Accelerometer Usage

gloomyandy

@tekkydave Have just tested mine and RRF 3.4 seems to think it is a LIS3DH, which is a pain as I really wanted a LIS3DSH to test with (I look after the RRF STM32/LPC port). Oh well looks like I need to try and order from a different supplier.#

tekkydave

@gloomyandy

Blacksheep99

@tekkydave said in Accelerometer Usage:

@gloomyandy I bought one from that seller. The board seems to indicate a LIS3DSH but the chip has 3CH on it which is a LIS3DH. Have you connected it up & what does the Duet report?

I bought two boards from different sellers. One should be a LIS3DH and the other a LIS3DSH. They look identical to me and both have 3CH on the chip.
The only difference is one has OED26 after the 3CH and the other OED27.
I have only connected the OED26 to the Duet so far.

They both look like this on the back:

I'm assuming that the Duet will report the correct chip.

Edit: I have soldered a connector to the other one (OED27) and tested it with the Duet. Also get same message as the other one
Accelerometer 0 type LIS3DH with orientation 14 samples at 1344Hz with 10-bit resolution, SPI frequency 2000000
Hmmmpphhh!

Edit:
Further tests connecting the device to an Arduino confirm it is a LIS3DH as it returns device id = 51 (0x33). So despite it being advertised as a LIS3DSH and the pcb indicating that it is - well it isn't. So be careful when buying these folks and check the code on the chip:
3CH = LIS3DH
DSH = LIS3DSH

I have ordered another one from the link given by @dc42 above. Let's see what turns up

I ordered one from the seller DC42 linked to and this is what I received.

undefined

tekkydave

@blacksheep99 should be a LIS3DSH from the chip id of DSH.

tecno

Got my 2 today from UK and they show up as:

M955 P0 C"io3.out+io3.in"
Accelerometer 0:0 type LIS3DSH with orientation 20 samples at 1600Hz with 16-bit resolution, SPI frequency 2000000

So what do I need to do next to get started? I have one for X and Z and the other for Y only.

jay_s_uk

@tecno if you have a bedslinger, then yes.
and then use the input shaper plugin to record the results

tecno

@jay_s_uk

Yes its my good old Chiron

Will try to see what I get here.

tecno

Re ORIENTATION

As i see this is orientation I20, right?

If I flip the board so VCC will be to the right upside down is the orientation then I60 ?

How do handle this in config.g ? 2 lines ?

M955 P0 I20 C"io3.out+io3.in" ;active
;M955 P0 I60 C"io3.out+io3.in" ; not active

tekkydave

@tecno Yes 60.

0 = +X, 1 = +Y, 2 = +Z, 4 = -X, 5 = -Y, 6 = -Z.

the chip's z will be pointing in printer's -z dir = 6
the chip's x will be pointing in printer's +x dir = 0

tecno

@tekkydave
Thanks, just want to be sure.

tekkydave

I have finally got my hands on a LIS3DSH. I ordered it from the link supplied above by @dc42

https://www.ebay.co.uk/itm/272706183865

Note the 'DSH' on the chip

2021-08-27 14.04.08_crop.jpg

And in DWC:

Screenshot from 2021-08-27 14-13-37.png

Blacksheep99

I got mine hooked up to the Maestro. Seems to be working. I've had some overruns so need to look at that. I'm not sure what to do with the data from the graphs as yet but reading it seems documentation is coming and the plugin will be updated to support 3.4b3 soon.

Does it matter where on the print head gantry I place the sensor? To test I just taped in on top of the extruder assembly.

tekkydave

@blacksheep99 I have mine mounted on the back plate of the extruder mount. It would benefit from being rigidly mounted rather than taped.

Blacksheep99

@tekkydave thanks. The tape is just while I know it works. I should be able to design and print a mount for it.

jbarros

I'm trying to connect accelerometers to my boards and on a 6HC and a Mini 5+ I'm unable to detect it.
Tried 3 accelerometers 2x LIS3DSH and 1x LIS3DH and 2 cables, one I built with 100mm and a short version with dupond cables.

Wasn't it supposed to receive some message with failure or success?
Ex:

Error: M955: Accelerometer not found on specified port

I get nothing when entering:

M955 P0 C"io2.out+io2.in"

Firmware: RepRapFirmware for Duet 3 Mini 5+ 3.4.0beta3 (2021-08-24)

GeneRisi

FWIW, twisted pair wiring is designed to use both wires in the pair. The signals we are carryng are not differential, so try connecting the "other" wire in the pair to ground at both ends of the wiring. You might find that it cleans up the signal quite a bit. I am using CAT5 unshielded cable for both the accelerometer and the Paneldue serial link without any issues (as far as I can tell)

jbarros

@jbarros said in Accelerometer Usage:

I'm trying to connect accelerometers to my boards and on a 6HC and a Mini 5+ I'm unable to detect it.
Tried 3 accelerometers 2x LIS3DSH and 1x LIS3DH and 2 cables, one I built with 100mm and a short version with dupond cables.

Wasn't it supposed to receive some message with failure or success?
Ex:
Error: M955: Accelerometer not found on specified port
I get nothing when entering:
M955 P0 C"io2.out+io2.in"
Firmware: RepRapFirmware for Duet 3 Mini 5+ 3.4.0beta3 (2021-08-24)

@dc42

I decided to test with 3.3 and I got an answer:

M955 P0 C"io2.out+io2.in"
Accelerometer 0:0 with orientation 20 samples at 1344Hz with 10-bit resolution

I was not getting a response on 3.4b3, either good or bad from M955 but I didn't actually check if data collection was working which it is, so it's just the missing response from M955 that's whacked

dc42

@jbarros I'm glad you solved it. The M955 command no longer returns a response if it is successful, just like most other G- and M-code configuration commands

jbarros

@dc42 said in Accelerometer Usage:

@jbarros I'm glad you solved it. The M955 command no longer returns a response if it is successful, just like most other G- and M-code configuration commands

Thanks, I was unaware of that.
What is the rationale behind this change?

dc42

@jbarros said in Accelerometer Usage:

What is the rationale behind this change?

It's normal that configuration commands when they succeed do not return a response. You can still get the details by sending M955 with ~~no additional parameters~~ just the P parameter.