New belt status test, completely different belt values.
 

New belt status test, completely different belt values.  

  RSS
vandenkeere
(@vandenkeere)
New Member

Hi,

 

I flashed the new firmware 3.9.0-3421. Just before flashing it, I run a couple of seltest, and looked at the belt status values. Immediately after the flashing, I run the new standalone, belt status test.

Before the flashing, the results were X-axis 284, Y-axis 268

After the flashing, the results are X-axis 239, Y-203. I run this test a couple of times, always with similar +- 4 results.

Now I’m confused about this big difference.

First the belts were quite loose, no they are overtighten, which values should I use?

Thanks.

Posted : 20/05/2020 1:35 pm
vintagepc
(@vintagepc)
Noble Member

Do you also get the same numbers with a full selftest on the new firmware?

Posted : 20/05/2020 3:17 pm
vandenkeere
(@vandenkeere)
New Member

Yes, self test or standalone belt test delivers the same result

Posted : 20/05/2020 3:36 pm
vintagepc
(@vintagepc)
Noble Member

Weird, there were experiments with changing the speed used for the test (which affects the numbers) but these were not in the final version that was merged for this reason.

When I tested it on my own printer prior to it being merged the numbers before/after were the same... maybe something else changed afterwards.

/shrug. 

Posted : 20/05/2020 5:02 pm
vandenkeere
(@vandenkeere)
New Member

I adapted my belts to the new values, I had to shift several teeth in the belt, so it is really a very different setting. Now I’m printing a large object, to see how it works. Seems to work fine for now. 

Posted : 20/05/2020 5:17 pm
vintagepc
(@vintagepc)
Noble Member

keep in mind those numbers measure motor load and not actual belt tension. There are a lot of things that contribute to motor loading such as bearings and speed. They can be an indicator of belt tension issues, but should not be taken as canon for that. 

So if your prints were fine before, I would not take too much stock in the actual numbers.

For example, mine generally read around 210-220 because I have 0.9 degree steppers. 

Posted : 20/05/2020 5:46 pm
chocki
(@chocki)
Prominent Member

Is the higher the number, the higher the current draw?

Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet....
Posted : 31/05/2020 10:40 pm
chocki
(@chocki)
Prominent Member

With 0.9 steppers on X and Y I have X 167  and  Y 186  Measured skew 0.01 if this makes any difference.

Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet....
Posted : 31/05/2020 10:46 pm
vintagepc
(@vintagepc)
Noble Member

The lower the number, the higher the current draw. 

The belt test value comes directly from the "Stallguard" register in the TMC2130 - and the other use for this register is the crash detection. 

Essentially, if the value in the register is 0, then the DIAG (crash detect) pin is set high (and your printer detects a crash). 

(Ergo, tuning stallguard/crash detect tweaks will also affect the belt tension numbers, and you want to tune SG such that your register is a ways above zero unless it is obstructed. I think this is why PR has the suggested range value; if you are too far above, CD may not work reliably, and if you are too low then you will get false crash senses.)

Posted : 01/06/2020 2:25 pm
chocki
(@chocki)
Prominent Member

Thanks vintagepc. I'm running the latest 3.9.0 but with the TMC setting and skelestruder alterations from Guys 0.9 branch ported over since he's too busy with important work for now which is much more important than maintaining firmware. 

This post was modified 4 months ago by chocki
Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet....
Posted : 01/06/2020 3:15 pm
gmurad
(@gmurad)
Active Member

I can confirm that the numbers being outputted are much lower now than before the firmware update, without any change to the printer. I know this because I spend a lot of time doing belt tension and running the self test just before the update and I had reached a stable point with both X and Y very close to 260. Now I'm getting values around 200.

Posted : 02/06/2020 3:46 pm
chocki
(@chocki)
Prominent Member

Tweaked my firmware and belts and now have X 198  and Y 214 yet at the end of the Belt Test it shows selftest failed which it also did previously, yet if I carry out full Selftest which includes fans homing etc, everything passes.

Any idea what is going on? and what should I check?, anything in particular from the pronterface output during these tests? or where in the firmware does it compare values returned during testing against what is supposed to be correct.

 

Thanks

Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet....
Posted : 04/06/2020 11:57 am
vintagepc
(@vintagepc)
Noble Member

The belt (axis) test fails if the length of the axis doesn't match either what's expected or the value from the two measurement passes it does (i.e. value from left dir != value from right dir). 

Usually that means belt slop or a loose pulley. I don't know what that means in the context of Guy's firmware, given his stallguard tweaks. I am running stock with my own 0.9s. 

Posted : 04/06/2020 2:28 pm
chocki
(@chocki)
Prominent Member

That is fine and I can see the results and if the value is too high the printer will fail the self test with a loose pulley message, I have tested this by slackening the tension, but this error is strange, the self test runs, completes without any issue waits approx. 30s I guess then states self test failed. This only occurs in the separate Belt Test routine, but does not fail during the complete self test routine which includes the belt test as part of it.

It's not actually stopping me printing, just seems like a glitch at the end of the belt test routine.

Any pointer to where the test routine is in the code so that I can debug this myself please.

I'll try stock stall guard settings tomorrow.

Thanks.

This post was modified 4 months ago by chocki
Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet....
Posted : 04/06/2020 4:59 pm
Share:

Please Login or Register