EDGE behavior $DEFA=1 about g2 HOT 15 CLOSED

Carl, Thanks for finding this. While there is no persistence (yet) in G2, $defa should work. Or at least not kill the USB. Noted.

from g2.

I am still debugging the after-effects. While I am able to reconnect with CoolTerm and do $ and $$ and individual parameter looks, I can no longer reconnect with Chilipeppr - the initialization script that the JSON server sends seem to put tinyG2 in a state where all it wants to respond with is {"ej":""}. So it may not be the USB client itself. Could $DEFA=1 corrupt flash?

from g2.

There is no flash on G2. I'm not sure what state the various pieces are in. Can you check out the board with Coolterm and at least isolate that part?

from g2.

I meant Program flash, there is 512K of that, correct? The memory where tinyG2 executable is placed by BOSSAC is what i am refering to as flash. I understand there is no EEPROM. My assumption is that the settings.h (in my case, a custom set of settings) is loaded as part of the 'flashing' of the executable. I further assume that changes to parameters are held in RAM. Or, more likely(?), the settings.h setting are copied into RAM that mimics what EEPROM provided in V8(call it "pseudoEEPROM"), and could be changed by executed commands, but not be persistent.
As best I can tell, the CoolTerm $ and $$ command dumps are correct , meaning same as what I fed the compiler with the SETTINGS= directive.
As you probably know, the Chilipeppr JSON server sends a string of json commands to tinyG after it connects. It is after receiving and responding to that string of stuff that tinyG is no longer responsive to Chilipeppr serial port commands such as $ and $$.

Assuming my "pseudo EEPROM" guess is roughly right, perhaps DEFA=1 is corrupting some 'hidden' stuff that also resides in "pseudoEEPROM" ?

from g2.

Maybe I should add this:
In it's curent state, after a hard reset(power cycle or reset button)of the tinyGv9k, operation with CoolTerm looks normal, but only testing $ and $$ responses.
It no longer functions with Chilipeppr; after the json startup script is loaded and responded to by tinyG2, anything sent to tinyG2 from the Chillipeppr Serial Port command line gets the {"ej":""} response.

from g2.

Thanks for the explanation. Have you tried restarting the serial-port-JSON-server? You could also send some JSON to the v9 from Coolterm and see if it responds with correct JSON responses. Try {"sys":n} and {"x":n}

from g2.

Here is another look at the dynamics.
Method:

1. Chilipeppr and json server stopped
2. Reset button on tinyGv9k; motor light flash, heartbeat begins [normal]
3. Start json server with -V option (verbose logging to linux CLI)
   4 Open Chilipepper, select /dev/ttyACM0 as port for json server, Serial command window starts to fill
   

My summary:
json server sends 10 individual json parameter setting comands
In the Serial Port Console, they get green checks - I assume that means jason server is happy
The next command starting with "{"sr":{"line":true,"posx":true, "posy":true,"posz":true, "vel":tr" gets yellow acknowledgement, I assume server not happy
Server resends the whole cycle, starting with the 10 individual json elements, looks OK again
The line starting {"sr":{"line":true,"posx":true, "posy":true,"posz":true, "vel":tr does not get good response, again.
json server tries again - nothing gets green acknowledge (locked up?)

Here is a scrape of the CLI log from json server startup
https://dl.dropboxusercontent.com/u/50261731/LOG%20scrape%20-%20failure%20.txt

from g2.

I was typing while you sent, I'll retry with coolterm as you suggest.
As you can see from above, it is the multi-parameter line that chokes the interface

from g2.

I'm not familiar with the startup sequence for Chilipeppr, but it does not look correct if the board is returning "" for values in the responses. Can you see what you get back from Coolterm when you send {"sys":n} and {"x":n}? If these are not correct then it may require the board to be reflashed. That would be mu next step.

from g2.

Looks OK to me in CoolTerm, but I am not real conversant in json
((I assume it is normal to not get the "tinyg [mm] ok>" after the json request-response sequence))

tinyg [mm] ok>
$
[fb] firmware build 71.04
[fv] firmware version 0.97
[cv] configuration version 5.00
[hp] hardware platform 3.00
[hv] hardware version 0.00
[ja] junction acceleration 1000000 mm
[ct] chordal tolerance 0.0100 mm
[sl] soft limit enable 1
[st] switch type 1 [0=NO,1=NC]
[mt] motor idle timeout 2.00 Sec
[ej] enable json mode 0 [0=text,1=JSON]
[jv] json verbosity 4 [0=silent,1=footer,2=messages,3=configs,4=linenum,5=verbose]
[js] json serialize style 1 [0=relaxed,1=strict]
[tv] text verbosity 1 [0=silent,1=verbose]
[qv] queue report verbosity 0 [0=off,1=single,2=triple]
[sv] status report verbosity 1 [0=off,1=filtered,2=verbose]
[si] status interval 250 ms
[gpl] default gcode plane 0 [0=G17,1=G18,2=G19]
[gun] default gcode units mode 1 [0=G20,1=G21]
[gco] default gcode coord system 1 [1-6 (G54-G59)]
[gpa] default gcode path control 2 [0=G61,1=G61.1,2=G64]
[gdi] default gcode distance mode 0 [0=G90,1=G91]
tinyg [mm] ok>

tinyg [mm] ok>
{"sys":n}
{"r":{"sys":{"fb":71.04,"fv":0.970,"cv":5,"hp":3,"hv":0,"ja":1000000,"ct":0.0100,"sl":1,"st":1,"mt":2.00,"ej":1,"jv":4,"js":1,"tv":1,"qv":0,"sv":1,"si":250,"gpl":0,"gun":1,"gco":1,"gpa":2,"gdi":0}},"f":[1,0,9,4505]}
{"sr":{"posx":0.000,"posy":0.000,"posz":0.000,"posa":0.000,"feed":0.00,"vel":0.00,"unit":1,"coor":1,"dist":0,"frmo":1,"stat":1}}
{"x":n}
{"r":{"x":{"am":1,"vm":16000,"fr":16000,"tn":0.000,"tm":280.000,"jm":5000,"jh":10000,"jd":0.1000,"sn":3,"sx":2,"sv":3000,"lv":100,"lb":20.000,"zb":3.000}},"f":[1,0,7,325]}

from g2.

I have held off re flashing in case there was something useful to find in here.

from g2.

I am really stumped now. I tried my Arduino Due board, and it reacts the same as the tinyGv9k.
I just posted a question to JohnL to see if by chance the initialization sequence changed since yesterday.
As with too many things these days, I am not sure where that script sequence originates from.
John just produced two very useful fixes/enhancement to the tinyG fiddle, almost as fast as he was asked.
That's the good news.
But what version of Chilipeppr am I running at his instant? I have no idea.

from g2.

You are getting all the right responses to the JSON. Somethings's probably up with the communication with serial port JSON server. I don't think you need to reflash.

from g2.

The short answer is http://chilipeppr.com/tinyg?v9=true runs on G2
platforms.
The server initialization script is modified.
Details soon.
And of course, why it ran OK yesterday is a mystery.

On 12/28/2014 07:20 PM, Alden Hart wrote:

You are getting all the right responses to the JSON. Somethings's
probably up with the communication with serial port JSON server. I
don't think you need to reflash.

—
Reply to this email directly or view it on GitHub
#38 (comment).

from g2.

Eliminate the impossible, and whatever is left, however improbable, must be the truth. Good to know.

from g2.