Focusrite FireWire interface headphone output repair

[vvangelovski]

I've go a Focusrite Saffire Pro 14 which works fine except for the headphone output on which one channel is dead. This is how the daughterboard looks like this:



Fairly simple thing, just the two coupling capacitors going from the molex that connects the mainboard to the output jack. Now, the dead channel capacitor doesn't pass any signal, the other one does. Also the dead channel cap doesn't pass a continuity test from my DMM while the other one does. So I'm 100% sure it's just the capacitor for the channel. The problem I have is I have no clue what the values for these caps should be. I can't get a consistent reading on the live one. Also, I've no idea what type of caps these are. Each device seems to be soldered at 4 points, 2 leads and 2 solder joints on the actual capacitor housing, no markings on the devices.
Any ideas on how to measure these and what type of caps to look for?

[Richard Crowley]

I doubt very much that those are "capacitors".  It seems much more likely they are filters.  Perhaps a choke (inductor) or maybe just hard connection in series (across the long dimension) and a capacitor to ground (across the short dimension).  If you are measuring "continuity" between the far ends, then that is not a capacitor.

I would try simply shorting across the device (the long dimension) with a tiny piece of wire. I would bet there is a 98% chance that will restore functionality with no significant side-effects.

[vvangelovski]

It could be that they're not caps, my only clue was the markings on the PCB (C14, C15). Yes, now that I think of it, a working cap shouldn't pass a continuity test. Also, there's no measurable resistance on the working one, so it could be some kind of coil. Also, there is continuity between both the short and the long ends on the working channel, but on the dead channel there's only continuity between the short ends. The short ends of both devices are connected together on the PCB.

[amyk]

SMD feedthrough caps.

[vvangelovski]

SMD feedthrough caps.

I opened my older USB interface from the same manufacturer to see if the headphones daughter boards are the same. Well, they're not the same, but pretty similar. The old one has through hole instead of SMD filters and they do have markings, they are a Murata 102/100v EMI Supression Filter. So I think it's essentially the same thing you posted, only 1000pF instead of 2200pF. I'm not sure how to measure the capacitance on the working part to get a ballpark value. Maybe measure between one of the signal terminals and the middle terminal?

[Richard Crowley]

The two far ends are ONE side of the capacitor, and the two center terminals are the OTHER side of the capacitor. The exact value is pretty much arbitrary. It is an RFI filter and has no effect on the audio performance. Unless you live in a high-RFI environment, you could probably remove and bypass it and you would never know the difference.  This is not a primary piece of the puzzle. Don't sweat the exact details.

[vvangelovski]

The two far ends are ONE side of the capacitor, and the two center terminals are the OTHER side of the capacitor. The exact value is pretty much arbitrary. It is an RFI filter and has no effect on the audio performance. Unless you live in a high-RFI environment, you could probably remove and bypass it and you would never know the difference.  This is not a primary piece of the puzzle. Don't sweat the exact details.

Thanks a lot for your help.

I'm pretty sure I'll be able to find the exact replacement by doing some measurements. Yes I know that just shorting these bastards won't affect the audio output since they are supposed to filter in the MHz range and you can't hear any EMI in that range. The thing is that it is a reasonably expensive piece of equipment and I'm a bit OCD about replacing the exact parts. I'm not too worried that anything before them in the signal chain is dead because there is signal up to the filter. The channel wasn't working out of the box, but I needed the thing at the time and didn't need to plug headphones in it, so I forgot about it. It's probably just a bad component out of the factory.

I think they are there to stop any EMI being picked up from the headphones getting to the mainboard, because it deals with very low latency communication with the PC, so any EMI will result in glitches.

[SeanB]

Ferrite bead on a small piece of wire, 470pf SMD capacitor at each end to ground and you have the substitute. You can even use just a SMD ferrite bead in place and it will work. Most likely that was placed there more for passing EMI testing than for protection.

[ConKbot]

Or a bit more simply, if you have your headphone cord acting as an antenna, if any RF/EMI isnt filtered, then when it gets to a non-linear point (I.e. a semiconductor device) it can act as a simple detector, and then you end up hearing the modulation of the RF, instead of the RF itself.

[Bardo]

Hello guys from the past. I know there's a chance this post never reach you, though it might help someone in the future. Im from Argentina so my english will have flaws, ill make my best effort to be as clear as possible.

I had the same issue that opened this topic. My right headphone output went dead, not all of a sudden. But along of what a think was a month, maybe more, maybe less, can't recall since it was 3 years ago. It wasn't until last night that armed with fierce determination to end this nightmare and a Fluke graphical multimeter that found this topic and was able to find a satisfying solution.

Measuring the ends of the so called functioning "RFI filter" with differents functions of the multimeter i was able to determine that:

-There's no capacitor placed in series with the signal, due to the fact there's continuity (less than 0.5 ohm) between the longest terminals.

-There's a chance that the longest terminal correspond to an small resistance (or inductor) in parallel with a small capacitor since the Lissajous figure was almost vertical until i used the highest testing frequency available on the device (18.75kHz), where it showed a little slope to the right cuadrant (clockwise) and an ovaloid figure. Im not sure of my interpretation of the graph.

-There's for sure a ground-referenced capacitor between a long terminal and a short one (any) but a couldn't get a good reading of the value. I can tell it is 1 nF tops. I suspect much less.

The comparison between the good filter and the dead one told me that the ground capacitor was still operating (in the dead one of course), however no continuity remained. For this reason audio couldn't get through.

I ended up leaving the half working filter in place while using a small wire to let the signal get to the output. I find no audible difference between left and right output. I dont know when ill be able to perform a frequency response measurement or else, im using the interface all the time.

Thanks for your thoughts, they helped me to figure this out way faster.

Hope this helps anyone too. Cheers up!