-
Quote
And you dont have to learn some awkward tool. Works perfectly fine for me !
I guess once you've learned the quirks of Altium, you've got no brain cells left. -
Difference between HT230 and LS230
According to Matweb.com there are some very minor differences between LS230 and HT230. Heres the comparison
The LS230 exhibits slighlty ( 2% ) higher Heat of Vaporisation, and slighty higher Specific Heat Capacity. ( 1.5% )
For all practical purposes Galden HT230 and Galden LS230 appear to be very very similar and for this application i don't think its going to be an issue.
I've found three commercial systems that specify Galden HT230 as their working fluid, as well as several DIY projects.
That being the case, I propose that getting some HT230 seems to be ok.
-
And then there was this. This is the 'official' word From the people who make it.
"Solvay Solexis understands that Galden® HT Grades are used as Soldering Fluids in Vapor Phase Soldering Equipment. These fluids have a wider molecular weight distribution and are designed for use as Heat Transfer and working fluids in closed systems. They consist of a mixture of elements belonging to the same chemical family of perfluoropolyethers having varying boiling points, including those with boiling points lower than the nominal values. These low boiling components preferentially evaporate during the operation of the Vapor Soldering Machines. This phenomenon leads to higher losses of the fluids as well as a shift towards higher vapor temperature with time – temperature shift.
As a result, the cost of the vapor phase soldering process and the fluid emissions to the environment increase. The latter defeats the purpose of moving towards lower environmental friendly lead-free soldering.
We believe that Vapor Phase Soldering produces the best quality soldered joints. In addition, it is the most economical, environmental-friendly and the safest soldering process compared to other technologies such as convection ovens that operate at higher temperatures due to the poor heat transfer efficiency of air and other gases.
The Galden® LS and HS fluids are specifically designed with very narrow molecular weight distribution for the Vapor phase Soldering process by a stricter control of the molecular weight distribution during their production. This eliminates the above mentioned problems.
In addition, in order to improve the performance of the Galden® HT grades in closed loop systems, Solvay Solexis has decided to modify the specifications of Galden® HT grades with emphasis on the properties specific to these uses, such as viscosity.
This would make the Galden® HT grades even less suitable for use in Vapor Phase Soldering Process. Solvay Solexis will not be guarantee to provide technical assistance
?to problems arising out of the wrong use of our Galden® HT grades in Vapor Phase Soldering Equipment.
At the same time Solvay Solexis will be at our customer disposal to make the transition form Galden® HT grades to Galden® LS and Galden® HS grades as smooth as possible.
Solvay Solexis also guarantees the supply of the Galden® LS and HS grades for current and future requirements..
We request you to kindly inform your customers of these changes and help them to move to a better process with lower fluid losses and lower emissions."
PDF attached, what do you make of this.. Is this a marketing thing, or is this for real.
-
PDF attached, what do you make of this.. Is this a marketing thing, or is this for real.
It seems to me like it's for real. As a chemical engineer, everything you quoted from Solvay Solexis makes logical sense. -
Ok, but to play devils advocate here:
Weller and others recommend HT. So it legitimately performs well. HT is 1/2 the price of LS. Let's assume that solvay doesn't have bigger margins on the LS product, but that might not be true given that they have no competitors (they do in heat transfer though: fluorinert, krytox, etc). If you have a sealed batch vapour phase system, then it is effectively a closed loop system so you should expect basically no losses except what sticks to the board.
I can see the that the LS would be better for an inline system which is perpetually open, but will the losses offset the savings from using HT in a batch oven?
I also think that having some lower boiling point fluids would be better; it will make the temperature gradient of the vapour cloud much less steep.
Can anyone shed any light on why fluid viscosity really matters in a batch oven? I can't think of anything.
Also a lot of this kind of reeks of marketing:
"The latter defeats the purpose of moving towards lower environmental friendly lead-free soldering."
This makes sense superficially, but it doesn't really make technical sense. The lead-free soldering move had nothing to do with ozone or CFCs. We can of course argue all day if it is actually better for the environment
, but lead free (afaik) uses the same amount of fluxes, generates the same amount of gas, etc if not more due to the higher temps required.
The actual fluid losses aren't quantified. I'm pretty sure if it was a huge amount, they would use the figure to make their argument. They have a technical audience after all. Something like "in our testing, we found that HT losses were 250% greater than LS at 230C". After reading the technical detail in the PDF, I'd be willing to bet that they've done the test and didn't see an enormous difference.
In the slides, they state that Galden is not a CFC, not toxic, accidental spills/release is ok (even on food!), etc. But when it comes to LS vs HT, HT is pretty nasty stuff for the environment? Um
tldr; do the losses by using HT really account for a doubling in price? My guess is no, or they would have been a little more technical about the whole thing. -
Now I'm trying to think about how to raise the board gently when potentially the balls are still liquid.
