High temp. boards

[Sarke]

Hey peeps,

I have been building boards for a couple of years now for baking in 130 °C (266 F) for several years continously using the SH260 material, but the last quote I just got from PCBway was over 2000 USD 

Does anyone know if an aluminium PCB would also be able to withstand the same temperatures for that duration without warping and outgassing? (There is no component heat generation, so the board will have a very homogenous temperature distribution for the entire duration. The applied voltage is like 4V max.) I also see there is a Tg170 variant now that wasn't there the last time I ordered, maybe this would also fare better than the Tg150 boards that were totally obliterated a few years back? 

[voltsandjolts]

Aluminium PCB just means a PCB bonded in some way to aluminium, so that's not a solution in itself. It may be worse. Sticking with polyimide is probably the best option. Why the requirement for SH260 when there are plenty other good polyimides? I've seen Isola P95 used in high temperature environments before but typically 18months max.

[thm_w]

AllPCB says the following:

3. Dimensional Stability

Aluminum substrate printed circuit board has apparently higher stability than the insulating material of the printed circuit board. When heated from 30 ° C to 140 ~ 150 ° C, the dimensional change of aluminum substrate is only 2.5 ~ 3.0%.

https://www.allpcb.com/pcb/aluminum_substrate.html

I don't see anything from JLC or others, maybe ask them.

[Sarke]

2.5-3% is kind of a lot tbh. In any case, I don't think an absolute dimension change is the main villain here, last time the board outgassed a lot over the years and warped, the opposite ends of it were probably somewhere around 8-10 degrees different. Several traces had peeled off.

Ended up ordering with just Tg170-180 material and praying for the best, if it doesn't work I will try Rogers 4003C which apparently has gotten quite a lot cheaper

Thanks for the replies!

[voltsandjolts]

What's the application, if not confidential?

[Sarke]

Mounting and connections for lifetime testing of parts at very high temperature 

[Doctorandus_P]

How big of a problem is outgassing?
Or delamination of copper tracks?


Have you considered just using bog standard FR4 and then only use internal layers, and putting it on a relatively solid frame with lots of screws and spring washers to prevent warping?

Do you need SMT, or can you limit yourself to THT for this?

[voltsandjolts]

Mounting and connections for lifetime testing of parts at very high temperature 

OK, you’re in Norway doing high temperature electronics, so reading between the lines I would guess you are in the oil & gas downhole electronics business. Cumulative time at temperature is the enemy, for higher temperatures and longer durations in a sealed chamber you do need to consider outgassing and restrict the materials you put in the chamber to reduce outgassing as much as possible. Perhaps you want to select plastic packaged parts with reduced bromine or perhaps you want to select hermetic sealed parts, or go full hybrid (à la Quartzdyne) for maximum endurance. But if all this is battery powered anyway, then there is no point in going for mega-durations, batteries will be the limiting factor, especially so at high temperatures >135C where self-discharge will really start to bite. And if you are working sub-150C you are on easy street anyway but polyimide boards are still your best bet for duration (I mean, polyimide cost versus the effort you’re putting into development is nothing). Use that and worry about the other stuff.

[Sarke]

OK,...

That was some decent guesswork However this is actually only for tests up to 130C in Binder heat chambers. We have somewhat decent control of the board mounted components we are using (however thank you for the suggestions on part packaging options, there were useful lessons in the papers I found when googling), as well as the boards we use for commercial applications. The only reason I'm interested in the potential use of non-polyimide boards is because the board cost actually becomes quite significant in some of the high-volume experiments with many different designs.

Do you need SMT, or can you limit yourself to THT for this?

Both, unfortunately.

How big of a problem is outgassing?
Or delamination of copper tracks?

Outgassing, delamination and massive warping is what happened to the first test in 2016 by our board design partner, running at 150C, after only a couple of months I believe that was TG120-130 material. Kind of a surprising mistake to make. I made this thread because I have the same question as you, for TG170 boards.

Have you considered just using bog standard FR4 and then only use internal layers, and putting it on a relatively solid frame with lots of screws and spring washers to prevent warping?

I guess the degassing in itself is not a problem, but the loss of stiffness and increased warping. Fastening the board in many positions to preventing could make sense, however it will be a lot of extra work. Will look into that, thanks.

[Doctorandus_P]

From another unconventional viewpoint:

How complex is the circuit that has to be in this chamber?

Is it feasable to glue the parts to some substrate and then just air wire it?

[Sarke]

From another unconventional viewpoint:

How complex is the circuit that has to be in this chamber?

Is it feasable to glue the parts to some substrate and then just air wire it?

I love unconventional! Almost everything I work with is firmly within that category  And you're right, it's very feasible and I sometimes just do it that way (with 2 component epoxy) but here's a list why I usually want PCBs, especially for 20+ samples experiments where I need consistent experimental conditions for each sample.

• Total flexibility for board mounting dimensions and methods
• Holes for strain relief, wire holding and metal strips usage wherever I need them
• It looks professional
• I can't just unscrew epoxy when I want to take the part back off
• Incredibly high level of consistency between each "channel" of the setup
• Some of my experiments require SMD components.
• Soldering a a long row of lined up, identical SMD components is incredibly time efficient
• When you think about it, PCB material is actually a pretty cost efficient substrate for what you are describing as well, for these ambient conditions. If I had a 3D printer that could print PEEK (which needs a very hot heat chamber) I would probably be replacing my boards with printed parts in many of my experiments.