LCD problems (solved) and Firmware Pin Mapping

[Tdeagan]

I went through 25+ hours of madness getting my firmware updated to the latest on my Printrboard Rev D. That's all working great. But when I plug in my Extrudrboard, I get the same problem as described in http://www.printrbottalk.com/forum/viewtopic.php?f=15&t=4063.

The bot will home the axis, but anything else gets:

Code: Select all

Printer stopped due to errors. Fix the error and use M999 to restart. (Temperature is reset. Set it after restarting)
>>>m999
SENDING:M999
Resend: 1
Error:2
: Extruder switched off. MINTEMP triggered !
Error:Printer stopped due to errors. Fix the error and use M999 to restart. (Temperature is reset. Set it after restarting)

If I plug in the LCD, the com port and connection disappear.

If I unplug the LCD and/or Extrudrboard the bot works as expected (though I haven't put it through a full regression test. Everything obvious works.)

I've watched Brook's video of setting these up and am following his pin connections. I've got a second PSU powering the Extrudrboard per spec.

Anyone have this working? Did it just work or did you have to do anything special?

[REPRAP SQUAD]

I finally got my firmware updated and finished wiring. I will be testing in the next day or so. Did you plug a thermisister into the empty 3rd extruder plug? Kf so you will need to.

[Tdeagan]

Interesting. A thermistor into the third extruder plug? I have my second extruder/heater/thermistor plugged into the first of the two connections (aka the second extruder connection set,) but nothing else but power and the ribbon cable.

Are there docs that describe the additional thermistor you're noting? Does it go into the 3rd thermistor socket? It's a two wire connection so it's hard to picture it going into the motor plug.

[REPRAP SQUAD]

Ya it goes into the 3rd extruders thermisister plug. Otherwise it tripps out and gives warnings. I believe Printrbot is going to add it to the kit. I know for the dual extruder kit production version not beta. It will be included and I would also assume it would be added to the extrdrboard kit as well.

[REPRAP SQUAD]

There are 3 videos that Printrbot posted on Twitter and 1 of them talks about it

[Tdeagan]

Thanks!!

I watched most of the second video (the parts that seemed relevant to just the Extrudrboard and LCD, but not the Printrbot stock extruder kit.). But I didn't watch the other two.

I get really impatient watching videos. They're so linear. My learning style betrays me with them. I can spend untold hours combing data sheets or reading manuals or web material, but watching a 3 minute video just kills me. I've been trying to learn Blender, but all the tutorials are videos and its like pulling teeth to sit through them. They're clearly where the world is going. Sigh.

I need to tie myself to a chair and toothpick my eyes.

[Tdeagan]

I watched the videos (which was like chewing glass.)

I burned the specific firmware (Printrbot-LCD-Extrudr-Firmware) in case the latest Marlin I was using was wrong.

I'm still struggling with a critical work/no-work app note like the third thermistor thing being tacked onto the end of a video. I only stumbled across those vids, it's not like they were listed on the product page as required. Seems like that would be something you would note on the product page or in a tech note sent with the product. Whatever.

I don't have a third thermistor (I've been a thermocouple guy up to now.) But it seems like a standard resistor would fulfill the same purpose. My intertoob search says it's an EPCOS 100k NTP thermistor, which suggests that I could use a 100K resistor across the third thermistor port.

Either way, attaching my LCD still causes the system to drop the com connection, so something is not right there.

I'll try out the dummy load on the third thermistor port and report back.

thanks again.

[REPRAP SQUAD]

Ya I agree with you. Supposedly the recent changes Printrbot has made is to get a handle on this sort of stuff. So hopefully they do. Thats a big reason for my blog, is to attempt to help and clear up as much confusion as possible.

[Tdeagan]

I whipped up a 100K Ohm resistor with the proper plug and plugged it into the third thermistor port. I still couldn't get the Extrudrboard to work.

I compiled and flashed the latest Marlin firmware, it's more recent than the special LCD/Extrudrboard version, says it supports the LCD and I wanted source.

I removed the dummy load and, for lack of a better term, 'fiddled' with the Extrudrboard, watching the VDC output of the heater (with Pronterface trying to set the T1 hotend to 185C.) And it started working.