I think your biggest challenge here is a smooth motion without rocking and too fast accelaration/deceleration.
A quick though about rocking/shaking is to use guidance rails like CNC machines do.
They are now so accurate the backlash is within 0,01 mm even when milling steel so not a problem for your system.
You have a lot of choice, for instance IGUS DryLin Linear Guide or even HIWIN Carriage HGH system.
For controlling the basket a simple steppermotor and driver would suffice than depending on the total weight you have to move up and down.
So actually you are implementing one axis of a CNC machine with two guidance rails.Quote from: jeremyOr perhaps since it is a hobby system and I only need one board, just turn off the heating and wait for it to cool. Suggestions welcome!
Would not recommend that, it could take an hour too cool down and I am afraid a lot of components can not stand 230oC for more than a few minutes (see datasheets) making it practically unusable. Than also the solderjoint will deteriorate since it is supposed to be cooled down, it might make it brittle since the flux has gone.Quote from: mrpacketheadPDF attached, what do you make of this.. Is this a marketing thing, or is this for real.
Probably real but don't forget these are industries running 24/7 in the week, you should ask yourself how often will you be using this system and how much will you then loose to evaporation. If you make the tank high enough and add cooling above than you should even have to worry less.
-
Now I'm trying to think about how to raise the board gently when potentially the balls are still liquid.
I think your biggest challenge here is a smooth motion without rocking and too fast accelaration/deceleration.
A quick though about rocking/shaking is to use guidance rails like CNC machines do.
They are now so accurate the backlash is within 0,001 mm even when milling steel so not a problem for your system.
You have a lot of choice, for instance IGUS DryLin Linear Guide or even HIWIN Carriage HGH system.
For controlling the basket a simple steppermotor and driver would suffice than depending on the total weight you have to move up and down.
So actually you are implementing one axis of a CNC machine with two guidance rails.
I was thinking this too, I have already have some linear guides and steppers waiting to go. My only problem would be contamination of the fluid with the factory-shipped lubricants, I will have to give them a really good clean beforehand. It would probably be chain driven (too hot for belts I think) as well so that will need to be cleaned too. I do have a spare short ballscrew floating around here somewhere though...
Perfluorocarbons are used as lubricants anyway, so should be no problem running them without lubricant. -
Another thing to test is if the carriage will expand the same amount as the guidancerail it self otherwise it could get stuck at higher temperatures.
A CNC guidance rail is so sturdy that you could let it end well above the higher temperature and just extend the platform. The picture with the external (seperate) lifting system of one of the previous pages comes into mind. -
Another thing to test is if the carriage will expand the same amount as the guidancerail it self otherwise it could get stuck at higher temperatures.
A CNC guidance rail is so sturdy that you could let it end well above the higher temperature and just extend the platform. The picture with the external (seperate) lifting system of one of the previous pages comes into mind.
It's like you're reading my mind! I was just trying to work out if the heat would conduct through the rail and carriage enough to make the problem negligable. I think it would make for a much simpler build if you didn't have to make any extensions. The problem is that this is looking like an all-metal construction as good borosilicate glass is very expensive in large internal diameters, and I think it would be good if we could devise something that the average hobbyist at home could build (aka without a cold saw and milling machine) -
Ok, but to play devils advocate here:
Lower viscosity - better run-off from the PCB?
Can anyone shed any light on why fluid viscosity really matters in a batch oven? I can't think of anything.
-
lower viscosity - better heat transfer to the center of large FBGAs
-
Ok, but to play devils advocate here:
Lower viscosity - better run-off from the PCB?
Can anyone shed any light on why fluid viscosity really matters in a batch oven? I can't think of anything.
Yes, this is about the only sensible thing that I could come up with. But I think this is again less of a problem when you are doing single boards in a slow batch oven. Inline systems sure; of course you don't want the fluid to be taken away by conveyer. I had a look, and both fluids have almost identical surface tension, so given how inert this stuff is (it doesn't really dissove) I'm sure it will be just as easy to pour off either one.lower viscosity - better heat transfer to the center of large FBGAs
I'm not sure this is correct; we're talking about the viscosity of the fluid being heated, not the viscosity of the vapour. Although they probably will be related. As a reference, the HT series, which is more viscous, still has a kinematic viscosity of 10 times less than olive oil. And ~3 times more than tap water at room temperature. Of course this will decrease with heating. -
heat is transferred to the parts by a liquid phase, not a gas. the vapor condenses on the surface, so the board is entirely immersed in liquid, less than 1mm thick. when the peak temperature is reached, the vapor stops turning to liquid. so vapor itself does not do anything to the components.
see the presentation: (the slide decks on this channel are extremely detailed)
especially the point on "low surface tension". Viscosity and surface tension are related, although the relationship is different for different types of materials. -
heat is transferred to the parts by a liquid phase, not a gas. the vapor condenses on the surface, so the board is entirely immersed in liquid, less than 1mm thick. when the peak temperature is reached, the vapor stops turning to liquid. so vapor itself does not do anything to the components.
see the presentation: (the slide decks on this channel are extremely detailed)
...
especially the point on "low surface tension". Viscosity and surface tension are related, although the relationship is different for different types of materials.
