Mk4 reported bad temperature vs real life bad temperature  

RE: Mk4 reported bad temperature vs real life bad temperature

dang it, I messed it up. It’s (actual - desired)/(actual), not (desired - actual)/(desired).  I always forget which is which.

However, that doesn’t change the fact that when using degrees Rankine or Kelvin, the answer is the same in either one.

Summary: If doing percent differences on temp, use an absolute temp scale (°R or K) and it’ll be valid. Using °C or °F won’t be valid, because as mentioned the 0 point is arbitrary.

RE: Mk4 reported bad temperature vs real life bad temperature

The absolute zero point is of no relevance for the question at hand. The temperature deviations are scaling not relative to the zero point but to the reference temperature, ie the room or enlclosure temperature which the heatbed is exposed to. 

For such a delta temp °C and K are the same. Percentage errors of delta temp are just a rough guide but they are relevant as a rough measure. 

RE: Mk4 reported bad temperature vs real life bad temperature

In numbers that would mean for the following
target: 110°C
actual: 104°C
Ref.: 30°C

Relative error in %:
((110-30)-(104-30))/(110-30)*100 = 7.5%

In a cold room with 20°C ref. temp it could look like that:
((110-20)-(104-20))/(110-20)*100 = 6.7%

And in a warm enclosure at 50°C like that:
((110-50)-(104-50))/(110-50)*100 = 10.0%

There is a theoretical reason for that reference, after all, if you input no energy into the bed, that will be your default temperature, by heating the bed you are deviating from that temperature and the more you do the bigger your losses will be. Same would be the case for cooling where it would be more obvious why basing such a relative error on absolute temperatures makes no sense. 

RE: Mk4 reported bad temperature vs real life bad temperature

Why would using absolute temperatures make no sense? Why would this become more obvious in a case for cooling?

Heat loss is proportional to temp difference. But we aren't talking about heat loss. We are talking about how much error there is in measured bed temp vs set bed temp.

Posted by: @thejiral

The absolute zero point is of no relevance for the question at hand. The temperature deviations are scaling not relative to the zero point but to the reference temperature, ie the room or enlclosure temperature which the heatbed is exposed to. 

For such a delta temp °C and K are the same. Percentage errors of delta temp are just a rough guide but they are relevant as a rough measure. 

Posted by: @thejiral

In numbers that would mean for the following
target: 110°C
actual: 104°C
Ref.: 30°C

Relative error in %:
((110-30)-(104-30))/(110-30)*100 = 7.5%

In a cold room with 20°C ref. temp it could look like that:
((110-20)-(104-20))/(110-20)*100 = 6.7%

And in a warm enclosure at 50°C like that:
((110-50)-(104-50))/(110-50)*100 = 10.0%

There is a theoretical reason for that reference, after all, if you input no energy into the bed, that will be your default temperature, by heating the bed you are deviating from that temperature and the more you do the bigger your losses will be. Same would be the case for cooling where it would be more obvious why basing such a relative error on absolute temperatures makes no sense. 

RE:

But we are talking about heat loss. What do you think the deviating temperature towards the edge is? Also, you can't measure on the upper side of the flex plate. The higher the temperature of the bed, the greater the heat loss between the location of your thermistor and the actual surface of your build plate. This is complicted by the fact that the heat loss depends a lot on the reference temperature so there is no easy way to compensate for it. There might be some possibilities but they might introduce more error than they do good.

If we are not talking about the real temperature of the bed but about the precision of the thermistor that is another discusion of course.

RE: Mk4 reported bad temperature vs real life bad temperature

Take a look on the picture - top surface reached 104C where prusa and my thermistor was located while down surface by my thermistor reported 102. Silver tape and capton tape might have influence but constant 10C difference at the center of the plate between set temperature and reported temperature by machines is something considering. Keep in mind 2 different prusa machines, from different years, with different beds and thermistors types, this can not be mistake.

