Finally got to the hardware store to buy glass for my Anet A8 :) I sticks a lot better then to p...

[G+_Edvardas Rimkus]

Finally got to the hardware store to buy glass for my Anet A8 :) I sticks a lot better then to painters tape.

I have connected OctoPi to it and it seems that the print speed declined 1,5-2 times. Is there some kind of a setting i'm missing?

 

And another question to those who print in PLA. have you found anything to smooth prints (like acetone is used for ABS)?

[G+_Jeff Gros]

1. Speed of serial vs SDCard:

 

The LR/DR here is that according to tests done by Tech2C (second link below) he got a max of 45 mm/s on USB and 75 mm/s on SD Card before encountering slowdown.

 

I'm not too familiar with the A8, but if it is a Ramps/Rambo board, then OctoPi is connected to your printer through USB. The USB is then translated to serial via a serial to USB chip.

 

The serial connection is going to limit your throughput. You can try cranking up the baudrate, but keep in mind that the higher you go the more likely the data can get corrupted in between the processor and the serial conversion chip.

 

I haven't spent too much time looking through the marlin firmware to see how it fetches the next data. However, since you can send commands to the printer on the fly, while it is running, it seems pretty unlikely that any of the gcode commands are queued. This means that while it does take a lot of time for the head to move to execute a command (which would potentially swallow any comm delay if a queue was used), since there appears to be no queue, the comm delay will be present with each gcode command.

 

The SDCard comms are probably done using SPI bus on the order of MHz. This means that communication could be many times greater using the SDCard vs Octopi.

 

These videos were pretty interesting. The first one talks about print speed, and the second one addresses the comms issue. Watching the first video also shows the delay where the printer is processing the next command, which goes along with my theory of there being no gcode queue:

 

 

2. Smoothing prints:

 

I've never tried this, as I've never needed to smooth my prints. However, this link looked interesting. That's a shiny benchy!

 

rigid.ink - How to Smooth PLA to a Mirror Finish [step-by-Step]