TMC2100 How hot?


Howard Dutton
 
Edited

It seems the SSS TMC2100 uses the QFN package which is rated for 1.2A (RMS) as opposed to the TQFP which can handle 1.4A.  So I guess if you have excellent cooling (fan/big heat sink) the TMC2100 SSS maxes out at 1.2A RMS and can fully power a 1.7A rated stepper motor using a Vref of 2.4V 1.7V.  [Note: the Watterott GitHub shows the TMC2100, TMC2130, and TMC2208 all as 1.2A RMS "motor phase current".  The TMC2208 datasheet says it's 1.4A RMS (and a lower RDSon makes me think they might run a bit cooler.)]


Howard Dutton
 
Edited

I'm still not understanding why the RepRap formula is wrong. 

I admit to not understanding the TMC2100 datasheet content on this subject entirely, I don't see how the exact values in the table on Page 23 (Section 7) relate to the "RMS RUN CURRENT CALCULATION" in Section 8 on the following page but the values are close to what I'd expect and for the SSS's 0.11 Ohm sense resistors (all RMS):

Table (interpolation:)  1.78A
Formula (RRCC): 1.74A
RepRap Wiki: 1.77A

Now those values are just the basis.  The SSS TMC2100 is designed with current scaling so that trim pot changes VAIN from 0V to 2.5V (20k fixed and 20k pot as a voltage divider) which then scales the current from 0A to 1.77A (about.)  This additional calculation is covered by the formula in "CURRENT REDUCTION BY ANALOG SCALING" in Section 8.

The RepRap Wiki formulas look right to me:

Their example shows 1V = 0.71A... so lets see (1V/2.5V=0.4, 0.4*1.77A=0.708A)

[I just spent some more time looking over the RepRap Wiki and even though the formula they show works their discussion just above in that section appears to be in error.  They claim Imax is 1.77A and Irms is 1.25A which goes against what I understand and also the Watterott FAQ which puts Imax at 2.5A and Irms at 1.77A]


Mark Christensen
 

Antonio,

Probably because they just cut and pasted from data from other stepSticks. That kind of thing is why I always go back to the manufacturer data sheets.
Speaking in general terms, there is a lot of useful information out there and there is a lot that is wrong as well.

Mark
--------------------------------------------

On Sun, 1/21/18, Antonio Velkov <dreamingdies@...> wrote:

Subject: Re: [onstep] TMC2100 How hot?
To: onstep@groups.io
Date: Sunday, January 21, 2018, 5:33 PM

Mark, as far as I can
understand the formulas in the wiki for TMC2100 are
wrong?
http://reprap.org/wiki/TMC2100

I've checked the datasheet
and your formula is correct. Why then reprap and all other
websites are calculating with wrong formulas?


Antonio Velkov
 

Mark, as far as I can understand the formulas in the wiki for TMC2100 are wrong?
http://reprap.org/wiki/TMC2100

I've checked the datasheet and your formula is correct. Why then reprap and all other websites are calculating with wrong formulas?


Howard Dutton
 
Edited

On Wed, Jan 17, 2018 at 10:28 pm, Mark Christensen wrote:
Also, in addition to the 1.4Amp (RMS) figure quoted on the front page (for the TQFN package) see page 26, which gives recommended current ranges.
I believe that you're correct and my take on this was wrong.  My misunderstanding was with which current is referred to.  The current into the driver is well below 1.4A RMS x 2 (per coil) at 12V.

On the question of power - motors generate their torque from current - the voltage (DC or chopped) is just a way of delivering the current. So calculations about power are not really the issue.
Sure voltage is just a way of delivering the current, but how fast do you want it delivered (power?)  A current of 2A at 1mV, for example, isn't going to operate a stepper motor with the same torque as 2A at 1V, and so on.  Somewhere both current and voltage enter into it so the driver must limit power and we know it does (that trim pot, sense resistors, current chopping.)  So the driver (current) spec. isn't for the 12V supply voltage into this stepper motor's 1.1 ohm coils which I assume it tries to do (but cuts off the current before it rises the 10.9A level?)  And it isn't for the overall limited current (per coil) at 12V which is 0.37A  It's perhaps based on the stepper motor coil resistance (which the driver can "know") and the current level we set on the driver (trim-pot) since those two can represent all that needs to be known about the stepper motor to drive it properly?


