ribbon cable current (2023 Update)

Ralph Mowery snipped-for-privacy@charter.net wrote in news: snipped-for-privacy@news.eternal-september.org:

Cheap jumpers typically also have very poorly crimped alligator clips on them as well. I used to rework mine all at once, the moment I open the package. My preference was actually to make my own sets for the labs I worked in.

Actually, potting would improve the thermal situation some, but we don't plan to encapsulate a $12,000 e/o modulator.

snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:

That is why we always used SPC (Silver Plated Copper). That stuff takes solder wetting better than nearly everything else. And oddly as well, we had a small amount of TPC and if it had a PVC jacket, it was almost as hard to tin as a still coated transformer wire. One had to use a very active flux to get it to wet and oh... lookout that you don't let the heat zoom up the wire and melt a 5D length of the jacket and more as it pulls back. That is why I also like teflon. PVC just has such poor workability for me. It would have to be a minimum of TFFN or THHN.

snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:

Place it in a can and pot the can to insulate. That way you retain servicability. Put (thick) Nomex paper stock inside the can as an internal insulator layer (ouside too really) and then pot it suspended in the center of a cavity that is the max size you want the entire "oven assembly" to be.

Any potting compound conducts heat hundreds of times better than air.

Almost anything conducts heat better than still air. Styrafoam. Fiberglass.

Here's what I'm thinking now.

The SMA feedthrus screw into the walls, to thermally bootstrap the short hardlines inside. I can use really skinny coax for the short jumps.

The oven box can have 1/2" thick walls all around. Maybe more.

snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:

What you are trying to control is rate of change or capability to obtain stability within your chamber keeping that rate at or near zero and even adjusting the temerature inside to match the temperature it was calibrated and characterized into your circuit at. Hence the term 'oven'. So it needs an internal heat source and a monitor circuit so you can energize the heat source as needed. The rest is about the box's ability to keep a temp for a long period. The heat source can bring it up as needed, but that has power costs, so battery fed designs require power decisions to be made, so you might want the best box available. You could pot it and place that inside a vacuum insulated enclosure. Far more real estate needed there.

Better than standing air... yeah, sure. But moving air carries heat away. It would not be better than that. The potting compound is a stable blanket. The liner of the cooler (oven). And there are compounds which conduct slowly. Slowly enough to be better than an open tin lid getting splashed with moving air. Putting it in a box IS making a better, more stable chamber. It all comes down to application and available (circuit board) real estate you have.

I mean you could use the insulation wadding the industry currently uses to insulate IR calibration oven walls so that they are very uniform in temerature. And a silica based "plaster of Paris" stuff to paste on the inner AND out layers of a basketball sized chamber (literally that's what we used). Then we suspend the sphere inside a huge steel cube also packed full of that insulative wadding. Put a very precise PID controller on it and VOILA, you have an NIST traceble IR calibration oven for your high temp IR process instruments.

You would just make a miniature version of it.

They make stainless hardline, that would have much lower thermal conductivity. I suspect they would silver plate the parts where the RF current flows. Also tiny diameters, like hypodermic needles.

On a sunny day (Tue, 17 Aug 2021 16:57:03 +1000) it happened Chris Jones snipped-for-privacy@spam.yahoo.com> wrote in <65JSI.797730$ snipped-for-privacy@fx13.ams:

If you want low / zero thermal conductivity use magnetic induction coupling. No idea what exactly he is heating, I see a metal housing, that conducts heat away to start with then you need more heating power.

Reminds me of Bill's thermal design with one I/O that was the point of high thermal conduction. He made that the point where the temperature was controlled if I recall. So put four heaters and sensors at these four points to maintain a constant temperature on the coax. Adjust the set point of those control loops to control the overall temperature. Oh, and consider replacing your big aluminum block with styrofoam. It is not at all clear how that will be a good thing if it is going to have thermal holes in it. You would be starting with a problem that is hard to overcome rather than designing a good solution from the start.

There is s/s hardline, some with a very thin copper layer on the inside of the shield to improve HF behavior. Sounds expensive. We'll probably just go with very skinny coax and thermally bootstrap it where it passes through the walls of the oven.

On a sunny day (Tue, 17 Aug 2021 10:34:13 GMT) it happened Jan Panteltje snipped-for-privacy@yahoo.com wrote in <sfg3oo$1jpj$ snipped-for-privacy@gioia.aioe.org>:

PS you can heat and power an object by shining with a laser diode (few $ on ebay) on it too. Maybe even measure its temperature remotely (IR). No need for any mechanical contact, and if it is an oscillator pick the signal up at some distance.

I doubt it would be expensive, in absolute terms, unless you are using many feet of it per box. The main nuisance might be soldering to it, but if you're having the cables formed and terminated for you, then they might have figured out a way of doing it that doesn't add much cost. But yes, using thin regular micro-coax might be easier, unless you have to handle a lot of RF power or otherwise need very low losses where small diameters are unsuitable.

I'm hearing an expensive electro-optical modulator that people want to be kept stable and isothermal to ca 1 mK. One version is a long skinny stainless-steel gadget that's a terrible heat conductor. I need to wrap it in a solid isothermal oven. A quartz crystal is tiny and easy to temp regulate by comparison.

A crystal needs two tiny wires to connect it, and I need four hardline coaxial cables. That doesn't help the thermals at all.

Fun problem. It would help isothermality if the oven box could be machined from isotopically pure diamond.

Two of the connections to the e/o modulator are bias, which is microsecond stuff, so we can use really skinny braided coax for those. Two are "RF", namely 10 GHz waveforms, so they need good cables. But only the last couple of inches, so we can use skinny cable there to trade a little loss against thermal conductivity.

It's surprising how non-isothermal nice looking metal boxes can be.

A set of long parallel heat pipes within the oven box but outside the modulator would flatten the longitudinal gradients down. This would be a kind of isothermal furnace liner. Unless one is making many such heat pipes, it's better to just buy them. There are many manufacturers.

Joe Gwinn

Interesting. Maybe we could buy tubular heat pipes and poke them into bored holes in the oven block.

Pity, that might raise the price.

You can get flat heat pipes that make excellent spreaders--startlingly better than copper.

Cheers

Phil Hobbs

I was poking around and found rectangular ones at Digikey.

We could mill some slots in the bottom of our aluminum-block oven and press or epoxy these in. That would iso-thermalize the gadget in its long direction.

Easy enough to allow for. I've ordered some to play with.

They used to be unreliable but I think that they are OK now.

Do you know if heat flow is linear on temp difference? Like a chunk of metal? I am trying to squash small gradients.

As long as the vapour pressure is high enough at the source side temperature, they work fine at low delta-T.

Cheers

Phil Hobbs