I think you have perhaps misunderstood my point, because although the video mentions that surface tension is important, the measured surface tension of both the LS and HT fluids basically are identical. So regardless of how important that particular quality is, there is no difference between LS and HT. In my understanding (which could be totally wrong! my chemistry is a bit rusty), with high surface tension the condensed fluid would be more likely to stick in droplets on various hot parts of the PCB in the same way that you see water "beading" on hydrophobic surfaces. These droplets would absorb some of the energy of the phase transition in their own thermal mass, thus making the transfer of heat into the board slower. You want the low surface tension stuff so that the fluid will be more likely to drip off and flow rather than stick together, leaving a higher thermal gradient to encourage the phase transition.
The viscosity of the HT fluid is likely much lower than water at ~230C, and washing boards in water does not ever seem to be a problem in terms of its viscosity.
I also don't think it is as simple as saying that the fluid does all the work. The vapour carries the potential energy via its latent heat of vapourisation. When the vapour comes into contact with the cold board, the board will absorb the heat as a counterpart to the phenomena that allows the vapour to transition to a liquid phase (in a pretty yin-yang kind of way). So it is really neither the vapour nor the liquid that does the work, but the action of transitioning between the two which releases the heat. This transition is prevented when the board is the same temperature as the liquid as the vapour cannot deposit it's energy anywhere in order to return to the liquid phase. I think heating merely from the liquid being hot would be much less significant, and could almost to the point of being negligable.
If you look at the lovely wikipedia picture I found (attached, taken from http://en.wikipedia.org/wiki/File:Energy_thru_phase_changes.png) you can see that for water, it takes only 8kJ to reach the phase transition zone, but then it needs a whopping 40kJ to actually turn into steam. So if we were using water as the vapour medium, even if the condensed liquid cooled all the way to 0C, it would only deposit 8kJ of energy (ignoring the phase transition energy). However, if the water vapour used the thermal gradient of a cold board to condense, it would heat the board 5 times faster, even if the water only cooled by 0.1C before dripping off the board. So it is the transition itself which is important
Thank you for sharing the video, it was very interesting. I will have a dig through his channel later.
And just to reiterate, I'm not arguing that HT is better in performance, I'm just arguing that it is perhaps better in performance/$ for a small batch oven. It is double the price after all.
-
I think you are focused too much on the cost of the fluid. it could cost 10 times as much and it would still be less than the cost of engineering a proper heat loop, cooling system, mechanical elevator, vapor detector, control system, user interface, and chassis.
-
I think you are focused too much on the cost of the fluid. it could cost 10 times as much and it would still be less than the cost of engineering a proper heat loop, cooling system, mechanical elevator, vapor detector, control system, user interface, and chassis.
I disagree. This is on my own time, so time isn't worth all that much and I think I already have most of the physical bits already (including linear bearings, motor drivers, chains, many devkits I can use as the controller, thermocouples, induction heating plates). So the major outlay for me will be the fluid, especially given that I'll probably mess up and lose some. This isn't a 24/7 machine, it just has to work while I'm looking after it. Even the 3M system which they used to solder expensive Intel CPUs is just a dumb heater, plus some water flowing around a big test tube.
Plus, I think it would be more applicable to others using the cheaper fluid. -
Well, don't let me be the bearer of bad news
One last video as cautionary tale:
Notice how vapor is pouring out of their setup. This means that the heat input is in excess of what is required for the small board they are reflowing. It causes tombstoning, and look at that BGA
this is what VPR was like in the 60s when it was abandoned. -
Notice how vapor is pouring out of their setup. This means that the heat input is in excess of what is required for the small board they are reflowing. It causes tombstoning, and look at that BGA
If i saw the presentation on the previous page they measure the temp on three vertikal places and lower the board when there is no vapour yet (so the Galden temp is around 200) than apply the extra heat and control it so the temp is following the reflow profile. That is how i intertreted it.
this is what VPR was like in the 60s when it was abandoned.
The video you showed is amateur time no smooth operating elevation, no temperature control what I can see and they are boiling the Galden even way before the pcb is lowered.
And even worse no lid to contain the vapour to let it come in contact with the coolong area and liquidize back.