Mark Christensen
 

Also, in addition to the 1.4Amp (RMS) figure quoted on the front page (for the TQFN package) see page 26, which gives recommended current ranges.
For external sense resistors they recommend no more than 1.4 amps. So while it may not be a hard limit the implications are pretty clear: Keep to 1.4Amps or less, or accept the risk.

Cheers,

Mark

--------------------------------------------

On Wed, 1/17/18, Howard Dutton <hjd1964@...> wrote:

Subject: Re: [onstep] TMC2100 How hot?
To: onstep@groups.io
Date: Wednesday, January 17, 2018, 8:53 PM

On Wed, Jan 17, 2018 at
02:42 pm, Mark Christensen wrote:

This formula is valid for the current adjustment
mode and to enter that mode you MUST  set the CFG3 pin (pin
5 of JP1 on the SilentStep Stick) to OPEN (as in open
Tristate logic). Again, read the TMC2100 spec sheet, page
23/24, which recommends a 1.5 amp motor or smaller be
used.
I don't think the datasheet is saying 1.5A stepper
motors are the limit there:  "The following  table 
shows  the  RMS  current  values  which  can be 
reached  using  standard  resistors  and motor types
fitting without additional motor current
scaling."

At
the bottom of the table it mentions using AIN current
scaling (the pot we adjust does this) for those examples
that were out of spec.


Mark Christensen
 

No, it is not the limit but if you read the very first page of the spec sheet it clear says for the TQFN package (which is what I see on the sticks I bought from Digikey) can handle 1.4A RMS. Yet on the same page it says you can use the stick with motors whose coils draw up to 2 amps. The two are not mutually exclusive. But it also means you don't want to drive them at 2 amps and you shouldn't need to.

And a bit of conservatism is always appropriate, so I wouldn't even get close to the 2 amp figure (or even 1.4 if I could get away with it).

The fact that JoAnn asked the question because the drivers were 'smoking hot' as she said means the Vref needed to come down a lot.

On the question of power - motors generate their torque from current - the voltage (DC or chopped) is just a way of delivering the current. So calculations about power are not really the issue.

Yes, you are correct: The CFG3 jumper is not connected. That's the problem with looking at too many spec sheets under too many tabs in Adobe.

I was happy to hear from JoAnn that she turned the Vref down and the motors still seemed to operate nicely and the drivers were cooler. Good-that is what should have happened. Short of putting a representative load on the motor (either a scope or some kind of friction disk with a dynometer to measure rotational torque) there is not much to be done. I wouldn't worry about it, frankly, since these motors have a lot of torque anyway and telescopes are heavily geared down.

Cheers,

Mark
--------------------------------------------

On Wed, 1/17/18, Howard Dutton <hjd1964@...> wrote:

Subject: Re: [onstep] TMC2100 How hot?
To: onstep@groups.io
Date: Wednesday, January 17, 2018, 8:53 PM

On Wed, Jan 17, 2018 at
02:42 pm, Mark Christensen wrote:

This formula is valid for the current adjustment
mode and to enter that mode you MUST  set the CFG3 pin (pin
5 of JP1 on the SilentStep Stick) to OPEN (as in open
Tristate logic). Again, read the TMC2100 spec sheet, page
23/24, which recommends a 1.5 amp motor or smaller be
used.
I don't think the datasheet is saying 1.5A stepper
motors are the limit there:  "The following  table 
shows  the  RMS  current  values  which  can be 
reached  using  standard  resistors  and motor types
fitting without additional motor current
scaling."

At
the bottom of the table it mentions using AIN current
scaling (the pot we adjust does this) for those examples
that were out of spec.