So now, I have the Extrudrboard working, pushing my #2 (T1) extruder with nothing plugged into the third thermistor port.

However, if I connect the LCD, the Printrboard goes offline and won't provide a USB connection.

Progress, but still frustrating.

[REPRAP SQUAD]

I dont have the most recent firmware. I have the one that says something like dual e /lcd hex. They may have fixed the issue. They really need to get a quality guide figured as well as a dedicated firmware guy just for pb on mac and Windows. Too confusing

[Tdeagan]

I just traced every pin on the LCD and there aren't any shorts. It's a _very_ simple board and I'm really familiar with the standard 16x2 LCD.

When I plug it in, the backlight comes on, Row 0 and Row 2 go solid black. The printrboard loses it's ability to provide a COM port. It's kinda acting like it's drawing too much power, but I pushed the RAMPS 1.4, 2 ABS heated extruders and a bed at 100C plus the LCD controller off a single 350W power supply. Now I've got an additional supply, so I'm highly skeptical that this little LCD is pulling too much current without a short.

I have noticed that the Firmware has some odd pin assignments. Here's the section from the latest Marlin Printrbot firmware for the Printrboard:

Code: Select all

  #define LCD_PINS_RS 9
  #define LCD_PINS_ENABLE 8
  #define LCD_PINS_D4 7
  #define LCD_PINS_D5 6
  #define LCD_PINS_D6 5
  #define LCD_PINS_D7 4

  #define BTN_EN1   16
  #define BTN_EN2   17
  #define BTN_ENC   18//the click


The Printrboard schematics show that the connections are to the following pins on the at90usb1286:
RS 34
ENABLE 33
D4 32
D5 31
D6 30
D7 29

EN1 35
EN2 36
ENC 37 //the click

There must be some abstraction/redefinition going on someplace that I'm missing. But incorrect pin assignments could cause very similar symptoms. It would put it right on top of the USB pins of the aat90usb1286.

The other pin assignments for things that are working don't map to the actual at90usb1286 pins.

[RetireeJay]

There must be some abstraction/redefinition going on someplace that I'm missing.

Yes, I ran into that - and set it aside for later deep study.

In my case, I was wanting to remap the Z motor output to different pins in order to use one of the headers on the Prinrboard. But when I traced through config.h to what the existing pin assignments were, there was no correlation to the physical pins on the Atmel chip. And in the case of the motor control pins, there was not even the mapping of Port X Bit y... Totally confusing. Lwalkera or Pxt (I forget which) explained that the Arduino ecosystem uses an arbitrary higher-level pin mapping to make it "hardware agnostic." But then there's a confounding factor of the "fastio" add-on which apparently re-maps again. So you've got to have a pretty high tolerance for intricate mapping, and you've got to be able to track down documentation resources (pin mappings) that I can't for the life of me find anywhere.

Good luck, my friend! If you find the magic table (tables) of mappings, please share them on this forum.

[Tdeagan]

I'm with you. Funny you should ask that, I just spent the last 7 hours making the spreadsheet below.

This is based on the firmware at https://github.com/Printrbot/Marlin
Google Doc Version

pins.jpg

I still see potential problems with the LCD Enable and LCD RS mappings. I'm thinking they should be to 32 & 33, not 8 & 9

[RetireeJay]

For what it's worth, the Pins.h from the Marlin github does not agree with the numbering you have in your chart.

Here is the section that defines the Printrboard pins (there are several other sections for other boards, but this is the Printrboard section.)

Code: Select all

****************************************************************************************
* Printrboard Rev. B pin assingments (ATMEGA90USB1286)
* Requires the Teensyduino software with Teensy2.0++ selected in arduino IDE!
* See http://reprap.org/wiki/Printrboard for more info
****************************************************************************************/
#if MOTHERBOARD == 81
#define MOTHERBOARD 81
#define KNOWN_BOARD 1

#define X_STEP_PIN         28
#define X_DIR_PIN          29
#define X_ENABLE_PIN       19
#define X_MIN_PIN          47
#define X_MAX_PIN          -1

