Firmware 2.02RC6 released

lolorc

@dc42 said in Firmware 2.02RC6 released:

@lolorc said in Firmware 2.02RC6 released:

Hi there,

I'm still not able to upload files to my duet wifi, I'm using latest firmwares, duetwifiserver is still the old 1.21.
Still using cisco access points and linux computers, still loads of tcp dup, tcp retransmission, tcp window full...
and no single issue with a raspberry 10cm away connected to the same ap...

I guess it's not enough just to use a esp8266 and expect it to work in a lot of different environments.

I've been uploading gcodes and firmware through the usb serial port for a while now, not the end of the world...

Happy Xmas.

I'm sorry to hear you are having problems with WiFi connectivity. It may be that the ESP12S module on your Duet is faulty.

it may be, that something I haven't though about. the board connects to the access point fine, I'm able to use the web service, the issues happen when uploading files.
I burnt my first duet board, this is a my second one, it's true I have never encoutered this issue with the first board. but it was 20 months ago...
Have you ever heard of that kind issues that were related to the esp itself ?

Best wishes.

dc42

@lolorc said in Firmware 2.02RC6 released:

it may be, that something I haven't though about. the board connects to the access point fine, I'm able to use the web service, the issues happen when uploading files.
I burnt my first duet board, this is a my second one, it's true I have never encoutered this issue with the first board. but it was 20 months ago...
Have you ever heard of that kind issues that were related to the esp itself ?

We know that not all ESP modules work perfectly during testing, but our testing isn't extensive enough to pick up all possible faults. However, another possibility is that your SD card has a very long write time. Have you tried a new SD card?

dc42

@frafa, please provide the config.g and homing files you are using, and confirm that the issues you have with the 2.02 firmware release are still as below. That error message should mean that from the current position, it is not possible to move in a straight line to X150 Y150, which is normally because a straight line to that point would pass too close to the proximal joint. You should still be able to use a G0 move to reach that point, assuming that point is reachable.

After print i test this macro file content to debug:

G28
G90 ;Absolute Positioning
G1 X150 Y150 F2500····
Return error on G1: M98 P"0:/macros/test" Error: G1/G2/G3: intermediate position outside machine limits

Second test:

G28
G90 ;Absolute Positioning
G1 X150 Y150 F2500
G1 X0 Y0 F2500
Error: G1/G2/G3: intermediate position outside machine limits 16:24:50 M98 P"0:/macros/test" Error: G1/G2/G3: intermediate position outside machine limits

the arm does the homing, then it makes a mistake, it moves in relative coordinates and not absolute in x150 / y150 (partial ok, not absolu)

but afterwards it is supposed to go to x0 / y0, and it goes to x0 / y300 ???

I think the concern is there is an inverted axis?

frafa

Hi David,
yes same problem,

now homing is ok
G28 no longer returns errors

But move absolute return error ...

Test Firmware Version: 2.02RC7(RTOS) (2018-12-22b2):

G28
G90 ;Absolute Positioning
G1 X0 F2500

Return error: 15:40:30 M98 P"0:/macros/test" Error: G1/G2/G3: intermediate position outside machine limits Error: G1/G2/G3: intermediate position outside machine limits

all movements in absolute mode return an error x/y/z...

My config and homing files:
config.g http://dl.free.fr/j7UVd31UF
homeall.g http://dl.free.fr/usibxoSpf (hominig error on "G1 Z10 F250" if i remove no errors)

Reminder everything works perfectly with 1.21RC3

dc42

What happens if you make a small G1 S2 move after homing (e.g. at the end of the homeall.g file) to bring the distal arm off the endstop a little? What I suspect is happening is that the end of the first segment of the move you are trying to make ends a little inside the minimum radius. The minimum radius is set slightly higher than the actual minimum radius when the distal arm is on the endstop, to allow for rounding error in the calculation.

Older firmware versions didn't check for intermediate points in a move violating the minimum radius, so for example they would let you attempt to move the nozzle right through the proximal axis, as long as the end point of the move was reachable.

frafa

Hi David,

Very strange behavior if I use G1 S2 after homing ...

