Extruder board not driving stepper

Hi all,

I'm almost there, on the home straight with a nicely moving Reprap in the three axes. Thanks for everyones help so far.

I'm stuck good and proper now with the extruder. Most issues I've been able to resolve with a little help or a good rummage around these posts but I'm at a deadend getting my extruder going.

Here is what I have .....

Reprap with Gen3 electronics
Geared Nema17 Extruder
Reprap-host-linux software, version 20081218. Tried the latest version but nothing worked.

I purchased the electronics with the firmware pre-installed so I don't know exactly what is on there. I've tried uploading the latest firmware but I'm also struggling with error messages that so many have had. I will be working to resolve this in the meantime.

However, as the 3 axes are working fine I would expect the extruder board to move the stepper ok. I have a thermistor and nozzle attached and both are working fine. The only oddity is that the nozzle heats up extremely quickly, within a matter of several seconds it has hit almost 190 deg C - is this normal!?

I have checked the connections to the extruder board, used different cables, checked I have everything plugged in, including the SCA/SCL to D9/D10. I just get no movement from the extruder stepper.

I 'occasionally' get a whining noise from the motor but it isn't the same as the other motors, much higher pitch. I have never had the motor engaged such that I can't move it by hand. There are 3 sets of connections on the extruder board to plug in the SCA/SCL, i've tried all 3 but not actually sure which is the correct one. Any tips please?

I've tweaked the potentiometer with no effect.

One strange point to mention. I dropped my multimeter over the connections to the stepper. (I've tried reversing polarity also). My x and y steppers give me approx 12v when static and my z gives me zero volts even when moving (magic?). The extruder stepper has given my 3 different readings .... 0v, 0.25v and 3.5v ish, depending on which time i have tested it. What should I expect to see?

Any tips on how to debug this would be greatly appreciated.

Thanks
Matt

I've confirmed it is not the stepper motor by driving it off the z axis board.

I thought I spotted somewhere that you can test the extruder board by connecting it to one of the 10 pin connectors for the x, y or z of the motherboard. I can't find this anywhere! Any tips?

Thanks

I dont really have a clue with the temp rise that fast, but maybe the heater is too close to thermistor so thermistor just reads the heather heating up instead of actually reading the barrel temp. I think it should be a couple of minutes, not few seconds, ofc it depends on setup / heater etc, but still a matter of minutes.

I think sda is step and scl is dir but better double check that. Have you checked the sda/scl lines with a multimeter, from the MB chip pins to the extruder mcu pins? Driving stepper from the a3949 h-bridge is said to be a nasty hack, and usually those chips burn pretty easy (probably should of been a3950 for more protection but that package was smaller pitch). So, maybe one of the a3949 chips are burned. Also put the extruder board pot at half and dont go over it else you may fry one of the a3949 as i did. Also make sure there is a firmware that has that hack on it, for the extruder board to actually run it. Dont need motor when testing, just look at the leds if they light up to indicate movement (if leds dont lit up, there is no current, then the motor cant possible move either).

Better yet use a 4th stepper boards and tie its enable to low/gnd (if needed, depending on chip), and put the step/dir where should be on the 10 pin connector, then you have a kind of better / proper setup.

Hi Noobman

Thanks for your input, greatly appreciated

There is a distinct possibility that I may have fried a chip as the pot will have been cranked up beyond 50% at some point during my desperation to bring it to life. Is there an easy way to confirm that I have done damage to either of these chips .... diagnosic tools limited to a cheap multimeter.

You knowledge of these a3949 and a3950 chips is way beyond mine and i may have missed a trick here. You mention 'make sure there is a firmware that has that hack on it, for the extruder board to actually run it.', is it this you are referring to?

If I either get my existing board running or confirm a fatality, have you seen anything online to get me going using a 4th stepper board. This would be preferable as they appear to be half the price of an extruder board. The thermistor I can read using other sources but i'm not sure about the control of the nozzle heater.

Sorry about more questions and thanks again
Matt

As i understand that you are not sure if the firmware on the board actually has the hack on it, or a way to upload other firmware, nothing is certain. Maybe the two a3949 have some "blowing" marks on them - or maybe you remember if they made some smoke at any point. If none is true, there is no need to call a blow component yet. Even so, determining exactly which one needs to be replaced may not be so easy if no clear indication can be made.

My top a3949 blew when i squeezed the pot about to 90%, but i dont think it will be proper to generalize that, as i did that board myself from a copper cald and even more, i had replaced some component values with some i had on hand.

Btw, dont overappreciate my knowledge which is rather thin .

There are some ways to locate the issue tho.

