[Solved] Software disconnects during baking or reflow

[ijprest]

Hi folks,

I just got my Voltera tonight. It was well packed and delivered quickly to Seattle, WA, and I had no problems with customs/duty owing like you mentioned might happen in your last email update.

So I've been following the 'Hello World' directions, and things have mostly been OK, but I've had a couple of issues with the software (Windows, v0.5.0, plus associated firmware update).

1. During my first print, the software (I think) stopped working; the top-level menu was still visible, but the frame was entirely filled with a white background. The print also stopped at this point.
   • I was playing with the software during the print (zooming & panning); not sure if this contributed to the crash.
   • Since it crashed in the middle of a print, restarting the software broke the nozzle tip when it homed the carriage; thankfully we have spares.

2. I'm now in the middle of my second attempt; the print worked fine (I made sure not to touch the software while printing). But now it's in the 'bake' stage and the timer isn't counting down (stuck at 43:29), and the graph isn't updating.

   • The heater is definitely on...
   • ... though it doesn't seem to be giving off as much heat as it did a few minutes ago. (I estimate I'm ~35 minutes or so into the bake... maybe it's in a cool-off phase?)
   • The software still seems responsive... just isn't updating (or maybe not receiving data from the hardware?).

Has anyone else experienced similar problems? If there's any particular data you want me to gather to help debug, let me know.

For now, I'm going to to let it go for 44 minutes (as closely as I can estimate it) before I click stop... I will report back when it finishes.

Thanks,

[jazozaya]

Hi ijprest,

I'm sorry you're experiencing some hiccups during your first print. I'm written some comments below:

First Print:

• What do you mean by top-level menu? Do you mean the standard windows menus like 'File' or 'View' or the software specific header like the stage progress (the green line at the top). There is a small glitch in the print preview currently but it only happens when you click and drag the circuit around. We've never seen it crash the system.

• After the software crashed, you mention the tip broke whilst homing. In what direction did the carriage move and at what point did the software start homing? The printer should raise first and then move in the XY whilst homing to prevent nozzles from breaking.

If the carriage ever stops during a print (either a software crash, power loss or E-Stop) the safest way to recover is to click the home icon in the top right of the software, this will force the system to raise and park the carriage.

During Baking:

• What temperature did the printer claim it was at? Did say it was at room temperature or 0°C?

• If you press the tilde key (~), a console will open up, this console will log all serial communication between the printer and the software. If the printer is baking as expected, a bedTemperatureUpdate should be recorded in the console every second. If there is no activity happening between the printer and the software then it might be the unit disconnected.

• What color are the LEDs glowing? If the LEDs are glowing red that means the temperature of the bed is above 60°C (regardless of the state of the software). If the LEDs are orange then the temperature is between 40°C and 60°C. If the LEDs are blue then the temperature is dropping.

Based on the symptoms you see, I suspect the unit disconnected after you clicked bake and was not able to re-establish a connection. If this happens, the V-One will continue executing the temperature profile and will return to room temperature once it is done.

---

To help us troubleshoot the issue, can you please create a support package? If you click on the envelope in the top right of the software a form will pop up which will ask for your name & email. The package will contain the relevant logs and configuration files that will help us narrow down the problem.

Please keep us posted,
Jesus Zozaya

[ijprest]

What do you mean by top-level menu? Do you mean the standard windows menus like 'File' or 'View' ...

... There is a small glitch in the print preview currently but it only happens when you click and drag the circuit around. We've never seen it crash the system.

Yes... the standard Windows File/View menu was still visible and responsive, but there was nothing else inside the frame (just a white background), and the print stopped.

This was during the print stage. I was definitely zooming & panning when it happened. It's possible I accidentally dragged something.

What temperature did the printer claim it was at? Did say it was at room temperature or 0°C?

It claimed to be at room-temp, with a slight uptick by a few degrees before it stopped updating.

What color are the LEDs glowing? ... Based on the symptoms you see, I suspect the unit disconnected ...

They were red for most of the bake; towards the end they went blue. I suspect the printer executed the temperature profile correctly.

At the end of the bake, I clicked the 'stop' button in the UI, and the software went into the 'crashed' state similar to my first attempt (file/view menu, but everything else in the frame was white). When I restarted the software, and it seemed to take a long time 'searching for printers'. I unplugged/replugged the USB and it was found right away.

So it definitely looks like some sort of communications problem. I'll try tonight with a different USB port and report back; maybe the USB port I used is flaky (it's a front-panel port on my desktop, so it's definitely subject to more wear & tear than the rear ports).

Thanks,

[jazozaya]

Hi ijprest,

Two things I forgot to ask:

• Do you have the printer plugged into a busy power bar?

The heater is 550W and will consume a couple of amps. We've noticed that if we have multiple monitors and computers plugged into the same power bar, when we start baking the printer will disconnect and reset itself.

If this is the case with your setup, I would encourage you to connect the V-One to a different outlet or disconnect unused devices.

• Did you plug in other USB devices when operating the V-One?