#define Y_STEP_PIN         30
#define Y_DIR_PIN          31
#define Y_ENABLE_PIN       18
#define Y_MIN_PIN           20
#define Y_MAX_PIN          -1

#define Z_STEP_PIN         32
#define Z_DIR_PIN          33
#define Z_ENABLE_PIN       17
#define Z_MIN_PIN          36
#define Z_MAX_PIN          -1

#define E0_STEP_PIN         34
#define E0_DIR_PIN          35
#define E0_ENABLE_PIN       13

#define HEATER_0_PIN       15  // Extruder
#define HEATER_1_PIN       -1
#define HEATER_2_PIN       -1
#define HEATER_BED_PIN     14  // Bed
#define FAN_PIN            16  // Fan

#define TEMP_0_PIN          1  // Extruder
#define TEMP_1_PIN         -1
#define TEMP_2_PIN         -1
#define TEMP_BED_PIN        0  // Bed

#define SDPOWER            -1
#define SDSS                26
#define LED_PIN            -1
#define PS_ON_PIN          -1
#define KILL_PIN           -1

#ifndef SDSUPPORT
// these pins are defined in the SD library if building with SD support
  #define SCK_PIN          9
  #define MISO_PIN         11
  #define MOSI_PIN         10
#endif
#endif


[Tdeagan]

Fascinating!

I took my Marlin Github pull from the Printrbot Github page's most recent link:
https://github.com/Printrbot/Marlin

Here's the pins.h section:

Code: Select all

/****************************************************************************************
* Teensylu 0.7 / Printrboard pin assignments (AT90USB1286)
* Requires the Teensyduino software with Teensy++ 2.0 selected in Arduino IDE!
  http://www.pjrc.com/teensy/teensyduino.html
* See http://reprap.org/wiki/Printrboard for more info
****************************************************************************************/
#if MOTHERBOARD == 8 || MOTHERBOARD == 81
#define KNOWN_BOARD 1
#define AT90USB 1286  // Disable MarlinSerial etc.

#ifndef __AVR_AT90USB1286__
#error Oops!  Make sure you have 'Teensy++ 2.0' selected from the 'Tools -> Boards' menu.
#endif

#define LARGE_FLASH        true

//Disable JTAG pins so they can be used for the Extrudrboard
#define DISABLE_JTAG       true

#define X_STEP_PIN          0
#define X_DIR_PIN           1
#define X_ENABLE_PIN       39

#define Y_STEP_PIN          2
#define Y_DIR_PIN           3
#define Y_ENABLE_PIN       38

#define Z_STEP_PIN          4
#define Z_DIR_PIN           5
#define Z_ENABLE_PIN       23

#define E0_STEP_PIN         6
#define E0_DIR_PIN          7
#define E0_ENABLE_PIN      19

#define E1_STEP_PIN        24
#define E1_DIR_PIN         25
#define E1_ENABLE_PIN      44

#define E2_STEP_PIN        26
#define E2_DIR_PIN         27
#define E2_ENABLE_PIN      45

#define HEATER_0_PIN       21  // Extruder
#define HEATER_1_PIN       46
#define HEATER_2_PIN       47
#define HEATER_BED_PIN     20  // Bed

// If soft or fast PWM is off then use Teensyduino pin numbering, Marlin
// fastio pin numbering otherwise
#ifdef FAN_SOFT_PWM || FAST_PWM_FAN
   #define FAN_PIN        22  // Fan
#else
   #define FAN_PIN        16  // Fan
#endif

#if MOTHERBOARD == 8  // Teensylu
  #define X_STOP_PIN         13
  #define Y_STOP_PIN         14
  #define Z_STOP_PIN         15
  #define TEMP_0_PIN          7  // Extruder / Analog pin numbering
  #define TEMP_BED_PIN        6  // Bed / Analog pin numbering
#else  // Printrboard
  #define X_STOP_PIN         35
  #define Y_STOP_PIN          8
  #define Z_STOP_PIN         36
  #define TEMP_0_PIN          1  // Extruder / Analog pin numbering
  #define TEMP_BED_PIN        0  // Bed / Analog pin numbering
#endif

