Full size raspberry pi in place of zero?  

Re: Full size raspberry pi in place of zero?

Someone has tried yet to connect the RPi and external power supply at the same time? Does it work?

I'm using a RPi 3B+. My plan:

The details on power are:

By doing it this way, the PI is part of the printer, no extra power to plug in. I'm getting a pretty efficient 24V->5V step down with the UBEC (no heat problems of a linear regulator), it will power on/off with the printer, and I still get the protection of powering the RPi thru microUSB instead of the RPi's GPIO pins.

Notes / Extra Detail:
You can power the RPI from it's GPIO pins using the stock "servo connector" that comes with the UBEC. Search for "RPI ubec gpio for which pins, but the happen to be all in a row and are the same pitch as the UBEC connector".

If you choose another UBEC, make sure it is capable of handling 24V, in R/C or battery lingo it will be marketed as "6S" (or maybe 2S-8S, which includes 6S). 6S is a 6-cell LiPo battery, which has an expected range of ~22V-25V. Also it needs to be capable of at least 2.5A. That should be in the description.

If you choose another UBEC, make sure it outputs 5V, most of them do, but there are some 12V UBECs. Also, there are many off-brand crappy UBECs out there. To me, "crappy" means: "when you draw 2.5A from this, the voltage drops from 5V to below 4.5V" That HobbyWing one I linked to is well known and used by lots of folks.

If you decide to chose a linear regulator, instead of a UBEC, be aware that dropping from 24V->5V while drawing >2A is going to generate some heat at the regulator. It'll need a heatsink. It's also less efficient. But probably gives "cleaner" power to the RPi. Unclear if this matters.

Re: Full size raspberry pi in place of zero?

So why you are not taking in account using the RPi Port on the Einsy? Why you want to connect it with the USB instead?

Re: Full size raspberry pi in place of zero?

So why you are not taking in account using the RPi Port on the Einsy? Why you want to connect it with the USB instead?

It wasn't that I wanted USB vs the EINSY header pins. From what I can tell, the communication ends up being 115K serial anyway. No difference.

What happened in my situation is that I've read the Pi Zero has problems with OctoPrint printing while streaming video (which I'll do eventually) because of the single core CPU. I decided I wanted a RPi 3, for more CPU horsepower. Once I decided on RPi3, then I had to decide whether to use the GPIO -> EINSY connection. It just didn't seem worth it to me to try squeeze the RPi3 over inside the EINSY case or right next to it. I could have used extension cables on the header pins, but I didn't see the value. When I found that short USB cable, it all clicked into place for me.

The power solution I described above may seem like a lot of work, but for me it it trivial. From my R/C hobby, I have 5V UBECs lying around. I know they take 24V in because a "2S-6S UBEC" is a pretty common part in the R/C world. Wiring up the microUSB->DC barrel connector is also easy, I've done a lot of "random DC power connector to 2.1mm barrel" in the last few months, so it's seemed like the easiest way forward for me.

Re: Full size raspberry pi in place of zero?

Ok I understand.

My plan seems to be lil different. Anyway I haven't yet found an answer for the duel power supply:

- Raspi 3+ in this case ( https://www.thingiverse.com/thing:2862375 )
- Connect the Raspi Pins ( https://bit.ly/2OoiqJ9 ) with a jumper wire to the Eimsy RPi Port ( https://bit.ly/2NMHdpm )
- Connect external Micro USB Power Supply to the Raspi due to the fact I want to connect a USB cam to the Pi which can't be power by the RPi alone

Aynway till now I wasn't able to find out if it is possible to connect the Pi to the RPi Port without the +5V Pin and connect the Micro USB Power supply instead.

Re: Full size raspberry pi in place of zero?

Aynway till now I wasn't able to find out if it is possible to connect the Pi to the RPi Port without the +5V Pin and connect the Micro USB Power supply instead.

WARNING! I haven't done this, but what I describe below should work:
What you would do is just not connect the 5V+ and ground (RED and BLACK in the RPI image you posted) pins to the RPI. With those gone, the only way the PI could get power is from the micro USB.

I'm not sure if the data pins require ground (black) connected, someone more familiar with the Pi would know. But I bet not.

Re: Full size raspberry pi in place of zero?

WARNING! I haven't done this, but what I describe below should work:
What you would do is just not connect the 5V+ and ground (RED and BLACK in the RPI image you posted) pins to the RPI. With those gone, the only way the PI could get power is from the micro USB.

I'm not sure if the data pins require ground (black) connected, someone more familiar with the Pi would know. But I bet not.

Thats the problem.... nobody wants to try 😀

Re: Full size raspberry pi in place of zero?

