[MK3S] Something wrong with firmware logic on temp control?
 

[MK3S] Something wrong with firmware logic on temp control?  

Page 1 / 2
  RSS
titus.o
(@titus-o)
Eminent Member

After upgrading from MK3 to MK3S I getting a-lot-more thermal runaway problem.
Today is another day that thermal runaway strike... AGAIN...
But today I see something really strange.

Some background info first.

E3D silicon sock installed
Mean Well NES-350-24 24V 350 Watt
PETG 240/90deg
FW3.5.3 & 3.7.1 (same problem)
PID tuned (both hotend and bed)
Filament sensor OFF
Power panic OFF

Thermal runaway happened at 2nd layer, where the parts-cooling-fan is yet activated, which rule out the possibility that parts-cooling-fan blow the temp down.

On both print (FW3.5.3 & 3.7.1 each), I can see both red LED on Einsy board (hotend and bed heating indicator) goes off for ~15sec even it's already under-temperature. It's not "not-flashing", it's "off" for 15sec!

The OctoPrint temp monitor also show same info, both hoteend and bed temp drop at the same time.

Is there something wrong with firmware logic on temp control?

 

I really don't want to convert back to MK3 and run really old firmware.

Best Answer by titus.o:

OK I think I got it solved...

It is... Power Panic...

YES... somehow I've power panic cable connected to Einsy... which I never did from first day I get MK3.
Must be a mistake when adding Einsy cooling fan.

After unplug power panic cable my MK3S back to work.
So... it's still a firmware problem, sort of...

Nope I don't use power panic at all since I always using OctoPrint.

This topic was modified 2 years ago by titus.o
...
Posted : 23/06/2019 2:57 am
titus.o
(@titus-o)
Eminent Member

Bigger image on OctoPrint

...
Posted : 23/06/2019 3:06 am
Nikolai
(@nikolai)
Noble Member

Looks like you have either loose or broken thermistor cable. I would replace it.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 23/06/2019 4:20 am
Dave Avery
(@dave-avery)
Honorable Member

actually i suspect a power supply problem because both the hotend and bed temps drop at the same time. 

Posted : 23/06/2019 4:41 am
titus.o
(@titus-o)
Eminent Member
Posted by: david.a66

actually i suspect a power supply problem because both the hotend and bed temps drop at the same time. 

This doesn't explain why hotend & bed heating LED stop flashing at the same time.

...
Posted : 23/06/2019 4:49 am
--
 --
(@-2)
Illustrious Member
Posted by: titus.o
Posted by: david.a66

actually i suspect a power supply problem because both the hotend and bed temps drop at the same time. 

This doesn't explain why hotend & bed heating LED stop flashing at the same time.

LMAO ... now that is funny. 

It is always wise to get more than one opinion... as for trusting Prusa? No way man....
Posted : 23/06/2019 4:58 am
titus.o
(@titus-o)
Eminent Member
Posted by: nikolai.r

Looks like you have either loose or broken thermistor cable. I would replace it.

If the condition is temp-drop while LED is flashing I would agree with you. But the point is: both hotend and bed LED is not flashing! Which means temp control logic is not sending power over to hotend and bed...

The only thing I can think about to cause such is firmware bug... 🙁

...
Posted : 23/06/2019 4:59 am
Nikolai
(@nikolai)
Noble Member

Power supply problem would most likely lead to printer reset.

Try to check if it has something to do with the position of the bed or x axes. In general I would always check from the tip to the base. Means check the cables first then the connections then board. 

I hope you flashed the right firmware and there is no slicer/sd card error. 

It's very interesting that both temps are dropping at the same time. 

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 23/06/2019 5:02 am
Nikolai
(@nikolai)
Noble Member
Posted by: titus.o
Posted by: nikolai.r

Looks like you have either loose or broken thermistor cable. I would replace it.

If the condition is temp-drop while LED is flashing I would agree with you. But the point is: both hotend and bed LED is not flashing! Which means temp control logic is not sending power over to hotend and bed...

The only thing I can think about to cause such is firmware bug... 🙁

Which LED are you referring to? On the einsy board? What about the bed LED? 

Firmware bug is the last thing I would think of.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 23/06/2019 5:05 am
--
 --
(@-2)
Illustrious Member

If power goes away - the LED's will not flash. Where do you think they get their power?

 

Nikolai, a simple 24v problem on the power bus won't affect the CPU too much. I doubt an open circuit fail will cause a reset that a short would cause.