#define TEMP_1_PIN         2
#define TEMP_2_PIN         3

#define SDPOWER            -1
#define SDSS                26
#define LED_PIN            -1
#define PS_ON_PIN          -1
#define KILL_PIN           -1
#define ALARM_PIN          -1

#ifdef ULTRA_LCD
  #define BEEPER -1

  #define LCD_PINS_RS 9
  #define LCD_PINS_ENABLE 8
  #define LCD_PINS_D4 7
  #define LCD_PINS_D5 6
  #define LCD_PINS_D6 5
  #define LCD_PINS_D7 4

  #define BTN_EN1   16
  #define BTN_EN2   17
  #define BTN_ENC   18//the click

  #define BLEN_C 2
  #define BLEN_B 1
  #define BLEN_A 0

  #define SDCARDDETECT -1

  //encoder rotation values
  #define encrot0 0
  #define encrot1 2
  #define encrot2 3
  #define encrot3 1
#endif

#ifndef SDSUPPORT
// these pins are defined in the SD library if building with SD support
  #define SCK_PIN           9
  #define MISO_PIN         11
  #define MOSI_PIN         10
#endif

#endif  // MOTHERBOARD == 8 (Teensylu) or 81 (Printrboard)


Which firmware version is your pins.h from?

[RetireeJay]

See this thread for a link to the version I was looking at... viewtopic.php?f=22&t=4427

...and a discussion of the difficulty of finding the right file and finding the right mapping... with no solution except go back and ask the experts! (I wanna be an expert too! *sniff, pout* )

[Tdeagan]

This is convoluted stuff. I spend much of my work-life futzing with issues like this. A few choice blog posts/notices on the Printrbot website would go a long way to clearing things up.

There are a couple contributing problems, the first of which is the number of places that state something like
"This version is for supporting Printrboard The copy at https://github.com/PxT/Marlin is the official source as shipped by Printrbot HQ."

That may have been true at a point in the past, but is (apparently) no longer accurate. In his videos, Brooks is promoting https://github.com/Printrbot as the official site.

The Marlin code has a gazillion forks. This is the network chart of the Printrbot site's code: https://github.com/Printrbot/Marlin/network You can see the fork by PxT a little ways down and by scrolling around you can see that PxT had commits as late as June 4 that haven't been committed to the Printrbot fork.

Both PxT and Printrbot were forked from Lincomatic which was forked from ErikZalm which is the top of the network (his network graph show the terrifying number of forks out there for the unwary.)

So it's a little weird that the PxT was the official code, but it's not where the new official code forked from. You can see who/when the contributions over the last month are coming from:
Printrbot contribs: https://github.com/Printrbot/Marlin/graphs/contributors
PxT contribs: https://github.com/PxT/Marlin/contributors
Printrbot is a lot more active and has lwalkera as a contributor. This thread notes him talking about the Printrbot fork http://www.printrbottalk.com/forum/viewtopic.php?f=22&t=2980.

Here's the comparison report between the two forks we're looking at, ots of diffs:(PxT:master & Printrbot:printrbot-prod)

This site helps a little https://github.com/blog/39-say-hello-to-the-network-graph-visualizer on explaining the network graph. It really only displays differences in code, not similarities, so it's not really a 'who-begat-who' chart. More of a 'how different have these forks gotten from each other chart.'

Warning: I'm a SVN user who s trying to wrap my head around Github and may be terribly terribly confused.

[Thew]

FYI. I have found the LCD I bought from Printrbot had the LCD wrong.
Should be a 5v LCD, Mine came with a 3.3v LCD. This caused all kind of problems.
see viewtopic.php?f=15&t=4903

Thew

[Tdeagan]

I also responded in Thew's thread.

He's dead on the money. The LCD part is a 3.3V LCD (I peeled the sticker with the partnum out with a dental probe.)

NHD-0420AZ-FL-YBW-33V3

Data Sheet

3.3V , 3.0V backlight. And this is _clearly_ being powered by a 5V pin off the Printrboard.

Aargh. (and great catch by Thew!)

