BBB + CRAMPS extruder control

[Pete McKenna]

My first attempts at posting haven't worked. I'll post the files if this post works.

I have an Ordbot I'm adding BBB+ CRAMPs control too. I have the motion control working reasonably well, but I can't get control of the FETs for the extruder and fan control, if I hack the hpg.pwmgen.00.out.xx.value I can enable them, but not with any pid / feedback control. ( I can confirm the FETs work and the extruder heats and the fan runs via hacking the vaule )

I have checked the thermistor, and it works as expected. What I can't figure out is how to get the commanded temp to go to anything other than 0. I have pre-heat settings, I've tried to use the remaped M codes for 104 and 109 (and 106 for the fan) but I get nothing. My configs are attached, if anyone has a BBB and CRAMPS working for 3D printing, or FET controls I'd love to get your advice.

Thanks

Pete

[Pete McKenna]

OK seems that firefox works, lets try the files.

[Pete McKenna]

Is there a trick to attach .hal and .ini files?

I'll try changing them to .txt I guess.

[Charles Steinkuehler]

There's not really enough detail to determine exactly what's wrong.

Since you indicate you can set the hpg.pwmgen.00.out.xx.value signals
and get expected results, it sounds like the PRU is properly generating
PWM signals. Note that by default the value signal is normalized, so
0.0 is "always off", 1.0 is "always on" and 0.5 is "on half of the time".

I do like to explicitly set the PRU period, but I doubt that's causing
your problems:
https://github.com/cdsteinkuehler/machine-configs/blob/master/configs/CoreXZ/CoreXZ.ini#L3

So if the PWM is working correctly the problem is likely with the PID or
temperature code. Use halmeter, halsscope, or just "halcmd show pin" to
see what the signals are doing. If you can't tell what's wrong, post
some additional debugging from these utilities and maybe we can spot
what's going wrong.

Finally, one thing that trips up a lot of CRAMPS users is you have a
bunch of different power supply connections. There are three different
supplies for the CRAMPS FET outputs:

P402: EXTPWR (E0, E1, E2)
P403: BEDPWR (BEDHTR output)
P404: AUXPWR V+ (FET5, FET6)

Make sure you have all the necessary supply rails connected. The LED
next to the FET output connector should light when the FET is on and
power is properly applied. For testing, you can use halcmd to set any
signals that are not being actively driven (eg: halcmd setp
hpg.pwmgen.00.out.02.value 1.0) without having to edit your HAL file and
relaunch.

--
Charles Steinkuehler
cha...@steinkuehler.net

[Pete McKenna]

Thanks Charles, and thanks for creating the CRAMPS!

I noticed as I was working with it last night that I have different aspects working depending upon, how I'm getting access to the BBB. I just got a monitor for it so I set it up as a stand alone device and had the monitor with keyboard, mouse on a USB hub, and power via V5 adaptor to the usb port no computer connected. 

In that state.

The E0 heater will come up as expected based on the bias setting, but there is no therm input, so it will just run away.

I switched back to the laptop powering the BBB over the USB connection using that as the network connection as well to run Axis over X. This resulted in no output from the E0 FET, but I have the therm input working again.

Power rails wise I have no bed and have 24VDC running to  P201, and P402 and the screw terminals, V+ and G.

The CRAMPS is v2.2 

I appreciate the help. The power connections were based on a picture of the board be wired up, that I found online. Its a very strange set of conditions, and it is repeatable when I switch from stand-alone to with a PC connected.

I can set the halcmd setp for the FAN FET, but when trying to do so for the E0, it tells me there is a pin assigned or something like that, but I'll try it again. and I see the LEDs for all of the FETs now so I'll pay attention to if I have the expected control of them when halcmd setp.

Thanks

Pete

[Charles Steinkuehler]

I'm not saying this is your problem, but the BBB (and most ARM SBCs) are
pretty sensitive to the power supply and cannot be reliably driven by
typical USB ports. I typically run my 12V CRAMPS systems using an ATX
power supply with the 5V standby power going to the +5V connection on
the AUXPWR connector (P404) with the BBB/CRAMPS board turning on the ATX
supply via P301 (which is basically the same as powering the BBB via the
P1 5VDC "barrel" power jack).

For my 24V systems, I have a separate 24V supply for the high power
rails and use a 5V "wall wart" to the BBB DC input (P1).

I also run the BBB "stand alone", with a physical Ethernet cable
connected to the BBB and using the HDMI output for video and a standard
USB KB/mouse for the console (sometimes with a serial link as well if
I'm debugging). Connecting via the USB gadget interfaces _should_ work,
but it's not how I test and I don't think it's a common configuration
for most developers.

It's strange the therm input doesn't work stand-alone but does when
you're powered from the laptop. I'm not sure what could cause that, but
some basic debugging ought to be able to root out the cause. Run
halmeter or halscope and monitor the temperature outputs from the
hal_temp_bbb component. You should be able to see the values change
with the temperature (eg: hold the thermister between your fingers vs.
letting it sit at room temperature). If that's not working, you likely
have a device-tree issue or something which is causing the ADC to not be
working properly. If you see the values change, there's probably
something wrong in your HAL file.

--
Charles Steinkuehler
cha...@steinkuehler.net

[Pete McKenna]

I stopped using the usb as the board power source, chased out some USB device failure issues, (hub and mouse, it seems) and then replaced my thermistor, as it got unstable when I tried to neaten up the cords. With those changes in I was able to run a couple of calibration cubes, before I jammed up my nozzle. So I think I've got it going and just need to work on my tuning. Thanks for the help.

Pete