Whenever you plug in additional USB devices (mice, keyboards, inkjet printers, etc). Windows has a tendency of power cycling all connected USB devices. So it is possible for you to inadvertently disconnect your V-One by plugging in a seemingly unrelated device.

---

Hopefully this helps,
Jesus Zozaya

[ijprest]

Do you have the printer plugged into a busy power bar?

There's two computers & two monitors on it (plus misc low-power stuff, like a modem, fan & some lighting), but I wouldn't say it was particularly loaded. The UPS didn't beep at me like it does when I use my laser printer, so I don't think the Voltera drew enough power to cause a power-dip or anything. (FWIW, the computers are on the UPS, but the Voltera isn't, since you're not supposed to put high-draw items on the UPS.)

Did you plug in other USB devices when operating the V-One?

Not intentionally... but USB in Windows is always a bit wacky, so who knows?!.

...

I'm happy to report that tonight's attempt was much more successful. I printed, baked, soldered & reflowed with no problems. I'm now the proud owner of a blinking LED thing.

The only difference was that I plugged it into a powered USB3 hub, rather than into the front USB port on my PC. So I'm going to chalk it up to a flaky USB port.

Two suggestions for your next software update:

1. More robust handling of communications interruptions... maybe display a message or something.
2. In a bunch of places in the UI you need to click a button (like 'calibrate'), but the button doesn't appear if the steppers have been turned off (it just says 'please wait' or something). I found this confusing.
   • Maybe you could detect this and activate the steppers automatically at the beginning of these stages.
   • At a minimum, maybe the UI/instructions could prompt you to click 'home', which will activate the steppers and enable the button.

...

Software teething pains aside, I'm very pleased with the Voltera. All I need now is pick & place! Next version, maybe?

[jazozaya]

Hi ijprest,

I am happy to hear your second attempt went much smoother.

If USB communications are interrupted, the software should let you know in two separate locations as indicated below. You won't be able to click any buttons (print, calibrate, bake) if the printer is disconnected. Looks like you've encountered a bug, I'll make a note of it.

In a bunch of places in the UI you need to click a button (like 'calibrate'), but the button doesn't appear if the steppers have been turned off (it just says 'please wait' or something). I found this confusing.
Maybe you could detect this and activate the steppers automatically at the beginning of these stages.
At a minimum, maybe the UI/instructions could prompt you to click 'home', which will activate the steppers and enable the button.

As you advance through the software, the printer will automatically prepare the corresponding tool as soon as you enter relevant step. For example, as soon as you enter the Probe step the printer will start moving

The tool preparation sequence will:
1. Raise the carriage first to ensure nozzles don't get damaged.
2. Home the carriage in the X & Y axis
3. Ensure the right tool is mounted (i.e - A dispenser is not mounted when it expected a probe)
4. Lower the tool to the bottom circular shaft (bottom Z switch). This sets our 0 height.
5. Perform XY alignment by descending into the larger hole and contacting the 4 limit switches
6. Raise the carriage and return to (0,0).

Whenever the motors are released, the tool preparation sequence needs to be executed again. In a typical print, it will only happen twice, once for the probe and once for the dispenser.

During tool preparation, the buttons will display 'Please Wait', once the sequence is complete the buttons will behave as normal.

---

Please keep your feedback coming. We love hearing about your experience.
Jesus Zozaya

[jazozaya]

Hi ijprest,

I just wanted to follow up - I looked into the small glitch in the print preview I mentioned. Turns out it wasn't as small of a glitch as I thought!

In the print or calibrate step, you are normally able to click and drag a box to select what traces you want to print (Shift adds to selection, Ctrl removes from the selection). I discovered that under certain circumstances it was possible to crash the software (software freezes, then white screen) just by clicking around in the preview. I'm not 100% certain but I think Alt-tabbing could also induce it.

I'm fairly confident that you encountered this glitch during your first print. We have fixed the bug and will be pushing a software update this week.

---

Thanks for telling us about it!
Jesus Zozaya

[Eriond]

Hi all!
I just wanted to add a +1 to the "software hangs during bake process". In my case; pretty much the same scenario. I was doing my first Hello World print, and all was good and well until somewhere in the middle of baking. With 22:53 left on the timer, there were no more updates. I opened the console, but couldn't see any updates there. re-plugging the USB, windows told me an unknown device was attached. After power-cycling the Voltera, all was back to normal.

I tried to do a support package, but then I got an "Uh oh" issue. Hence this post.
Secondly, I would appreciate if it was possible to use a different key than tilde to get to the console. On my Swedish keyboard layout, I have to press AltGr + "key-next-to-return-with-a-lot-of-symbols-incl-tilde" and then followed by space. It took me a while to figure that one out...

[jazozaya]

Hi Eriond,

Thanks for letting us know, if the printer resets right as you start the bake process then it is probably the power draw problem I talked about earlier. However if you were already 20 mins into the bake then it is likely something else.

We've encountered the issue you are describing ourselves only a handful of times. It's a skittish bug that tends to go away the next time you are baking, if you find a way to reliably reproduce the disconnect please let us know. Knowing how to reproduce a problem is always the first step to fixing it.

