top heater uneven heat

Support forum for Achi rework stations.

Re: top heater uneven heat

Postby gkop » Mon Mar 27, 2017 5:04 am

code0102stinks wrote:Why do you think having TC touch the bottom of the board is going to hurt the pcb? Many very experienced BGA guys do that and it controls the temp better. IR plates
put off infrared that heats the pcb more that it will heat a shiny metal TC bead. More accurate. You top heat is far above the BGA chip and it's controlling TC is placed
on top of the board by the BGA. Does it fry the pcb or chip because it is so far from the source of heat? Nope. Here is a pic of Deusjevoo's of his modded bottom TC that
touches the bottom of the PCB. He has good control if the temp.


I think the idea behind not touching the board stems from the boards ability to act as a heatsink when in contact with the probe. It is a fact that some thermal loss will occur via dispersion from board acting as a heatsink. How much is unknown and part of me wants to think that it is probably a trivial amount. to prove or disprove this theory we would need to know the following:

1) where did this logic begin (not necessary, but if it did in fact start before IR was in use and was designed/implemented for bottom hot air stations then we can assume that the argument for its necessity may be false logic based on habit of those who came before IR was common place.)
2) better under standing of IR thermodynamics and its potential for loss via dispersion vs its absorption. this requires the following subsets
a. ir wavelength in common use for bga
b. absorbtion rates of IR verses material and thickness of application.
c. way to calculate and standardize expected loss of heat based on room airflow which can be measure, but not with cheap over the counter solutions.
d. determine if thermal loss from air gap is greater than thermal loss from direct contact.
e. another thermodynamic problem...does the air gap allow for greater heat absorption by creating a pocket of stabilized hot air space vs potentially isolated readings from direct contact that reflect the temperatures of small area instead of the larger whole of the board.
***there is probably a lot more that can be considered. anyone know a good physicist with a focus in thermodynamics? :facepalm:

or, instead, you can say "f" the analytics and just go with what works best for you. get a dummy board and run profiles till your high on flux. thats been my method, cause I dont have the time, attention span or wallet for the real science stuff above. :mrgreen: In all honesty, for the bottom of the board, below or touching is probably negligable in the overall process as building the profile to suit your method will most likely offset the difference.

For the top, I always have mine touching. One issue I have noticed with top temps is that tc probe can be easily tricked by heat ramp up. top of tc probe is blasted with IR and final phase of profile needs to move quickly. The speed of this can cause false reading where tc probe gives higher output of heat from above before it actually warms board to required temp. This results in a premature advance of steps and completion of profile. Never in all of history has being premature in finishing been a good thing. Im still working this one out and have some ideas, but its a PIA to get tc probe where I want it and be able to thermally block overhead heat so that reading of temp is solely based on board temp and not moving air above. Also, in my experience, the closer your IR plate is to your BGA target, the greater the likelihood that temp probe will receive premature readings.

Im done ranting for the night. I come back tomorrow and write the rest of my novel. :D
gkop
Valued Member
Valued Member
Progress to next rank:
50%
 
Posts: 35
Joined: Tue Nov 22, 2016 8:40 pm
Location: Georgia
National Flag:
United States of America
Has thanked: 7 times
Been thanked: 2 times
Set up: ACHI IR6500 Modded 3500w, Hammer-Smash

Re: top heater uneven heat

Sponsor

Sponsor
 

Re: top heater uneven heat

Postby code0102stinks » Mon Mar 27, 2017 1:00 pm

IR plates put off infrared guys. The pcb absorbs IR. The air doesn't. The air gets heated by convection heat from the plates below and also convection
heat coming back off the pcb above after it has gotten hot from absorbing IR from the plates. You seem to realize the top heat TC needing to be touching
the pcb to be accurate, but gloss over the bottom TC. Both top and bottom heaters are IR and both top and bottom are needing to control the temp of the
pcb. Try using hot air on top for your heat and you will see completely different temps at top TC just because the two types of heat are completely
different.
code0102stinks
VIP Member
VIP Member
 
Posts: 687
Joined: Sun Dec 09, 2012 5:18 pm
Location: WA, US
National Flag:
United States of America
Has thanked: 78 times
Been thanked: 232 times
Set up: DIY Hot Air Top Heater and Nozzle. Modded T8280 Preheater

Re: top heater uneven heat

Postby Deus » Mon Mar 27, 2017 2:02 pm

Wanted to react last night, but was sleepy, code beat me...
Totally agree with code0102stinks on this.

