ribbon cable current (2023 Update)

I am looking into similar situation, but with 3M ZIF socket. The pin itself should handle 3A, but the socket is only rated for 1A. It really depends on the contact area of the socket. So, i cut the tabs on the socket to allow pins to insert more fully, and also double up on the pins. I am hoping to drive 3A through the pin/socket.

More specifically, it counts on the current rating of the female part of the connector pair. This determines the contact area, requires the use of non-deal conductors and suffers most in plating and physical degradation in repetative use.

Avoid connectors where this is important. The female parts should be considered as maintenance replacements, and not be on the fixed or captive part of the cable assembly.

Some zif flat cables have good ratings and thermal limits at the connector, but there's a lot of rubbish as well.

RL

On a sunny day (Fri, 13 Aug 2021 13:32:39 -0700) it happened John Larkin <jlarkin@highland_atwork_technology.com> wrote in snipped-for-privacy@4ax.com:

Hard to tell, I have 3 variants of those flat cables here, the normal ones bought locally, the Chinese ones form ebay (no idea whet the conductors are made of?? but those are much thinner, and the flat colored ones that I got from a local raspi shop.

I would trust the Chinese ones with perhaps 50 to 100 mA at most.. or better 10 mA for LEDs only ;-) The other ones 100 mA to 200 mA???

Copper is expensive! You may not get as much from noname chinese vendors as you get from 3M. Read "Poorly Made In China."

We use #28 AWG stranded ribbon cable, but #30 is common and I think there is un-stranded too. CWG?

I think I'll go for 1 amp per wire, 24 volts out to my heater assembly.

Looks like we'll make a full oven, machined out of big chunks of aluminum, totally enclosing our e/o modulator. I'm thinking four TO247 mosfets on the bottom as the heaters, and thermistor Wheatstone bridges inside as the sensors. I'll post pictures eventually.

Getting the hardline coaxes in and out is thermally interesting. I don't want four cold spots on the side of the EOM.

I have seen what appears to me aluminum wire coated with copper. That is on some 16 gauge stranded low voltage wiring. I did not check to see if the wire was larger in diameter to make up for the aluminum or not.

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

Max out. 5 bucks a foot.

We plan to use tiny FFP flat ribbon cables to sneak the thermistor bridge connections out of the oven with minimal heat conduction. But the RF paths have to have 10 GHz bandwidth, so they need serious hardline or at least semi-hardline coax. I'm thinking about SMA feedthrus screwed into the wall of the heated box, and then short custom coax jumpers inside, to sort of bootstrap the heat flow.

All that sounds expensive.

Ovens are tricky, and often done badly. The usual chunky-looking aluminum box won't be very isothermal in real life. Copper is a bit better but hard to machine.

Am 14.08.21 um 19:58 schrieb snipped-for-privacy@highlandsniptechnology.com:

cheers, Gerhard

On a sunny day (Sat, 14 Aug 2021 08:05:04 -0700) it happened snipped-for-privacy@highlandsniptechnology.com wrote in snipped-for-privacy@4ax.com:

In cases like when I needed some real current I always used a separate power connector Dunno how long your flat cables are, but at 2A you will see some voltage drop.

One thing occurs to me is a balanced open-air twinline transmission line bridging the hot-cold gap. Just two thin copper wires spanning the gap, with a balun at each end. With 50-ohm input and output, a convenient twinline impedance is 200 ohms.

There is no shield. Or if there is, it can be two electrically overlapping pieces having no direct metallic connection.

This greatly reduces the amount of metal spanning that gap.

A alternative is a single dielectric-clad copper wire with launch and receive horns, and no shield. The copper wire guides the waves, so they don't radiate away. This was used in some military communications links between nearby components in the field.

Joe Gwinn

That's cool.

They don't seem to mention feedforward compensation, where ambient temp is sensed and a small tweak is applied to the setpoint. SRS does that in their (badly designed) OCXOs.

We will be able to do that, with a cal factor that scales ambient into the oven setpoint. It might well be (duh) that heater power is a better proxy for heat loss than ambient temp.

That's copper clad aluminum, or CCA. Lots of it on amazon and ebay, sold by actual size so up to the user to derate compared to real copper wire, but much cheaper than pure copper wire so sucks in lots of uninformed buyers. Besides the lower ampacity it tends to squirm and loosen in copper crimps and screw terminals due to the different thermal expansion coefficients of aluminum and copper, and I've been told it also tends to corrode over time.

Ribbon cable has quite high resistance too. I once had to fix a design where 5V 0.5A was 3V at other end of cable.

Am 14.08.21 um 20:45 schrieb snipped-for-privacy@highlandsniptechnology.com:

From the HP Santa Clara bunch.

Ambient temperature is badly defined. Which side? air/contact? Outside influence should be avoided. Peltiers are useless therefore to control temperature tightly.

In my SR-620 counter there is a Wenzel OCXO.

Maybe you can make compromises with the signal quality across the oven boundary with things like a Kapton ribbon and an eye opener chip inside. After all, our 10 GB/s XFP modules had card edge connectors.

In your search for a job title the word "senior" is missing.

Gerhard

I bought some of the cheap alligator on both ends jumpers from China off ebay. They were only about 2 feet long and small wire. I was not using them for much current but my calculations seemed way off on some equipment I was testing. Come to find out those jumpers had a very high resistance for the length. Replaced the wire on a few of them with # 24 and the calcuations were much closer.

There's lots of "tinned copper" wire sold on amazon and ebay, for jewelry making. It's very hard, doesn't solder well, and doesn't conduct well.

Easy, though, to draw or spin. Hemisphere shapes are relatively trivial to make a form, for spinning on a lathe. Cu, too, is easy to weld (maybe easier than Al).

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

The semi-rigid stuff should encapsulate well enough for you.

You have seen the paralell sets of runs inside the quantum computer designs? They have to "keep those signals separated" -The Offspring

Gerhard Hoffmann snipped-for-privacy@arcor.de wrote in news:sf90ri$aj5$1 @solani.org:

Gotta "keep 'em operatin'" -Offspring