDT, eugenecisneros <elc@...> wrote:
I am trying to order some TMC5160 drivers and the supplier is telling me that there is a new version.  The "new" version has a different pin configuration from what I can see.  Hopefully, someone can advise as to whether or not these will work in the MaxSTM 3.6 board.  Thanks in advance..

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.


eugenecisneros
 

Dave, you are right.  I didn't consider the physical size.


eugenecisneros
 

Size is not a deal breaker, as workarounds are possible.  The real question is if these are electrically compatible.  I see them being sold as QHV5160.


Howard Dutton
 

I'm wondering about the goal, what motors are you driving that you need a TMC5160, what step rates do you want to hit, etc.

There might be other options.


eugenecisneros
 

Howard, I realize that there may be options, but I don't want to err on the low side and have to repeat the exercise.  My mount is quite heavy, about 150 lbs on the polar axis, so quite a bit of inertia.  I already have a Nema 23 Bipolar 1.8deg 2.4Nm (340oz.in) 1.8A 4.95V and would like to use it.  I thought that I would try running it with 2130s at about 0.9A today and see if it makes significant power.  I plan on a 5:1 planetary gear box and 1:1 belt drive to my final 200:1 worm drive.  I would like to have tracking of at least 1 arc sec and a reasonable goto rate of at least 2 degrees/sec.  Does any of this sound reasonable, or am I chasing my tail?  Thamks!


 

@Howard
Will MaxSTM 3.6 support Fysetc TMC2209 v3.1 in UART mode on OnStepX? If not, then it may be cheaper for him to replace board with something that work with standalone drivers and then use TMC2209 as they can deliver required power and you can buy handfull of them for a few bucks.


Howard Dutton
 
Edited

On Mon, May 9, 2022 at 03:53 AM, Vladimir wrote:
Will MaxSTM 3.6 support Fysetc TMC2209 v3.1 in UART mode on OnStepX?
Not sure as it hasn't been tested.  But in theory, yes.  SoftwareSerial mode and not with status feedback though, and if used all four axes must be TMC2209's.

If not, then it may be cheaper for him to replace board with something that work with standalone drivers and then use TMC2209 as they can deliver required power and you can buy handfull of them for a few bucks.
No reason a MaxSTM3.6 can't work with stand alone drivers.  You might want to add a pair of enable pins to the pinmap for Axis1 and Axis2, I'd map them over those axes CS pins.


eugenecisneros
 

On Mon, May 9, 2022 at 06:31 AM, Howard Dutton wrote:
No reason a MaxSTM3.6 can't work with stand alone drivers.  You might want to add a pair of enable pins to the pinmap for Axis1 and Axis2, I'd map them over those axes CS pins.
Is there any plan to make the MaxSTM3.6 compatible with stand alone drivers whereby microsteps can be changed for goto speeds?


Howard Dutton
 

On Mon, May 9, 2022 at 09:56 AM, eugenecisneros wrote:
Is there any plan to make the MaxSTM3.6 compatible with stand alone drivers whereby microsteps can be changed for goto speeds?
No.

You are free to configure OnStep to do what you will with it though, and it can do that kind of thing AFAIK.  How you map those features onto the available otherwise unused pins/connections of a MaxSTM3.6 is up to you.

The STM32F411 can step at 62.5 KHz in square wave mode (best for external stepper drivers.)  That's enough for about 1.75 deg/s in 64x microstep mode in your case, I believe.  Or double that speed (3.5 deg/s) in 32x mode and that's without mode switching.  Naturally that doesn't mean the motors/drive are capable of reaching those speeds.


 

On Mon, May 9, 2022 at 03:31 PM, Howard Dutton wrote:
No reason a MaxSTM3.6 can't work with stand alone drivers.
Didn't knew that, I think Max series wiki page state that only SPI drivers can be used with MaxSTM but I didn't go for details...


eugenecisneros
 

You are free to configure OnStep to do what you will with it though, and it can do that kind of thing AFAIK.  How you map those features onto the available otherwise unused pins/connections of a MaxSTM3.6 is up to you.

The STM32F411 can step at 62.5 KHz in square wave mode (best for external stepper drivers.)  That's enough for about 1.75 deg/s in 64x microstep mode in your case, I believe.  Or double that speed (3.5 deg/s) in 32x mode and that's without mode switching.  Naturally that doesn't mean the motors/drive are capable of reaching those speed
Thanks Howard, I think that the above is beyond my abilities as I have been out of the game, retired for 20 years now.  Since TMC5160s are not available, my progress has stopped, so I will have to look for another option.  


Howard Dutton
 
Edited

On Mon, May 9, 2022 at 11:24 AM, Vladimir wrote:
Didn't knew that, I think Max series wiki page state that only SPI drivers can be used with MaxSTM but I didn't go for details...
Best left that way and to generally not speak of hacks like this in the Wiki as that is not within the scope of the intended (and tested) use.