[lwalkera]

The USB connection disappearing is more a result of Brook just ordering the cheapest LCD he could find, not checking the specs, and ending up with 3.3V LCDs. Fortunately, the only thing different is the value of the current limiting resistor for the backlight. You can disconnect LCD backlight and stop the overheating and browning out by unsoldering the solder jumper on the back of the LCD to the right(when the front is facing you) of the pin header.

[REPRAP SQUAD]

This kind of stuff really needs to be caught early on or at least have a page with workflow notations as far as what happened, how it was fixed etc. I am a die hard Printrbot guy all the way but, if this kind of issues aren't figured out soon I feel that a lot of people will go somewhere else. Printrbot says they are going towards a more user friendly bot so I guess only time will tell. In the meantime I guess I am learning a ton just trying to do basic modifications or at least what I consider, "basic" to be. Within the next few days my belts and pullys should be here and I will start printing with the dual extruder setup so ill keep you guys updated on the progress as I will be adding an lcd after I chase down the dual e issues.

[RetireeJay]

There are a couple contributing problems, the first of which is the number of places that state something like
"This version is for supporting Printrboard The copy at https://github.com/PxT/Marlin is the official source as shipped by Printrbot HQ."

That may have been true at a point in the past, but is (apparently) no longer accurate. In his videos, Brooks is promoting https://github.com/Printrbot as the official site.

The Marlin code has a gazillion forks. This is the network chart of the Printrbot site's code: https://github.com/Printrbot/Marlin/network You can see the fork by PxT a little ways down and by scrolling around you can see that PxT had commits as late as June 4 that haven't been committed to the Printrbot fork.

Both PxT and Printrbot were forked from Lincomatic which was forked from ErikZalm which is the top of the network (his network graph show the terrifying number of forks out there for the unwary.)

So it's a little weird that the PxT was the official code, but it's not where the new official code forked from. You can see who/when the contributions over the last month are coming from:
Printrbot contribs: https://github.com/Printrbot/Marlin/graphs/contributors
PxT contribs: https://github.com/PxT/Marlin/contributors
Printrbot is a lot more active and has lwalkera as a contributor. This thread notes him talking about the Printrbot fork viewtopic.php?f=22&t=2980.

Can you explain how different forks would use different "pin" numbers for the very same hardware? Especially in light of the fact that supposedly the next layer up in the logical chain is Arduino sofware which is "hardware-agnostic" meaning the same logical pin number always has the same function, no matter what hardware it's running on? So we have constant Printrboard physical hardware and constant Arduino logical hardware, but our pin numbers are still changing???? I don't get it. (But then, I've never programmed in Arduino. I've programmed in assembly language on the 6502 [the heart of the Commodore 64 and the Apple II], and I've programmed in several other languages, but never in C or its variants.)

[Tdeagan]

I'm with you on the confusion. I left a comment on the pins.h checkin that changed the pin numbers in the most recent printrbot firmware asking what abstraction layers are in play.

I not that lincomatic got frustrated on this topic and talked about it in his blog at http://blog.lincomatic.com/?p=537. Both of the forks we've been discussing are forks of lincomatic's code, so we may be dealing with the very issue his blog post is talking about (or not, I'm having a _very_ difficult time sussing this out.)

I'm now digging into the core_pins.h file that lincomatic talks about. It's not in the actual firmware source, it's in the arduino-????\hardware\at90usb1286\cores\at90usb1286 subdir that the teensyduino stuff sets up.

I'll add what I find to the matrix I'm building.

[Tdeagan]

I've added the core_pins.h definitions. Clearly a lot of those are in use. One explanation for the pin numbering switching we see is the use of different def files. But I'm still not clear on which ones/when/how.

The 'key' between the different definitions is the PIN, i.e. PINA0 is what I'm matching every instance of something referring to port A, bit 0 to, wherever something refers to it.

Google doc Version

pins.jpg

[Tdeagan]

The USB connection disappearing is more a result of Brook just ordering the cheapest LCD he could find, not checking the specs, and ending up with 3.3V LCDs. Fortunately, the only thing different is the value of the current limiting resistor for the backlight. You can disconnect LCD backlight and stop the overheating and browning out by unsoldering the solder jumper on the back of the LCD to the right(when the front is facing you) of the pin header.

This is incredibly helpful!!