i know this is a wordy first post, but anything to confirm and help. just take your time. its does work nicely. i will do a full write up with pictures on it soon...
i will confirm, using a RPI3 headless at the moment, works so much faster than the pi0w. (havent tested the rpi3b+ but i hear its not worth it for a couple of extra cycles and some conflicts)
key things you MUST note when connecting it. you are plugging EVERY matching pin EXCEPT the 5v pin. got that. you are NOT to connect the 5v pin to the PI/einsy. period.
your pins have to be the long type as stated for the pi0w becuase they go thru a board, and into a header on the other end. if its not long enough, contact will not be made or reliable to the einsy.
pins 6,8,10,15,16 ONLY are being connected. NOT pin 4 (stressing that for a reason)
match up the diagrams avilable for connecting the rpi0w, and then visualize removing that top pin (the 5v pin)
you must have pin 6 connected (ground) as it is a reference for the GPIO communication. when using other voltage sources like a RPi approved wall wart (what im doing), the grounding/reference is not the same on differing power supplies and information gets trashed .wont get overly technical but this is a common problem if you try to power neopixels with an external power source, and dont give the microcontoller a ground reference to communicate properly with, you are gonna have a bad time.
you will know your ground is not connected as the pi and printer toss gabled junk back and forth to each other and octoprint will tell you it doesn't understand what is being said.
i will investigate the prusa supply on my equipment and see if the ubec/buckConverter method is safe. honestly, id rather err on side of caution and not ask the power supply to push out a couple more amps beyond all the heating elements than it was designed for.
a final note. when connecting the pins, be aware your have to think of it all as kinda backwards. the pins from the rpi, when looking at it with the USB block facing you have even number pins on the right (outside), and odd on the left . when connecting to the prusa, you have to visualize that, with the printer facing you, its the holes on the einsy that are furthest away, take your time to visualize that. i actually did'nt when i first tried,
when plugging in to the einsy, if you are just using jumper wires to test , the two holes on the bottom should be pins 15 and 16. skip two places up (4 holes total), and match up 10.8.and 6 ...pull the sd card out of your pi0w ...and put it in your pi3, and it should just work. at least it did for me. (adjust your router settings to the new MAC address of the Pi3 if you use port forwarding or DHCP IP reservation to remote in (not recommended until octoprint adds some security options to block all access when not logged in).
PS, if you are having octoprint crashes on your external browser, check your plugins, i had one that was crashing often. removed it (title bar display) and have had ZERO problems since.

Re: Full size raspberry pi in place of zero?

i will investigate the prusa supply on my equipment and see if the ubec/buckConverter method is safe. honestly, id rather err on side of caution and not ask the power supply to push out a couple more amps beyond all the heating elements than it was designed for.

I can confirm using a UBEC works. I think it is safe, here's why:
Let's round up and say The Pi3 is going to use a max of 2.5 at 5V. That's the limit of the HobbyWing UBEC I linked to above. 2.5A * 5V = 12.5W. Let's round up to 15W to account for some inefficiency in the UBEC. So the input side needs to supply 15W from the 24V PSU. That's about 0.625A from the 24V PSU. That's not much.

Re: Full size raspberry pi in place of zero?

I have some nice results for you. I connected my Raspi 3+ via the Einsy Board with all Pins connected. The Pi is using Wifi connection. To have finally somebody on this world to give some real numbers, I measured the actual current using this connection method. One is plain and the other even a Logitech C270 USB Cam connected to the Pi. The current was measured between the Raspi 5V Header Pin and the Einsy. Here are the results:

Pi Wifi connection without Cam:
Boot 0,807A PEAK
Print 0,503A PEAK
Idle 0,404A AVG
Shutdown 0,906A PEAK
Stress Test 100% CPU 1,084A PEAK

Pi Wifi connection with Cam:
Boot 0,976A PEAK
Print 0,650A AVG
Print 0,740A PEAK
Idle 0,625A AVG
Shutdown 1,100A PEAK

With the cam I didnt do the stress test but actually it can be calculated. I didnt wanted to stress the board to much and to be honest for a unrealistic situation. You never going to have 100% CPU usage.

So I also guess should be no problem at all to connect and power the Pi3+ even with a USB cam connected.

Re: Full size raspberry pi in place of zero?

I measured the actual current using this connection method. One is plain and the other even a Logitech C270 USB Cam connected to the Pi. Here are the results:

Thanks! Where were those currents measured? Are on the 5V connection? or from the 24V PSU?

Re: Full size raspberry pi in place of zero?

Directly between the Raspi 5V and Einsy Board connection.

Re: Full size raspberry pi in place of zero?

Today i tried Rpi 3 on my printer. I can tell that speed comparing with zero is waaaay bigger. I will leave out power type question for now (i think i'll use 24/5V switch PSU module), but regarding connection - USB or serial thgourh pins - i noticed one important difference: if i connect Rpi via USB then printer restarts each time octoprint connects to it. This doesn't happen if Rpi connects to printer via serial - with this connection on "connect" only files on SD are re-arranged.

Lately i printed from SD and using octoprint only for monitoring. So, if octoprint accidentally disconnects my print will stop when i open web page and re-connect...
Ok, i know that correct way is printing from octoprint, but still..it's annoying issue which actually disables option to print from SD while monitoring state through RPi...

Re: Full size raspberry pi in place of zero?

Lately i printed from SD and using octoprint only for monitoring. So, if octoprint accidentally disconnects my print will stop when i open web page and re-connect...
Ok, i know that correct way is printing from octoprint, but still..it's annoying issue which actually disables option to print from SD while monitoring state through RPi...

I'm connected via USB. In my setup, I have OctoPrint setup to autoconnect on startup. When I open the web page, The OctoPrint server is already connected. It does not reconnect every time I open the web page. Also, it is my understanding that the Prusa folks are working with the OctoPrint folks to avoid this "restart printer on USB connect" issue. But that will be a future release.

Re: Full size raspberry pi in place of zero?

Your're absolutely correct, octoprint indeed connects at startup (ok, you have to enable that...). What i'm talking about is if, just in case, octoprint accidentally disconnects (say, rpi crash... anything) in the middle of the print. Then re-connecting will cause printer to reboot.

But, if you're correct that guys are working on this issue i'm somewhat satisfied.