This post was modified 2 years ago by --
It is always wise to get more than one opinion... as for trusting Prusa? No way man....
Posted : 23/06/2019 5:06 am
Dave Avery
(@dave-avery)
Honorable Member

if the power supply dropped from 24 to 12 v for a time both heaters  would drop in power but the logic would still work (5v is derived from the 24 V so would still function as long as the power supply put out over 5v)the 2 heaters are fed from separate feeds from the power supply and are low side controlled ( FET to ground with the high side of the heater connected to power thru the board fuses ) 

Posted : 23/06/2019 5:07 am
--
 --
(@-2)
Illustrious Member

LEDs get the same power as the heaters ... power loss is power loss. 

But when it is all said and done, the only way to know if the power supply is drooping is to measure it.

It is always wise to get more than one opinion... as for trusting Prusa? No way man....
Posted : 23/06/2019 5:16 am
--
 --
(@-2)
Illustrious Member

And these's always a chance something is messing with the AC Fault detection.  That's happened before...

It is always wise to get more than one opinion... as for trusting Prusa? No way man....
Posted : 23/06/2019 5:19 am
Nikolai
(@nikolai)
Noble Member

@Tim & David: True. Easy to test this case with a voltmeter. You might also see a flicker or dimmed led in this case (because of the schematic).

My bet is still on something simple. It worked before and now after the upgrade issue appear. So check whatever was touched/changed. 

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 23/06/2019 5:20 am
--
 --
(@-2)
Illustrious Member

Mean Well NES-350-24 24V 350 Watt

I presume there is a new over powered power supply involved.  Not much simpler than that.  Perhaps when wiring the wrong connections were used and the power supply compromised in some way? Cross connected the various rails, etc?

 

This post was modified 2 years ago by --
It is always wise to get more than one opinion... as for trusting Prusa? No way man....
Posted : 23/06/2019 5:25 am
titus.o
(@titus-o)
Eminent Member
Posted by: nikolai.r
Posted by: titus.o
Posted by: nikolai.r

Looks like you have either loose or broken thermistor cable. I would replace it.

If the condition is temp-drop while LED is flashing I would agree with you. But the point is: both hotend and bed LED is not flashing! Which means temp control logic is not sending power over to hotend and bed...

The only thing I can think about to cause such is firmware bug... 🙁

Which LED are you referring to? On the einsy board? What about the bed LED? 

Firmware bug is the last thing I would think of.

These 2 marked with purple box. I know the one on left is hotend, the one on right is bed.

...
Posted : 23/06/2019 5:36 am
titus.o
(@titus-o)
Eminent Member
Posted by: Tim

Mean Well NES-350-24 24V 350 Watt

I presume there is a new over powered power supply involved.  Not much simpler than that.  Perhaps when wiring the wrong connections were used and the power supply compromised in some way? Cross connected the various rails, etc?

 

I been rolling with MW NES-350-24 from the 2nd day I recieve my MK3. The first thing I did on MK3 is print the parts for NES-350-24 and replace the stock PSU right away.

That configuration been rolling for 6mons or so with 5kg printed. Nothing bad. Hard to believe it's MW NES-350-24 problem.

I'm doing another print (same g-code) now and LCD monitor displaying voltage. Let's see... 

-----

EDIT:

And it just thermal runaway... 23.8V no change on PSU...

This post was modified 2 years ago by titus.o
...
Posted : 23/06/2019 5:40 am
--
 --
(@-2)
Illustrious Member

Depending on the sampling timing, you won't see these sorts of defects on the LCD; it takes a meter that reads several times a second; preferably hundreds of times a second and captures the min and max readings of those samples.

And thermal runaway has nothing to do with simultaneous temp drops of both heaters.  Totally different error.  So you are fighting at least two things not working right.

Thermal shutdown is very common with the MK3S nozzle fan shroud design.  Heater just gets out of sync with the controller when the part an clicks on. Socks often help at PLA temps.  

Thermal errors, on a MK3, typically a failing thermistor. Common and simple.

ps: why do you have Power Panic disabled? Getting a lot of false triggers?

This post was modified 2 years ago by --
It is always wise to get more than one opinion... as for trusting Prusa? No way man....
Posted : 23/06/2019 10:22 am
Dave Avery
(@dave-avery)
Honorable Member

unless you take the power panic assy out of the old prusa supply or buy a clone power panic board the meanwell doesn't come with power panic sensing. i expect lots of users with replacement supplies don't have power panic 

Posted : 23/06/2019 1:44 pm
titus.o
(@titus-o)
Eminent Member

OK I think I got it solved...

It is... Power Panic...

YES... somehow I've power panic cable connected to Einsy... which I never did from first day I get MK3.
Must be a mistake when adding Einsy cooling fan.

After unplug power panic cable my MK3S back to work.
So... it's still a firmware problem, sort of...

Nope I don't use power panic at all since I always using OctoPrint.

...
Posted : 23/06/2019 3:43 pm
Page 1 / 2
Share:

Please Login or Register