If this happens again and you still have ink that needs curing, as of software version 0.5.0, it is possible to jump directly into the curing step by clicking Reflow and then Bake. You don't need to go through the Print sequence again.(see below)

I believe you were not able to submit a support package because of a permission issue with the email service we are using. I'll look into it today.

Good to know about the console shortcut, I'll look into changing it to Alt+C or Alt-G to make it more accessible to our Scandinavian folks

---

Please keep the feedback coming!
Jesus Zozaya

[Eriond]

You called it a disconnect issue, but there is a thing that puzzels me.
I suppose you use an FTDI USB-to-serial solution, and when disconnecting and reconnecting the cable, it should just have removed and then recreated the associated virtual serial port. It shouldn't have given me the warning about an unknown device, unless the FTDI chip itself hung for some reason. What I'm saying is that maybe it isn't a Voltera issue, but an FTDI one.

Another thing that briefly flashed by was when I was about to reflow the board, the GUI button was labeled "BAKE". I guess it should have been "REFLOW". Picky me...

[lintilla]

+1 to the "software hangs during bake process" here too. Happens every time.

The software do not say disconnected, but the timer stops.
I just close the program and let the machine run it's cycle.
When led turns blue I can open the software again and it connects just fine.

[lumeffects]

From my experience, the ftdi makes a virtual com port but the application program has to connect to that. If the ftdi is forced to close and reopen the com port, the app is often unaware and will automatically loose the link, even if the com port is reestablished as the same com port number. This is a fundamental usb com port connectivity problem which my company forever had problems with.
Reestablishin a broken serial usb link is achievable but can be tricky and problematic. It's most probably the applications responsibility to detect and reestablish the link.

[Eriond]

I agree that keeping a COM port link can be tricky, and that you have take special care for the disconnect event. With FTDI, there is a second option: to use the D2XX DLL. This gives the programmer more options as it connects directly via the USB stack. Not sure which method the Voltera app uses, but if sudden disconnects are an issue, maybe the D2XX could be a way forward? (hint, hint )

[HBijlsma]

The same / really similar problem here.. So I can confirm the problem
On a windows 10 computer while baking the time remaining stops and the v-one is non-responsive.

At first I thought it maybe was the FTDI driver, because I used an already existing driver installation. After changing the driver to the one on the voltera website the board still performs the same.

Random time stops after 1-5 minutes of baking.
It happens every time, even after system reboots etc.
(Software version v0.6.5)

[HBijlsma]

Giving some experimentation it looks like the printer becomes unresponsive on the moment the heating elements re-engages.

Tried at a 'manual' setting of 160degrees celcius, the printer was cooling from the 200+ degrees of the bake..
After dropping below the setting (160) the printer again becomes unresponsive.

[jazozaya]

Hi @HBijlsma,

On a windows 10 computer while baking the time remaining stops and the v-one is non-responsive.

Can you clarify what you mean by the V-One is non responsive? Do you mean that:

1. The software freezes completely and eventually white screens and crashes?
2. The software remains responsive, but sending more commands like homing the system, or outlining a print do nothing. (The printer does not move)
3. The software remains responsive, but the buttons are not visible and instead display "Please Wait" or similar.

It's worth mentioning that the entire heating profile is loaded onto the printer as soon as you hit bake. The printer will continue baking even if the software crashes or you physically unplug the unit. If the printer stops 1-5 minutes into the bake, it is possible that the serial connection broke, but the printer is still running fine. When a bake cycle finishes and the printer is cooling down, the LED's will flash blue, so you can use these as an indication.

With that said - the printer will reset if you restart the software, this will wipe out the baking profile and you will lose the progress.

We are still looking into the source of this issue, we made some changes to how serial communication is handled last week and are hoping to release the next SW version this week.

---

Jesus Zozaya

[HBijlsma]

The time on the screen freezes...
The program stays responsive, the printer keeps baking (the leds still pulsate and the temperature is good)
Stopping the bake makes the software application crash. (Or any other action which communicates with the printer)

When restarting the program software, it cannot connect to the printer anymore. (keeps searching)
Only a hard-reset (on/off/on) from the printer makes communication possible...

[HBijlsma]

Additional info:
My problem is only for the 'bake' and manual settings..
The reflow -> automatic worked perfectly

[troy]

This may have been fixed in 0.0.7, i.e. "Resolved an invalid function call that would have left the app in a semi-responsive state." Some of the log files I have point to that toward that issue/fix.

@HBijlsma if you submit a support package for this i can read the log file to confirm that it contains the pattern I am expecting (if not then it helps anyway). I know you are already submitting a package for another issue, but please submit again for this one. I think that will make it easier to track the resolution.

For those who already submitted a package, i have likely already scanned your log files.

For everyone, if you still see this in 0.7 please report again, I'll need a new log file too.

Cheers

[HBijlsma]

Just sent the packages (0.6.5 version and 0.7.0 version)

Hope you can make sense of it

The 0.7.0 version seems to have the same problem, but is more robust. (the program doesn't crash afterwards)