Repeter just stops

arturnoole163 wrote:

Hi,

I have now been using Repeter for a week and I have developed a problem. I can print just fine, it connects to reperter host when all of a sudden buffer goes to 0 and print stopps. Repetere stays connected to the repeter host. Is my bautrate too high (250000)?

I have tried everything to fix the same issue... I ended up just putting the gcode on the SD card and printing from the SD Card on the system LCD. I got an extension cable and mounted my SD card inside the cabinet so I could get to it easy.

I was doing some testing when I was having the issue and it seemed that the higher your comport speed, the sooner the same print would stop. Based on that it seems there is a bug of some kind between the firmware, and the data flow. To me it seems that the host is sending data to the printer as it normally should. However Repetier host does not work the same as the stock software. With stock it sends the complete gcode to the printer where it is saved on the SD card and printed from there. You can actually unplug the usb and turn off the PC and the printer will still complete the job.

In the case of Repetier, the software is supposed to send G-code to the printer until the printer says hay wait my buffer is full. The software wil then wait till the printer says ok send more code. With this setup you have active two way comm between the printer and the host until the file is done. If the usb is unplugged or the computer is shut down the job will only run to the buffer is empty.

What I think is happening here is on the buffer is only 32K, I need to look and see if that can be changed but I suspect it is a hard limit of the processor. The Host is sending code to the printer but the printer is not telling the host when it's buffer is full. Instead the host just keeps streaming the code to the printer until the host sees the job is done and goes idle. All the code that is sent that can fit in the buffer is lost. The buffer is first in, first out so naturally it will print properly until the contents of the buffer is gone. Since the data that should have gone in was just streamed in to digital oblivion due to the printer not sending back wait or ack commands the host is just doing what it is supposed to. At first I was not having this issue and it could be related to the fact the EEPROM is on the SD card. There may be some issue where it causes something to be corrupted in the firmware if you write something else to the SD card as I did. I put a gcode on it as I wanted to see if it still printed from the card. After that was when I started seeing these stalls.

The reason this makes since to me is as I said it seems to happen faster at higher port speeds. Lowe port speeds seems to last longer. The only way around this is to set it to such a low port speed if possible is so that the data is used real time and never goes to the buffer. The only other alternative would be to increase the buffer size so that the entire gcode could be placed in it.

As I said my work around was just to put the file on the card and print from the printer LCD. It has no issues with that and completes every time.