Howard Dutton
 

On Wed, Jan 17, 2018 at 02:42 pm, Mark Christensen wrote:
This formula is valid for the current adjustment mode and to enter that mode you MUST  set the CFG3 pin (pin 5 of JP1 on the SilentStep Stick) to OPEN (as in open Tristate logic). Again, read the TMC2100 spec sheet, page 23/24, which recommends a 1.5 amp motor or smaller be used.
I don't think the datasheet is saying 1.5A stepper motors are the limit there:  "The following  table  shows  the  RMS  current  values  which  can be  reached  using  standard  resistors  and motor types fitting without additional motor current scaling."

At the bottom of the table it mentions using AIN current scaling (the pot we adjust does this) for those examples that were out of spec.


JoAnn
 

On Wed, Jan 17, 2018 at 02:42 pm, Mark Christensen wrote:

GND: An internal (as opposed to the value you are reading on the pot) voltage is used with the sense resistors. With a .132 ohm resistor (0.11external + 0.020 on chip) that means your current should be limited (see page 23 of the TMC spec sheet).

 

Vio: Internal Sense resistors are used and the two external ones must be shorted as this mode requires that BRA and BRB (pins on the chip itself) be grounded. So unless you mod the silentStepStick you can’t use this mode. Vio is usually taken to be the voltage coming from an output port of the processor, not the V+ (5 or 3.3V) supply.

 

OPEN: The pot value for the voltage (that you are setting to 1V) and the external sense resistors (which happen to be 0.11 ohms) are used. Again, OPEN means floated tri-state which may or may not work with any particular processor. And without testing  one I wouldn’t know.

The CFG3 on the TMC2100 is always OPEN on a SSS TMC2100 (Watterott OEM.)  It has an open solder bridge that accomplishes this regardless of what is done to the CFG3 pin on the carrier.
OK, So I dialed down the pot to .22 roughly and that gives me 1.18 amp roughly. No heat issues at all even w/out heat sinks. Motors are turning fine though they are not under any torque. Is this OK? Is there a way to test them for torque if they are not physically connected to anything? I could get 2 new motors but rather not spend another $25. Also, note that I ma using this for a power supply: 

LE Power Adapter, Transformers, Power Supply For LED Strip, Output 12V DC, 3A Max, 36 Watt Max, UL Listed



Howard Dutton
 

On Wed, Jan 17, 2018 at 02:42 pm, Mark Christensen wrote:

GND: An internal (as opposed to the value you are reading on the pot) voltage is used with the sense resistors. With a .132 ohm resistor (0.11external + 0.020 on chip) that means your current should be limited (see page 23 of the TMC spec sheet).

 

Vio: Internal Sense resistors are used and the two external ones must be shorted as this mode requires that BRA and BRB (pins on the chip itself) be grounded. So unless you mod the silentStepStick you can’t use this mode. Vio is usually taken to be the voltage coming from an output port of the processor, not the V+ (5 or 3.3V) supply.

 

OPEN: The pot value for the voltage (that you are setting to 1V) and the external sense resistors (which happen to be 0.11 ohms) are used. Again, OPEN means floated tri-state which may or may not work with any particular processor. And without testing  one I wouldn’t know.

The CFG3 on the TMC2100 is always OPEN on a SSS TMC2100 (Watterott OEM.)  It has an open solder bridge that accomplishes this regardless of what is done to the CFG3 pin on the carrier.


Howard Dutton
 
Edited

On Wed, Jan 17, 2018 at 02:42 pm, Mark Christensen wrote:
1.       The motors you have chosen are rated at 2Amps, which is the above the RMS max  continuous rating for the TMC2100 (1.25amps RMS, 1.77 peak). In tracking mode you are possibly going to draw this. These are really beefy motors. Why? This choice will also seriously limit your battery life (if that matters).
The current rating is based on a voltage of 2.2V.  Power, the ability to do work, what heats the stepper motor, what creates the magnetic field is P=IV.  That is power = current * voltage.

