flexible ribbon "PCB" to regular PCB

I've seen them soldered directly to a PCB- it was very unreliable so don't go that route (on an LCD to PCB backlight)

--
This email has been checked for viruses by Avast antivirus software. 
https://www.avast.com/antivirus

OK, good advice. I can use the recommended connector, and after mating the ribbon, seal the entrance and cover it all with epoxy. I see 1.25mm six-wire is only 9 mm wide, but I assume it's strong.

--
 Thanks, 
    - Win

Pass it through a slot in the board first to give it more strength. Epoxy it there too.

then you're out of luck, solder isn't that robust. You can feed it through 2 slots to improve things, but it still won't be long term tuggable. Maybe 8 slots :)

NT

Awesome idea!

--
 Thanks, 
    - Win

I'd prefer to use clear silicone- a bit of flexibility/resilience

--
This email has been checked for viruses by Avast antivirus software. 
https://www.avast.com/antivirus

Yes, but it peels off easily. Hmm, hot glue is slightly flexible, and doesn't peel quite so readily.

--
 Thanks, 
    - Win

Depends on type, of course. I would also recommend 3M VHB tape. Maybe that, inbetween two slots, with the connector on the exit side, tape on the opposite side and the cable routed between all three in a zig-zag fashiion, would be best.

Incidentally, mind EMI. I2C is not "inter-printed circuit" for a reason. Even with a ground plane on that cable, don't expect much performance!

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Design 
Website: https://www.seventransistorlabs.com/

God ideas, thanks!

We'll be using a slow data rate. With a choice of pin order, how would you arrange +3.3, SDA, SCL and GND?

--
 Thanks, 
    - Win

+3.3, SDA, GND, SCL, GND. Or actually, putting 3.3 in the middle and GNDs at the ends may be better for hot-plugging or cable-wearing purposes, but not relevant electromagnetically (as long as there's a bypass cap from 3.3 to GND at each end, which there usually is; for the same reason, always connect all GNDs and supplies at the connector, whether you need the current capacity or not).

I forget if 5-pin cables or connectors are common. Maybe you're stuck with

4 for space, or 6 for availability?

Low data rates may not be enough, because of noise generating false clocks. I2C is generally tolerant of fast noise, but a cable exposes a much broader bandwidth than PCB traces do; it can pay to add ferrite beads (for damping and filtering) and small caps (as many pF as you can afford, usually

100-470) at both ends of the cable. Uh, beads facing the cable, that is.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Design 
Website: https://www.seventransistorlabs.com/

Beads on a kapton cable? (Or "flexible PCB" as it was described.)

Or make it wide enough for a hole or 2 and screw it down. Glue below and washer on top.

Or if PCB slots or washers are too sharp on the edges, and might cause it to break, a few large guage wires with insulation, can be mounted like jumpers closely in parallel. Then you can zigzag around them, and glue. That should give it the least stress.

"Tom Del Rosso" wrote in message news:qacn7r$kvg$ snipped-for-privacy@dont-email.me...

Ah, on the signals that is. :)

You would be able to put a clamp bead over the cable though, which is also a good idea if you find noise to be a problem.

Did basically that, five years ago, on a device with a mainboard and a HID board, connected by a 6" ribbon cable carrying I2C. It failed radiated susceptibility at ~200MHz (basically the resonant mode of the two boards joined by the cable). A clamp bead fixed it (damping the resonant mode --> less V dropped across the cable at F_res).

That same project, had a number of issues that we strongly recommended against, but which the customer demanded regardless. The elephant in the box was they wanted USB over unshielded cables... Apparently it managed to pass radiated, but EFT knocked it out. We were forced to use shielded enclosures, cables and connectors (instead of nice cheap plastic automotive connectors), nearly doubling the product cost...

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Design 
Website: https://www.seventransistorlabs.com/

One possibility (that i've never seen used) is to put holes in the flex, to engage pins, for mechanical positioning, then apply a clamp over a flex/ zebra-stripe/PCB sandwich. This isn't much of an improvement over 'using a connector', though, it has loose parts to be assembled. With the right plastics manufacturer, it'd be two; zebra elastomer and snap-over plate (holes in PCB and flex engage the plate's plastic snaps).

Win - ribon cable, ala Belden, etc, is a "golden oldie" for such applications. Typically, though, a ground lead for each signal lead is required.

Hul

W> Looking for good ways to connect / solder a flexible ribbon PCB to a regular

Thanks, but this time we need the super-thin aspect. It sneaks between two "supers" = bee-hive sections. We only need four I2C wires, but the smallest stocked Molex 1.25mm Premo-Flex has six.

--
 Thanks, 
    - Win

does it have to be molex and 1.25?

or

No, it could be either of those, but then I'd have to be confidant of the mating connector. We do need super thin. Haha, I'm not really sure of the Molex Premo-Flex thickness spec. I can see it's really thin.

--
 Thanks, 
    - Win

lar

ink,

c., to

I2C

They make a flat cable termination that is intended for soldering to a PCB. It has pins for through hold mounting, very secure. I want to say the in dividual pins in the body crimp to the flex, but I'm not certain of that. It might need to be made part of the cable. Do you not see anything like t his in premade flex catalogs?

--

  Rick C. 

  - Get a 1,000 miles of free Supercharging 
  - Tesla referral code - https://ts.la/richard11209