If IR is bad at one thing it's at heating up air, it doesn't, only heats up objects absorbing it.
Ofcourse heat will build up between IR elements and PCB which the TC will read.
For me, best option is to read or the elements temp and test what setting gives what result on the pcb.
Or even better, like I do it, TC touching bottom of PCB.
And use HB option, max 5 should do usally.
I know, most say it needs to be higher blablabla but if your heaters can't follow the set Ramp they will give out 100% and glow like hell ...
That is what HB is meant for, to pause heater when it can't follow.

I run both top and bottom together but this way I can hold my top temp lower than bottom untill bottom reaches max temp.
Top heater kicks in a bit later, no need to calculate timings.
Ok, I used to do a lot of testing and made a lot of graphs, saved solme in a document as reference giving me an idea if I want to choose a specific profile.
All this testing, and it was really a lot, you benefit from in the end.

Posted about a bottom tc mod long time ago, best mod ever for it (apart from upgrading heaters, adding different control options etc... ;-) )

Measuring real temps maks it so much easier, profile graph example:
Image
Deus
VIP Member
VIP Member
 
Posts: 479
Joined: Sat Feb 02, 2013 4:18 pm
Has thanked: 13 times
Been thanked: 136 times

Re: top heater uneven heat

Postby demonspride » Mon Mar 27, 2017 3:11 pm

@code, Fair enough,
@Deus what are you running for profile settings?
demonspride
Elite Member
Elite Member
Progress to next rank:
40%
 
Posts: 84
Joined: Sat Apr 11, 2015 12:52 am
Location: America
Has thanked: 1 time
Been thanked: 3 times
Set up: LY/ACHI IR6000 2400Watt upgrade Elstein.

Re: top heater uneven heat

Postby gkop » Mon Mar 27, 2017 4:45 pm

@code0102stinks

I agree with you on absorption of IR, but there are more factors involved than just absorption. You have refraction and heat dissipation (dissipated heat fills the air around object). Is a near area air reading more reliable than direct contact due to radiating heat temps being a more accurate and/or stable temp reading than surface temperature. These may affect temp readings as it would create a cumulative heat effect that plays in relation to how reliable probe readings are in contrast to the absorption and dissipation rates of the board. Remember that boards are design to radiate and dissipate heat

The biggest variables we dont know are entropy of the object we are heating and effectiveness of the IR spectrum that we are using. Also, we would need to know the same info regarding the tc probe so that an offset could be created based on its thermal properties, as it to will absorb IR waves. Until we know these things it is hard to say which is better definitively. This is why I stated that it is what has worked better for me and that possibly either method could be best with a profile build that is catered to your use.

In regards to my admittance of using direct contact on the top, it is an entirely different scenario. Top heat has to be generated and measured much faster which is why direct contact makes more sense to me. We simply dont have the time to allow stabilization for fear that thermal run-away could push past apu tolerances. However, if direct contact is the best method, it is far from accurate as I have mentioned that I can consistently prove that vast temperature readings can be found if you simply move the ir away from the target. This means that the probe is heating faster than the board, which makes sense that it would. :tumbleweed: If only manufacturers were kind enough to install temp probes into board substrate that we could jumper on to. :tumbleweed:

In summary, there are too many variables for us to truly know which is better, which is why you should do what works for you. At least, until someone with time and money to kill wants to do some lab quality temp measurements from all over the board, surrounding air space and even inside the board. Guesstimate is that will need somewhere around 500+ temp probes, multiple config setups and environment scenarios to truly make an educated determination between currently available tools and their effects on temp reading reliability.

