2

Would a 20-30cm microSD extension cord suffer any kind of noticeable transmission loss? This is a passive extender - literally, a microsd card with wires coming off of it, and a card slot at the other end. Simple ribbon cable connecting the two ends.

EDIT I'm not making this - it's a store-bought thing. Here's a picture of what I'm talking about:

enter image description here

kolosy
  • 2,308
  • 6
  • 29
  • 52

2 Answers2

3

I can't say with any certainty, but I think that for 20 to 30 cm it should work, as that length is just a small fraction of the wavelength of the data frequencies (around 10Mhz). But consider to use a ribbon cable with twice as many signals as you need (plus a couple of more) and connect every other wire to ground so that they decouple the electric field between the adjacent wires (each signal wire sees a ground wire at each side).

Use a couple of extra wires for the power and I suggest to put a small ceramic (or plastic) capacitor (something like 100n to 1u) between the VCC and GND at the microSD card's end to supply the drivers with some energy to make clean transients.

The cable will have some electromagnetic emissions, I'm afraid. But probably not so much that it would become a problem.

PkP
  • 7,677
  • 2
  • 24
  • 42
  • I'd also add that if you are able to control the SD interface's bus clock, and if you're seeing a high error rate (lots of CRC fails, for example), then try slowing the clock down. – markt Oct 30 '14 at 23:00
  • I would need some explanation for what's meant by "emitting ESD" if I were to understand what your thinking is? – Rolf Ostergaard Oct 31 '14 at 12:44
  • Rolf, I'm thinking that there's a lot of digital signals toggling at rather high frequencies and there are a lot of hopefully quite sharp edges at those signals, so since a ribbon is not shielded, I think it's going to have some electromagnetic emissions. To say that it emits "ESD" was a silly mistake made by very tired brains... sorry. I'll fix that in the text. – PkP Oct 31 '14 at 13:43
  • It looks like these things were done, so I should be good. – kolosy Oct 31 '14 at 16:13
0

Transmission losses should be the least of your worries at those data rates.

However you need to think about all those other things that may be much more likely problems:

  • Reflections. When you extend the length of the electrical path, you are more likely to get enough reflections that requires termination. With very short path, you may get by without termination if the electrical path is sufficiently short compared with risetimes. In order for termination to work well, you need a cable with well defined impedance. A ribbon cable may be okay, if signals/grounds are carefully placed. Loose wires all over the place may (most likely) not be okay.

  • Crosstalk. With a longer path and especially in a cable, you may get more crosstalk for two reasons:

    1. The conductors may be closer together compared to your PCB (and for a longer run, which would matter if you compare to really short traces on a PCB)
    2. The ground is further away and most likely not a solid plane (unless you have ground as a shield) creating better opportunity for field lines to go around two signal wires.

  • Ground bounce. If the cable has insufficient ground return wires, you may see ground bouncing up and down at the card end due to L * di/dt in the (too few) ground wires.

Oh, and on top of this comes timing. But I assume you have that analyzed already both with and without the extra flight time of about 2ns.

Isn't it funny how digital SI always come down to these few basic things (plus a few others including loss when signals are really fast; think 5GBps+)?

PS: As for the cable "emitting ESD" as some other poster noted - that does not make sense to me or resonate with any experience or theory I can come up with.

Rolf Ostergaard
  • 4,528
  • 19
  • 24
  • You raise valid points, could you confirm or debunk some of my reasoning? 1) Reflections: in RF theory in my years at the university I was taught that you don't need to worry about reflections when the overall transmission line is less than 10% of the wavelenght, which should well be the case here (wavelenght of 10MHz is 30 meters) 2) Crosstalk: having each other signal at GND should help with this. 3) Ground bounce: see (2). Personally I'm more worried that the signal degradates due to capacitive loading on the signals and that the VCC might bounce down and up. – PkP Oct 31 '14 at 13:54
  • 1
  • Wavelength is not relevant for digital. Rise time is. 2) Sure, but again maybe not enough. Analysis required. 3) See 2 :-) And you are right about VCC ripple at the card being a potential problem as well.
    • – Rolf Ostergaard Oct 31 '14 at 15:20
  • Thank you Rolf, I agree with you on all points. I'd like to be optimistic and I think that his design has a good chance to work, at least if he can control the SD host, so that he can choose the speed at which the card works so that the digital signals are good even with rise times that are somewhat lengthened due to capacitive (and inductive) loading. Though writing to the card might require especially strong VCC and more capacitance since the microSD needs a lot of current when writing to. – PkP Oct 31 '14 at 15:30