-
Notice how vapor is pouring out of their setup. This means that the heat input is in excess of what is required for the small board they are reflowing. It causes tombstoning, and look at that BGA
If i saw the presentation on the previous page they measure the temp on three vertikal places and lower the board when there is no vapour yet (so the Galden temp is around 200) than apply the extra heat and control it so the temp is following the reflow profile. That is how i intertreted it.
this is what VPR was like in the 60s when it was abandoned.
The video you showed is amateur time no smooth operating elevation, no temperature control what I can see and they are boiling the Galden even way before the pcb is lowered.
And even worse no lid to contain the vapour to let it come in contact with the coolong area and liquidize back.
And if you find the long version of the Video,you'll find that they actually poured water into the tank. ( why we don't know, but these were engineering students, anything is possible when you know everything ). Galden vapour is transparent, you can't see it, so that vapour is probably boring old H2O.
-
Some more data, that i was able to obtain tells me that;
The Distillation spec range of the HT230.
10% > 210C, 90% < 250
Compared to the LS which is
10% < 222, 90% < 235
The viscosity of the HT is between 3 and 7, and for the LS its 5-6 cSt.
The base material of the two products is the same thing, its how it is processed.
I am in a very interesting discusion with a supplier who can further distill the HT230,, and bring the spec very close to that of the LS-230. 7kg of this 'tweaked' product, will be around $1000 for 7kg, which compares favourably to a the $1400 or so for 5kg of LS-230. I'm now very tempted to proceed on this.
-
And if you find the long version of the Video,you'll find that they actually poured water into the tank.
That does sound awfull, mixing a liquid with a boiling point of 230oC with one that has a boiling point of 100oC is rediculous, so scrap that video as any serious source of information.
-
And if you find the long version of the Video,you'll find that they actually poured water into the tank.
That does sound awfull, mixing a liquid with a boiling point of 230oC with one that has a boiling point of 100oC is rediculous, so scrap that video as any serious source of information.
Heres the long version.
I love the combination of AC mains, very loose wire, water, and working out the front of the garage.. What do you make of the platform. Its awesome. Listen at 1:20.. " that could be all you galden esacping.. I hope not.. Do you want to put a bit of water in there? "
I wonder if what they were trying to create some kind of 'cover layer'?
I would'nt completely right off this video. I think these guys should get some points for trying. And they did actually get something soldered. I'm really curious to know what the motivation for the water was..
@6:43.. what the heck was that.. Did you kick it.
@7:18.. I better help that stepper motor.
@7:47.. oh "f...k" it soldered.
-
Interesting thread guys, hope it all comes together for you.
As I see it a deep vessel will be very important to contain and not waste vapour.
This is presumably where the pre-heating comes in to play with the different temp zones as you go deeper.
Is there agreement that there should be a cooled lid?
Had a look at the info linked and it's unclear how much vapour would be produced per say each litre of Galden.
Wondered if a circular domestic stove element would work as a heater?
Say a 6" one with it's thermostat, real easy I think and they're only ~8mm thick plus bracket for a total say 15mm. Should easy cover it with a litre.
-
Listen at 1:20.. " that could be all you galden esacping.. I hope not.. Do you want to put a bit of water in there? "
too much credit, if they were afraid that their costly Galden was evaporating their thought probably was too cool down the fluid so it stopped boiling/evaporating, hence add some cold fluid, like water.
I wonder if what they were trying to create some kind of 'cover layer'?
...........
I'm really curious to know what the motivation for the water was..
What they should have done was to turn the temperature control down and rework their control loop. But hey that needs some time to think.
-
Interesting thread guys, hope it all comes together for you.
As I see it a deep vessel will be very important to contain and not waste vapour.
This is presumably where the pre-heating comes in to play with the different temp zones as you go deeper.
Is there agreement that there should be a cooled lid?
Had a look at the info linked and it's unclear how much vapour would be produced per say each litre of Galden.
Wondered if a circular domestic stove element would work as a heater?
Say a 6" one with it's thermostat, real easy I think and they're only ~8mm thick plus bracket for a total say 15mm. Should easy cover it with a litre.
I think we are all in agreement that there needs to be some cooling at the top. Whether that is a cooled lid, cooled sides or something else (cold mesh perhaps?) I think still needs to be determined.
The amount of vapour you get will somewhat be a function of the spatial distribution of heating and injected energy, as well as possibly your own individual batch of fluid. So I'm not sure it is easy to just give an amount of vapour per litre. I think the more useful characteristic is the vapour "height", which seems mostly to be controlled by temperature/injected energy. The better you can control the vapour height, the better your oven will be.
I'm sure many sorts of heaters would work, including the one you mentioned. It seems to be more a problem of controlling the fluid loss while still getting adequate vapourisation.