***From talking through this, one notion Ive just had is to take a junk board and drill out a pocket and channel underneath the apu. Install 4-5 temp probes for under apu readings, then reball apu back to board. wait for cool-down and then run profiles for temp reading. This may provide some interesting data as it wouldnt give the whole picture, but certainly would give more accurate info on thermal transfer to the actual area we want heat...where the solder actually resides. this would have to be counter measured with tc probes at marked locations on the board and then found to be reproduceable within a margin of tolerance. This last part is due to needing a reference point for profile build of temps at key steps under APU in comparison to actual locations which can be probed on actual repairs.
gkop
Valued Member
Valued Member
Progress to next rank:
50%
 
Posts: 35
Joined: Tue Nov 22, 2016 8:40 pm
Location: Georgia
National Flag:
United States of America
Has thanked: 7 times
Been thanked: 2 times
Set up: ACHI IR6500 Modded 3500w, Hammer-Smash

Re: top heater uneven heat

Postby Deus » Mon Mar 27, 2017 5:14 pm

My profile might not help you as I don't have an IR 6000.
Mine is a moddedI Achi IR Pro, total of 3600 watt: 3200w bottom, 400w Estein RFS 80 for top.
On the graph in that pic profile was:

0.65 - 120 - 0 // 0.5 - 140 - 75 // 1 - 200 - 0 // 0.5 - 225 - 0 // 1 - 220 - 90 // -0.01
HB value of 5
Top heater 5cm above
PID settings: 19 - 2 - 45
It also had an offset value of 15 as my TC was not reading correctly.
Up to about 150c real temp is in fact 15c lower, above that it gets closer to +/- correct readings in the last steps.

The red line is bottom PCB temp, TC is touching pcb.

Here's my bottom TC mod:
http://bgamods.com/viewtopic.php?f=23&t=5113&start=10#p45289
Deus
VIP Member
VIP Member
 
Posts: 479
Joined: Sat Feb 02, 2013 4:18 pm
Has thanked: 13 times
Been thanked: 136 times

Re: top heater uneven heat

Postby Deus » Mon Mar 27, 2017 5:41 pm

gkop wrote:I agree with you on absorption of IR, but there are more factors involved than just absorption. You have refraction and heat dissipation (dissipated heat fills the air around object). Is a near area air reading more reliable than direct contact due to radiating heat temps being a more accurate and/or stable temp reading than surface temperature. These may affect temp readings as it would create a cumulative heat effect that plays in relation to how reliable probe readings are in contrast to the absorption and dissipation rates of the board. Remember that boards are design to radiate and dissipate heat


For me only 2 options: or TC in (there are types with internal TCs) or on IR element, or TC against the object to heat up (pcb), be it top or bottom.
In theory slower preheating should give more consistent results.
Some more theory:
As you already said. the result has to do with the target surface, distance, absorption, reflection, amount of copper in pcb etc.
When TC is in or on IR element and knowing the curve of your heaters you can set it to the temperature needed for the IR wavelength best suited for the surface you target.
Theorethically this is how one should use IR. (Setting radiation type and distance for the target body).
When the TC is hanging in between elements and pcb, think reading will variate much more, not a good thing for repeateability.

In practice, at least for me, using TC touching target, e.g.bottom lets me control the process pretty well too.
For bottom heater, by changing some parameters I can go slower or faster, or start fast, go slower on higher temps by simply changing them manually in process as needed or pauze/run manually.
Immediately see the result on the targeted surface/body, TCs touching really helps in getting pretty close results every time.
And seeing effective temps prevents me from burning thing on the bottom too.
For top I can set these thing in the profile ofcourse.
Deus
VIP Member
VIP Member
 
Posts: 479
Joined: Sat Feb 02, 2013 4:18 pm
Has thanked: 13 times
Been thanked: 136 times

Previous

Return to Achi

Who is online

Users browsing this forum: No registered users and 1 guest