All NEMA17 stepper motors with similar body length fall roughly in that power range regardless of the current rating.  There's two coils driven but micro-stepping drivers come in at 0.7 times (RMS) and you end up with a better approximation... 2+2=4... 4*0.7=2.8A...  2.8A*2.2V = 6.2 watts.  If the voltage goes up, to say 12V the current has to drop otherwise you're melting your stepper motor.  So wattage stays the same 6.2, voltage is known at 12, and current is 0.52A (and that's for BOTH coils the stepper driver spec. is per coil.As you can see that number is well within what the stepper driver can handle.  Really the stepper driver isn't going to get into trouble with any NEMA17 stepper motor, they just don't come close to it's power delivery capacity.
[I might be wrong about this so I'm retracting that in suspicion that the stepper driver Vref is like a power measurement?) Regardless, the SSS TMC2100 Vref setting method in the RepRapWiki has worked well during my limited testing.]


Mark Christensen
 

JoAnn,

The sense resistors are in two flavors, the two internal to the chip (the thing on other the overside of the heatpad 'vias' - which is what the little plated through holes are called in the trade) and the two external sense resistors (on the stepStick but external to the chip). By the setting of CFG3 (GND, Vio, or OPEN) you select which is used. To use the pot (the Vref adjustment) to set the current limit (in conjunction with the sense resistors) you use the OPEN configuration and the formula.

It is unfortunate that the manufacturers of the stepSticks provide little how-to data on configuring these. For the older models that used the Avago chips there was more configuration info available. Most of us have had to piece together the configuration from the schematics for the silentStepSticks, while referencing the TMC documents, which I appreciate most normal human beings will regard as being written in Greek. And then, like yourself, breadboard and experiment to verify our understanding. Add to that the confusion of how to achieve the OPEN tri-state condition and it take a while to get it right.

Mark C.
--------------------------------------------

On Wed, 1/17/18, <coachbosswife@...> wrote:

Subject: Re: [onstep] TMC2100 How hot?
To: onstep@groups.io
Date: Wednesday, January 17, 2018, 7:04 PM

[Edited Message
Follows]
Mark Christensen, Thanks
for the reply. Most of which I don't understand :) 
I believe I have the pin CFG3 open as I have not
connected it to anything (I am still breadboarding). I
really had no idea what I was doing when ordering stuff so
those are the motors I got. Is there a way to use these?
With the formula (Vref/sqrt(2))/.132 even if
I set the pot to .5v it is still over 2 amps. 
I don't understand how I select the sense
resistors. 

Name is JoAnn
BTW :)


JoAnn
 

Howard, I did read that but I have no knowledge of electronics. I only gathered from the video how to set the pot, not how much. The data sheets are really greek to me at this point. 


JoAnn
 
Edited

Mark Christensen, Thanks for the reply. Most of which I don't understand :) 
I believe I have the pin CFG3 open as I have not connected it to anything (I am still breadboarding). I really had no idea what I was doing when ordering stuff so those are the motors I got. Is there a way to use these? With the formula (Vref/sqrt(2))/.132 even if I set the pot to .5v it is still over 2 amps. 
I don't understand how I select the sense resistors. 

Name is JoAnn BTW :)


Mark Christensen
 

CoachBossWife (name?),

 

1.       The motors you have chosen are rated at 2Amps, which is the above the RMS max  continuous rating for the TMC2100 (1.25amps RMS, 1.77 peak). In tracking mode you are possibly going to draw this. These are really beefy motors. Why? This choice will also seriously limit your battery life (if that matters).

 

2.       The common references to setting Vref are questionable, see

http://reprap.org/wiki/TMC2100#Setting_the_reference_voltage_.2F_motor_current

 

As I read that web page (found by simply Googling), a 1v Vref will set the current limit at 0.71 amps. But then you should not be seeing much heat at all. I have driven TMC2100 silentStepSticks with 1.68 amp motors and while it gets warm it isn’t bad. I think I set the Vref to under .5V. It’s been a while.

BUT in any case the 1v Vref figure conflicts with the TMC2100 data sheet from TMC, which I would trust more than a general website that probably is cut and pasting from A4998 data.