First, to rule out the mb, you can connect the sda/scl lines to a regular stepper board, with half of a 10 pin idc cable that has 10 pin idc on a side, and 3 pins 0.100" on the other side. Two pins to connect the sda/scl to step/dir in the correct place (make sure to double check that), and the 3rd to connect the gnd (near sda/scl) to enable. Alternativelly you may choose to hack into the board to manually connect supply gnd to enable, but this is to avoid that. I dont know what chip is on your stepper board, but read its datasheet and see at Enable, how it works, in other words, does it need to be high or does it need to be low to have the outputs working. What i said assumes that for the outputs to work, the enable pin has to be low (to gnd) thats why you put enable to gnd. This would be the best setup to work the extruder motor with a 4th stepper board.

Second, to check if the extruder board works at all, has firmware, and does what is needed:
- if you put a thermistors pins into the terminals, does it report the temperature changes - like put your fingers on termistor and see if temp increases to your body temp or smth like that?
- when you turn on the fan from software, does a led light up on the 6 pin terminal outputs?
- same as the above, does a light come on when playing in software with extruder heater?

If all the above works, then you have a setup. Fan is optional anyway. Heated bed just needed with abs, and see the related wiki page. You can forget the a3949 at all, burned or not, and just use a 4th stepper board instead. AFAIK this is much better anyway.

If you dont want to go with a 4th stepper board, and you want to revive the a3949:
-metering lines for continuity (offline) the the step/dir pins from the MB mcu to the Extruder board mcu - all ok? on EC2.2 extruder board mcu, as you look with the crystal below the atmega168, its the 3rd and 4th pins on the right ic side (counting from top down). On the MB1.2 as you look with the sd card slot on the lower left, its the same, the 3rd and 4th pin from the right, counting the same, from top down.
-any signs of being burned on any of the two? tho it can burn without explicit marks and less noticeable smoke or maybe not at all, dunno
-do you remember any of the leds near the a3949 ever changed on/off when they received instructions to move the extruder?
And nothing else comes across my head further on. You could take a look at a3949 datasheet tho maybe some inspiration comes.

Edited 1 time(s). Last edit at 01/19/2011 11:35AM by NoobMan.

you are a gent ..... plenty for me to explore.
Thanks

to make the 10 pin idc to 3 pin .100" / 2.54mm cable:

idc cable numbering starts with 1 being the red wire, the one near it 2, etc

at one side:
-insert a 10 pin idc cable into a 10 pin connector as normally would;
then at some reasonable range,
->pins 1&2 can be kept on the stepper gnd, but not connected to mb gnd, just dont cut them but leave their ends open on noise consideration (because having 2 gnds on a board with an ic on it isnt good practice, the stepper board already has the gnd from supply, but you can connect them aswell shouldnt be much difference tho, distance is short enough to worry about noise or turbionary ic current, but i just generally choose to respect the mere few principles i know)
->keep pins 3,4,5: pin 3 is step, 4 is dir, and 5 is enable (double check if you wish)
->cut again the rest 6-10

the other side, 3 lines left, put a 4 pins male header:
- only 3 pins of header will be used, but the 4th (vcc - so cut it near header or isolate it) can be smth like a "key" to remember position, and make a note on pcb not to inser it the other way around;
- scl goes to dir (double check)
- sda goes to step (double check)
- gnd goes to enable (assuming that w/e ic the stepper board uses, needs the enable to low in order to enable the outputs - if it works that way you will see some of the 4 leds light up, otherwise none of them will be lit = all outputs disabled)

the cable is attached in the picture, pointing at it with a filament piece

sry if i overdone it with the explanations
and good luck to it

The explanation is perfect for a true noob like me, thanks!

I see from your picture that you still have an extruder board being used for the thermistor and nozzle and if i'm working this out correctly, those signals go to the MB via the RS485.

With this i've now learnt that SCL/SCA is telling the stepper about the step and direction which is your case is going to a separate board. Hence I guess this is why is could be easy (and cheaper) to driver the extruder stepper via a normal stepper board.

This is where it would be cheaper for me to source a SB if my EB is fried, but still retain the working features of my EB. So, the cable you are pointing at in your photo (many thanks) and the detailed explanation should help determine if there is a fault with my EB.

Just to be 100% certain, is this the connector on the EB that i'm going to pick up on?

Thanks!

Edited 1 time(s). Last edit at 01/20/2011 05:16AM by mattuwattu.

Having a separate stepper board to run the stepper motor, is not because my chip burned (i did replaced it), but because it (at least should) give better results - at least in terms of noise the motor makes, adjusting the values to motor, allows microstep, its simpler in firmware, etc. The stepper board is actually designed for a stepper, while that h-bridge was designed, i believe, primarily for dc motors.