Example with G1 S2 X1.0 F2000
the arm moves to X: -19.66 Y: 279.36 (while the homing position is X: 30.84 Y: 51.59)

After I send:
G1 X10 F2500

Do not move and forward error:
G1 X10 F2500
Error: G1 / G2 / G3: Intermediate position outside machine limits

The Homing position is inside the bed, (Homing: X: 30.84 Y: 51.59)
the problem would not be there?

dc42

It's the distal arm you need to move off the endstop, so in the small G1 S2 movement use Y instead of X. Also, use G91 before it and G90 after it.

frafa

HiDavid,

if I execute this macro file:

G28
G91
G1 S2 Y1.0 F2500
G90 ;Absolute Positioning
G1 Y10 F2500

Return:
M98 P"0:/macros/test" Error: G1/G2/G3: intermediate position outside machine limits

homing ok, movements in S2 ok,
but the command G1 is not executed
and after more nothing answers
no more absolute or relative displacements ? ...

if I run a G28 I can move again in relative movements
but not in absolute movements ...

dc42

I presume you are homing the distal axis to a minimum angle endstop. Is that correct? If so, please try Y2.0 or Y3.0 in your macro instead of Y1.0. If you are homing distal to max angle, use Y-1.0 or -2.0 instead.
if that doesn't help, please provide your config.g and homing files, and I'll run them on my bench system.

frafa

Files are just on post from above:
https://forum.duet3d.com/topic/8155/firmware-2-02rc6-released/17

Files:
config.g http://dl.free.fr/j7UVd31UF
homeall.g http://dl.free.fr/usibxoSpf (hominig error on "G1 Z10 F250" if i remove no errors)

dc42

I've just had a thought. Your homing position is probably outside the M208 limits you have set up, and it's probably checking the intermediate positions against the M208 limits. Please send the following command to disable the M208 limits:

M564 S0

and then try again. You may still need to do the small G1 S2 move to move away from the distal endstop.

Also, please try a G0 move after homing. You shouldn't need to use M564 to disable limit checking this time as long as the endpoint of the G0 move is within the M208 limits,; and you shouldn't need the small G1 S2 move either.

frafa

humm ... I can not do anything good (humm google translate sorry ...)
G0 X0 the arm went upside down and blew the belt
I repair the mechanics: p

frafa

Tests...
Unlock
M564 S0

G1 Y300
ok

g1 x10
ok

g1 x0
ok

G1 Y0
Error: G1/G2/G3: intermediate position outside machine limits

rehoming
g28
ok

g1 x300
ok

g1 x100
ok

g1 x30
ok

g1 x20
no error but not move ...
Apparently impossible to move X below 30?

Mangy_Dog

RC6? What? i thought you already released the full update final?

dc42

@frafa, I think this issue is this:

In firmware 2.01 and earlier, when you asked to do a G1 move, RRF checked only that the end point of the move was within machine limits. In firmware 2.02 it checks that every intermediate point in a G1 move is within limits, to make sure that it can execute the move in a straight line (which is what G1 means).

Typically, the XY homing position of the machine will be outside the machine limits set by M208. If this is the case, a G1 move executed when the proximal and distal arms are in the homed position will fail (unless you use M564 S0 to disable the M208 limits).

Therefore, after homing the machine, you should execute a G0 move to an endpoint that is inside the M208 machine limits. You can do this within the homeall.g file. If you are using a Z probe to do Z homing, then in homeall.g it's probably best to home proximal and distal arms, then do the G0 move to a suitable position (e.g. the centre of the printable area defined by M208), then do the G30 move to home Z. That way, you will be homing Z at a position within the normal print area, which is usually preferable to homing Z at a position outside the normal print area.

Alternatively, you could use a G1 S2 move after homing to move the head to a position within the M208 XY limits. But a G0 move is easier.

frafa

I read again tomorrow, not having understood everything ...

Example M564 S0 then G1 X60 then I made a G1 Y300, it does not make a straight line, in the center it makes a curve ...

dc42

@frafa said in Firmware 2.02RC6 released:

I read again tomorrow, not having understood everything ...

Example M564 S0 then G1 X60 then I made a G1 Y300, it does not make a straight line, in the center it makes a curve ...

That probably means that a straight line movement between the initial and final positions that stays within the M208 bounds isn't possible.

frafa

if I do the same thing with version 1.21RC3
I move well in a straight line ...
I just tested

I have about an area of X60 Y300 that does not seem usable with the new firmware
While nothing can stop you from reaching this area ...

after M564 S0
if I move in G1 X100
and that after I made a G1 Y300 there I have a straight line with the new firmware

frafa

Hello David,

My simplified kinematics
http://www.openhardware.eu/tmp/scara-kinematics.png

With M564 S0 G1 the red area is not accessible.
G1 The orange area does not make straight lines.
G1 only the green area is normally accessible with M564 S0...

I have 1/3 of the surface of the bed that is not accessible ...

JamesM

I was having a similar homing issue as well with my Taz6 printer after installing the duet wifi and installing the newest firmware. I don't know if it was the right thing but I just increased the negative move distances in the home x and home x files larger than the total possible movement of the axis. Seems to have fixed the issue.

EDIT: I spoke too soon. Last night when I went to start a print my Taz6 went to do the initial homing of all axis and it failed. The issue always seems to be with the y-axis not making it to it's endstop and then a failure stating insuffient axis homed.