 

First, according to the TMC2100 data sheet (TMC2100 DATASHEET (Rev. 1.07 / 2017-MAY-15, page 24) there are three options for the pin CFG3 that sets the current limit mode: GND, Vio, or OPEN. Getting the OPEN setting to work is tricky, as that usually requires you condition the IO port to be in the float tri-state condition and power Vio off of the output of a processor port pin. That is not always possible with some processors.

 

What is the difference between the three conditions?

 

GND: An internal (as opposed to the value you are reading on the pot) voltage is used with the sense resistors. With a .132 ohm resistor (0.11external + 0.020 on chip) that means your current should be limited (see page 23 of the TMC spec sheet).

 

Vio: Internal Sense resistors are used and the two external ones must be shorted as this mode requires that BRA and BRB (pins on the chip itself) be grounded. So unless you mod the silentStepStick you can’t use this mode. Vio is usually taken to be the voltage coming from an output port of the processor, not the V+ (5 or 3.3V) supply.

 

OPEN: The pot value for the voltage (that you are setting to 1V) and the external sense resistors (which happen to be 0.11 ohms) are used. Again, OPEN means floated tri-state which may or may not work with any particular processor. And without testing  one I wouldn’t know.

 

 

In summary,

 

If you have grounded the pin (5 on JP1) then you are using the internal (2.5V) value and the pot is ignored.

 

If you have tied the pin to +5 (or to a port set high) then the two sense resistors on the chip must be used AND you must short the other two (R4 and R2).

 

If you are able to float the pin (the OPEN mode) then you can use the pot as you want to.

 

The RMS current when turning (running, page 24) is give by:

 

(Vref/sqrt(2))/.132 when the CFG3/CSN line is OPEN, which should be the case with the 0.11 ohm (R0805/0R11) sense resistors (R4 and R2 on the SilentStepStick schematics).

 

That results in over 5 amps, so effectively you’re running at full current: 2Amps RMS which is not recommended, even with a heat sink, for continuous operation. And is probably not necessary. Plus, as I said above (1) these motors are probably oversized for your application and are more current than appears in the table on page 23 of the spec sheet.

 

This formula is valid for the current adjustment mode and to enter that mode you MUST  set the CFG3 pin (pin 5 of JP1 on the SilentStep Stick) to OPEN (as in open Tristate logic). Again, read the TMC2100 spec sheet, page 23/24, which recommends a 1.5 amp motor or smaller be used.

 

Hope this helps you.

 

Mark Christensen

 

 

 

From: onstep@groups.io [mailto:onstep@groups.io] On Behalf Of coachbosswife@...
Sent: Wednesday, January 17, 2018 11:33 AM
To: onstep@groups.io
Subject: Re: [onstep] TMC2100 How hot?

 

I have just adjusted them to 1v. TBH I have no idea what they should be set to but at 1v the motors turn and do not stutter. these are my motors https://www.omc-stepperonline.com/hybrid-stepper-motor/nema-17-bipolar-45ncm-64ozin-2a-42x40mm-4wires-w-1m-cable-and-connector-17hs16-2004s1.html. The driver still gets HOT so I am not going to mess too much until I get the heatsinks. 


Howard Dutton
 
Edited


JoAnn
 

I have just adjusted them to 1v. TBH I have no idea what they should be set to but at 1v the motors turn and do not stutter. these are my motors https://www.omc-stepperonline.com/hybrid-stepper-motor/nema-17-bipolar-45ncm-64ozin-2a-42x40mm-4wires-w-1m-cable-and-connector-17hs16-2004s1.html. The driver still gets HOT so I am not going to mess too much until I get the heatsinks. 


Terry Fishlock <telfish@...>
 

Have you adjusted the trim pots to the right reference voltage?

On Tue, Jan 16, 2018 at 10:05 PM, <coachbosswife@...> wrote:
Uhmm, disregard this ridiculous question :) I just bought some of the nice big blue heat sinks and get that they go on the gold top area not directly on the processor. Luckily I found a US seller on eBay. 



JoAnn
 

Uhmm, disregard this ridiculous question :) I just bought some of the nice big blue heat sinks and get that they go on the gold top area not directly on the processor. Luckily I found a US seller on eBay. 


JoAnn