The cable i talked about is from motherboard sda/scl directly to stepper board, does not pass through EC at all, has nothing more to do with it. The arrow on your image points at the "quadrature" connector, which is where connects an eventual encoder, the purpose being to run the extruder with a DC motor which has this encoder attached. As you are using a stepper for extruder, you wont be needing it.

For good practice i wanted to share what happened to close out the problems i described in this thread.

I sat down this weekend to do some testing described by Noobman but before i started making cables i tried the software.

What i have is a set of gen 3 boards with firmware pre installed. As much as i have tried i am unable to update the firmware. I therefore decided to downgrade the RepRap host software. I downloaded the latest version then worked backwards through the old versions. I eventaully got to one that fired up the extruder when i hit the extrude button. Simple!

Thanks for your help on this Noobman.

Hy again - wow its been some time.

I presume you are still running that extruder with a3949 bridges. I dont know why one couldnt update the firmware on either EC22 or MB12. On MB 12 there is a jumper which might prevent programming (via clearing) if its not on, but on EC22 i cant figure why - maybe there is a short or something wrong on 5v level area(s). Or, maybe the programmer just doesnt work.

Have you tried the usb cable for programming? Dunno what programmer you used, but try with arduino and go to preferences file (e.g. windows on \doc&settings[user]\arduino\preferences.txt) and try change upload.using=bootloader and then try uploading the same way but just using the standard cable that you use to talk to motherboard. I hope i remembered the setting ok. Its for the normal cable and the 6-pins in-one-line connector. Praying to work.

Unfortunatelly i cant help much with running via a3949 chips, as i have just deleted that entire part from my own firmware. I saw you just downgraded the host versions to get it working. First things that went into my mind:
- Host versions switching: maybe you did this already, but i will mention it in any case. The host soft makes a preference file that is not located in its directory. The file comes with different versions a reprap.properties.dist. In host, turn Debug to true (besides having comms debug also true), and re-start the host. If you see a message about having weird parameters that are or dont exist on preference file, you might need to locate that file (windows try docs&settings\{user}\.reprap directory) and delete it. Restart the host 2 times and the last time it should say preferences distro and file match and "this is good". This happens alot while changing around different host versions. See the top part of wiki preference file below.
- again i dont have experience using a3949 but as usual im full of (maybe wrong) impressions. If an older host version works then probably its no hardware problem maybe its just something that is different. Maybe in the default settings of older host versions. So if that is tuned it might work with last version aswell. First thing in my mind, if you look at [objects.reprap.org] there is a parameter Extruder0_ExtrusionPWM(0..1)=-1 which sets the extruder motor current, with 0 its disabled and with -1 it should use the ec22 pot. But since we dont know what is on that EC22 firmware, we dont know how, or if it reads pot at all (with no current on board set it to half and avoid changing it as tampering with it on my board resulted on one chip burning). It says a value of 1 gives max current, so maybe its divided from what i vagly remember from gcode variant (not sure). Have you tried changing that parameter at different levels? Maybe setting it to full can burn something, things seem rather sensitive, see [objects.reprap.org] command M113. Or maybe it reads the pot but maybe the pot is currently so low that misses most of the steps - long shot.
- also there are more parameters. One is the extruding time period, for how many ms to extrude and one is a retraction time that is done after each extrusion. Maybe if these are close to each other, it might look like in that video you posted in another tread. Increase lets say the extruding time Extruder Purge to 30000 and Reverse to like 400-500 and see if at the start of a gcode that was made with host (not sf) it goes to dump point and after heating up it extrudes continious and properly at that point. If at this point does ok but on the rest of gcode does bad, maybe the firmware estepspermm is extremelly low which could make the extruding time seem close the same with reverse time. But i think this is very unlikelly, like a long shot, i cant figure as why some1 would set it like that (maybe was used with a cable like that and a driver at full step, or had a big diameter at filament engaging point).

- also, as an overall check, maybe the guy you got the electronics from, used them with replicatorG. In this case the firmware has slight different gcodes interpretations see [replicat.org]. Including totally different HB codes and their M109 doesnt wait while reprap 109 waits for extruder temp. I tried at some point to change my process file to match theirs so i could use it also with HB control. Bottom line, if the firmware was for replicat.org, then try that and see if it works any better. There is a pwm button in control panel there see how to make use of it, and beware when you click something it usually acts on machine asap without confirmation.

Sorry if i give you so many variants, i try to explore each possibilities, but i guess you could evaluate each and see which one is more probable to go with.

Good luck!

Edited 2 time(s). Last edit at 04/26/2011 07:44AM by NoobMan.