That said once TMC2209 SoftwareSerial mode is checked out and OnStepX is released I will likely add the TMC2209 and TMC2226 to the list.  In software serial mode only the drivers themselves need modification (cut off their RX pins) otherwise they should just plug in and work.  Well, a single resistor (say 10k or so) between each drivers TX (aka the sockets CS) and Vio pins should be added as that prevents them from coming up powered for a moment before OnStepX gets control (the MaxSTM3.6 doesn't have stepper driver ENable pin control as it isn't really needed for TMC drivers.)

That gets you the ability to set any mode on any driver and to mode switch ... but no status info. from the drivers.


Howard Dutton
 
Edited

On Mon, May 9, 2022 at 12:05 PM, Howard Dutton wrote:
That said once TMC2209 SoftwareSerial mode is checked out and OnStepX is released I will likely add the TMC2209 and TMC2226 to the list.  In software serial mode only the drivers themselves need modification (cut off their RX pins) otherwise they should just plug in and work.  Well, a single resistor (say 10k or so) between each drivers TX (aka the sockets CS) and Vio pins should be added as that prevents them from coming up powered for a moment before OnStepX gets control (the MaxSTM3.6 doesn't have stepper driver ENable pin control as it isn't really needed for TMC drivers.)
In TMC2209/TMC2226 driver HardwareSerial mode one would instead cut off the MS1 and MS2 pins of the drivers to program addresses 0, 1, 2, and 3.  One driver has none cut off (address 3.)  One has just MS1 cut (address 2.)  One has just MS2 cut (address 1.)  And finally one has both MS1 and MS2 cut (address 0.)
Then leave the RX pins intact, you want them to plug in and all connect together using what would be the SPI MISO line.
Then connect the RX1 and TX1 pins (STM32 hardware serial port) on the MaxSTM3.6 DB15 connector to the Axis1 drivers RX and TX pins (RX1 to RX and TX1 to TX in this oddball case.)
You still would want a pullup resistor on the TX pin of Axis1 to disable the drivers (one resistor would take care of all drivers.)
The pins that would otherwise be MOSI and SCK are automatically pulled HIGH (to help setting the driver address.)
The pins that would otherwise be CS and MISO are not initialized so they are OPEN and don't interfere with operation even though there are unused connections back to the microcontroller.

...that's the theory anyway as this too hasn't been tested (except similar on the MaxESP3 which works fine.)  This allows very high baud rates (460800 bps) and bidirectional communication.  The latest MaxSTM3.6 designs have a RX1/TX1 header added on the PCB so you don't have to wire into the DB15.


Howard Dutton
 

Note that TMC2209 support like this requires using the FYSETC TMC2209 version 3.1 as they have an onboard 1k resistor between TX and the TMC2209 chips PDN pin (and so RX pin.)  That is required and they are the only ones with it.

Read the TMC2209 datasheet to understand why it is required if you like.

Naturally one could design a different pinmap and add their own 1k resistor to make other TMC2209 drivers work.  Its just a matter of getting the wiring right.


George Cushing
 

The MaxPCB2 & 3.6 and MaxSTM3 have unused pins that can be put to work. The MaxESPX have a limited number of GPIO pins.


George Cushing
 

TMC5160 available $40 a unit.


Howard Dutton
 

On Tue, May 10, 2022 at 01:52 PM, George Cushing wrote:
TMC5160 available $40 a unit.
These look like a better deal but more risk as I don't know about the manufacturer and not 100% sure TMC5161's are just like TMC5160's as far as OnStep is concerned... I think they are but not aware of anyone who's tried them???

https://www.amazon.com/XUXUWA-Printing-Accessories-Printer-Motherboard/dp/B08QV6LH5W/ref=sr_1_5?crid=QLH45IQ2A8C1&keywords=tmc5161&qid=1652216135&sprefix=tmc5161%2Caps%2C54&sr=8-5


eugenecisneros
 

I have 2ea S5160s $19 ea on order and will report back when I test them.  


Robert Benward
 

I ordered a pair of these QHV5160 for $17 each.  They are the high voltage variants of the TMC5160, but they have extra pins that need to be cut away and they overhang on one end.  There might be sufficient pin length that you can leave one end not quite seated and overhanging the next driver.  This assumes light duty and you are not using the heatsink.  Check your board to see how these are oriented, if they are end to end or side by side.  My board is side by side so it was not an issue for me.  They claim to have 3000 pcs in stock.

https://fr.aliexpress.com/item/1005004195966600.html?spm=a2g0o.order_list.0.0.43d55e5bdLyGZ9&gatewayAdapt=glo2fra

Howard,
Those 5161 are strange beasts.  The external MOSFETS are on a board below and I suspect the chip output is being used to driver the gates of the mosfets, but since the upper board is available alone, I don't know how they break the pin-to-socket-to-pin connection to route the outputs to the gate inputs.  The Fysetc wiki shows the upper boards as available separately.  The upper board alone can handle 3.5A.   I wonder where the power board comes from?

https://wiki.fysetc.com/S5161_V1.1/

Bob