I take heat loss on edges as something standard for consumer type machine but still wonder how much expensive would it be to produce more even heated bed 😉

RE: Mk4 reported bad temperature vs real life bad temperature

Posted by: @thejiral

But we are talking about heat loss. What do you think the deviating temperature towards the edge is? Also, you can't measure on the upper side of the flex plate. The higher the temperature of the bed, the greater the heat loss between the location of your thermistor and the actual surface of your build plate. This is complicted by the fact that the heat loss depends a lot on the reference temperature so there is no easy way to compensate for it. There might be some possibilities but they might introduce more error than they do good.

If we are not talking about the real temperature of the bed but about the precision of the thermistor that is another discusion of course.

The main question was deviation of hottest (center) of bed to the bed temp setpoint, not deviation to the edges. We don't know that Prusa isn't applying a compensation for it. In fact, after a bit of searching on github, I found in the Configuration_MK4.h file the following code:

// Bed temperature compensation settings
#define BED_OFFSET 10
#define BED_OFFSET_START 40
#define BED_OFFSET_CENTER 50

Sure looks like to me that there's some sort of temp compensation being done, even if it's just a fixed amount. I need to download the entire project though so I can search through all the files and find out where they are actually using these variables to figure out what they are doing with it. I'll probably do that tonight.

RE: Mk4 reported bad temperature vs real life bad temperature

Interesting. Let's see if you figure out what these numbers are about.
What I am also wondering is if that lower bed temperature is something all the Mk4s show or if it is just some. If it is a very general deviation, it might be also a not very well set temp compensation in the end.

RE: Mk4 reported bad temperature vs real life bad temperature

My experience with MK4 bed temperature. PETG part lifts when printed in the lower right hand corner. I have cleaned the sheet with dish washing detergent, warm water, using 3m green scrub pad, and keeping my paws off the plate. Also tried using NANO liquid adhesive on the plate with same results.

Same part does not lift when printing at the center of the sheet. Understanding that adding brim to this part may lessen the chances of lifting. Consistent bed temperature would lessen the chances of part lifting. I would like to see temperature zones on the bed to mitigate heat losses.

Attached is a excel spread sheet showing the temperature measurements on the surface of the sheet. I took three measurements each of the nine points over a period of 6 minutes. Of interest, the right side of the bed temperature is dropping. While other points on the sheet remained relatively consistent. see attached file

Sheet: PRUSA textured

Set bed temperature: 90C

Temperature Measurement Instrument: Fluke 62 MAX+, Emissivity set value 82

RE: Mk4 reported bad temperature vs real life bad temperature

as all of the temperatures are reading low, have you considered setting the printer temperature high to compensate. 
Thermistors are sensors not thermometers, they interact with Analogue to digital ports on the Printer controller, so they are not massively accurate. 
and they only sense one point on the heatbed. the right front corner is possibly furthest from the thermistor 

regards Joan

RE: Mk4 reported bad temperature vs real life bad temperature

Increasing bed temperature is a possible workaround to compensate for inconsistent bed surface temperatures. Doesn't the XL have zoned heat control across the surface of the bed. Is PRUSA possibly waiting for their competitors to come out with a low cost A1 printer with zoned heat control.

RE: Mk4 reported bad temperature vs real life bad temperature

Summary:
Prusa was trying to help, I've got new thermistor which did not resolve any thing , then new bed which was 1C better then original. Those replacements did not solve anything and for some prints making bed hotter leads to first few layers discoloration so material was overheating.
Scrolling forum I found topic about air turbulence generated by printer and its seems it was main issue.

Solution:
https://www.printables.com/model/621536-improved-mk4-fan-duct
https://www.printables.com/model/566871
Magigoo PC and Magigoo for other ASA (yes this is a must unfortunately).

RE:

That makes no sense. If your bed is cooler than it ought to be, across the board (instead of target temp in the center or wherever and eg 20°C lower temp on the edge.), then compensating that 10°C difference won't lead to overheating but merely establishing the temperature it ought to have. 