Might have been nice to hear this from Printrbot before I repeatedly tried hooking it up to my bot and flashing/reflashing firmware...

I put in a broken part request to Printrbot. Given that it was a $65 component, I'm not over-eager to just fix it (or semi-fix) myself. A backlight is pretty useful on these displays. Almost every other LCD/controller board with identical functionality/parts on the market costs less than $55 and many less than $45.

• Option 1 - Wait for Printrbot to replace this with a viable 5V unit (Ouch, how long will this take?)
• Option 2 - See if Printrbot will send me a 5V LCD and I'll desolder/resolder it myself (Less ideal, but possibly a lot faster)
• Option 3 - Buy my own 5V LCD (or steal one from another project) and desolder/resolder it myself (Doesn't make me happy for the $65 I spent)
• Option 4 - try making a patch cable that lets me use my LCD controller from my RAMPS 1.4 (Still doesn't make me happy for the $65 I spent, but could hold me until Option 1 materialized)

None of these options are terribly palatable.

[RetireeJay]

I understand Port A Bit 0 etc, and I can see from the schematic and from the Atmel data sheet how that maps to a physical pin on the chip. That's the kind of programming I cut my teeth on back in the day.

And I understand the concept of - and the reasons for - a level of abstraction. I have used it myself.

But you still have to have unique identifiers, don't you? It looks like you have duplicates for most of the range 0 - 9

[Tdeagan]

Got my RAMPS 1.4 LCD working with the Printrboard by making a simple patch cable out of breadboard jumpers.



Patch list looks like:
PB.1-EXP1.9
PB.2-EXP1.10
PB.5-EXP1.8
PB.6-EXP1.7
PB.7-EXP1.6
PB.8-EXP1.5
PB.10-EXP1.3
PB.11-EXP1.4
PB.12-EXP2.5
PB.13-EXP2.3
PB.14-EXP1.2

That's the $45 board working where the $65 Printrbot board didn't.

[Tdeagan]

I understand Port A Bit 0 etc, and I can see from the schematic and from the Atmel data sheet how that maps to a physical pin on the chip. That's the kind of programming I cut my teeth on back in the day.

And I understand the concept of - and the reasons for - a level of abstraction. I have used it myself.

But you still have to have unique identifiers, don't you? It looks like you have duplicates for most of the range 0 - 9

I don't want to BS you, I'm nowhere near understanding how these defs are getting back to the actual physical pins. My fantasy is that if I spend ridiculous hours documenting everything I can find, I'll level up and get the 'clue' achievement. (Or someone in the know will take pity and just explain it )

[REPRAP SQUAD]

Quick question as lcd related things are by no means my strong part but, couldn't you buy an I2c lcd with 5v and wire it to work the same as buying a Printrbot one. I ask because I am looking for a cheap solution. I know you can buy them for less than 10 bucks and even get them with a built in sd card. Any info would be great.

[Tdeagan]

The firmware is set up to expect a 4x20 LCD (driven in 4-bit mode). I haven't looked at the other marlin options, there may be an i2c option in there, but it wouldn't be happy coming off the Printrboard ext2 pins without a lot of new code. It's also looking for a rotary encoder switch for menu selection.http://www.adafruit.com/products/377 The LCD and the rotary encoder could be bought for as little as $15 combined and breadboarded to work just as well as the Printrbot LCD controller.

I used a standard RAMPS 1.4 LCD smart controller with a very simple patch cord. I see those on the market for as low as $40. They work exactly the same as the $65 Printrbot LCD controller.

I am addicted to the LCD and controller at this point, having that control makes printing a good bit easier for me.

[REPRAP SQUAD]

Your talking me into it....I think I might just end up getting a ramps one since I have a ramps controller also. So I take it that its about the same setup using the ramps lcd as the printrbot one..?

[Tdeagan]

Check out my post above. I used some breadboard jumper cables and figured out the patch and it works great! Not all the ramps controllers are exactly the same, but they're all almost identical. All of them should be pretty close but I'd check the schematic to make sure.

The one I have is a clone of the RepRap discount smart controller http://reprap.org/wiki/RepRapDiscount_Smart_Controller. It includes an SD card slot, but I just ignored those pins.