I have never seen material discolouring due to setting bed temp 10-20°C hotter than recommended. Whitening bottom sounds like ABS or ASA. If you are talking about either of those two, whitening is a real thing indeed, but it has nothing to do with material getting too hot. It is stress marks which you get if bed adhesion is actually too strong. This can happen quite easily. 

Like you said yourself, the solution here is Magigoo, Dimafix or similar adhesives, which do not only help with adhesion during print, but importantly also with easy release after print. 

RE: Mk4 reported bad temperature vs real life bad temperature

My printer is enclose and I don't print anything from abs/asa/pc/pa11cf until chamber reaches around 29C.
By discoloration i meant different shade, sometimes melt artifact. When i was conducting testing i end up with bumping temperature up to 20C, but still have lifts from shrinking even if used magigoo.  For test i've printed part from prusa pa11-cf on designated sheet and every thing was perfect so that makes me thing that it might be different issue then bed itself. I've started to run different fun ducts and other covers for cooling fans and that was it. With out magigoo i still risk lifting but after applying every thing works as expected 100% success rate in this area with out need to bump bed temperature. Before was not able to print anything with prusa PC and right now, no issue.

RE: Mk4 reported bad temperature vs real life bad temperature

Yes, the Mk4 does seem to have that heat break fan issue, which isn't a problem for PLA but does cause issues for draft sensitive materials. The mod you linked to above, seems to resolve that specific issue. 

29°C is still pretty cold for ABS/ASA and especially PC/PA. But I do understand that the Mk4 can't easily reach enlcosure temperatures which are optimal for these filaments. Even without hitting optimal enclosure temperatures one can get decent prints out of it, as long as one keeps part cooling and drafts to the absolute minimum.

From my own experience. If you have enclosure temperatures of ~50°C, moderate cooling isn't such a big deal for ABS/ASA and if you have 60°C or higher you can also get away with stronger cooling without major impact on layer adhesion. But the Mk4 is not designed for such temperatures. 

RE: Mk4 reported bad temperature vs real life bad temperature

In summer (ambient 32) I was able to hit 42 in enclosure, normally where its around 23 it max out on 35. To get ~50C i would have to install heater in it. That's true was not design but after parts reprint... 😉 When I finally have time will install heater controlled by dev board and automatic power switch to keep ambient around 50C.

RE: Mk4 reported bad temperature vs real life bad temperature

If you want to reach temperatures that high inside of your enclosure, you should also consider moving the xBuddy Board outside of the enclosure to avoid "problems" of the beeping and crashing kind.

RE: Mk4 reported bad temperature vs real life bad temperature

Just be careful with active heaters, especially cheap chinese heaters can be a fire hazard. 
With compact CoreXY machines with enclosures one can reach 50°C even without active heater. Sadly, Prusa doesn't have anything on offer in that category. 

RE: Mk4 reported bad temperature vs real life bad temperature

Not sure what the discussion about percentages are, but in this thread there is clearly discussed, that the displayed temperature and the real temperature on the bed surface is significantly different.

The end user and the filament as well are not interested about the temperature between the heatbed and the sheet, but about the temperature on the sheet surface, which is far from reality.

To increase the bed temperature does not solve the issue, that the MK4 does not display the correct values.

RE:

I think it was established here that the Mk4 firmware probably needs some changes for temperature control as the results do seem too far off. However, your statement that it all doesn't matter as the temperature on the upper side of the build plate matters fails to acknowledge complexities of reality. That location can not be measured directly and the cooling depends not only on the precise build plate used (there are 3 stock options from Prusa alone) and more importantly on the ambient temperature which can vary a lot, as both scenarios are realistic, inside an enclosure and outside. The corners will always be cooler than the centre as well. 

While I agree that Prusa would need to bring the actual temperature of an examplary stock set-up within less than +/- 10°C, expecting much higher precision is missing the point IMHO. If you really want very high precision and optimal conditions, don't use stock settings (or only use them as a starting point) and calibrate your very specific set up with the filaments you are using. Create your own settings and then it really doesn't matter what the machine shows, as long as the measurements are consistent.