ART DTI Box not as roadworthy as I'd like...

[BigRedX]

About 5 years ago when I put together my current live rack, after reading various posts on here about isolating my gear from any nasties from the PA and possible ground loops I invested in an ART DTI Box for the main outputs of the audio interface.

For the last 18 months I have been suffering from various intermittent problems on the LH channel manifesting as either buzzing or distortion on the signal. Each time after it had occurred I would take the rack apart and try and see where the problem was, test all the leads without finding anything and put it back together for it to work perfectly again for a month or two before the problem returned.

Finally a few months ago I thought I had tracked it down to a dodgy cable on one of my home-made TRS jack to XLR leads inside the rack (flexing the one for the problem channel would cause pops and crackles) so I replaced both of them with new good quality ready made ones. This appeared to have fixed everything until my most recent gig where it returned worse than before.

I have another gig tomorrow so yesterday the whole rack was taken apart one more and each item was tested individually without finding anything amiss. Put it all back together and the problem is now worse with a constant low level hum on the LH channel and clicking manifesting when the track was playing.

Listened to the signal going back to the headphones socket on the Mac (fine) the headphone socket on the Focusrite interface (also fine) the channel outputs of the interface going direct to a headphone amp (still fine). Discovered that problems occurred once again when the DTI Box was introduced into the signal path. Again the symptoms could be "cured" by flexing the LH channel lead going into the DTI Box. Finally with everything except the DTI Box connected it was fine and no amount of cable manipulation was introducing any problems. Therefore it could only be the DTI Box.

Getting it out of the rack and plugging in cables where I could see what was happening it was obvious that there was movement of the of the LH input XLR socket on the DTI Box and that was causing the problem. Unscrewed the input side cover and found that I was able to move the socket a lot more while the RH socket seemed firmly anchored to the PCB. I also discovered that in order to remove the PCB so I could inspect the track and solder joints would involve removing the output side cover and earthing tag which required a spanner I don't have. I put everything back together and somehow I have managed to stop the LH socket from moving and it all appears to be fine for now and also all the sound gremlins have gone.

As soon as I can I will have the box apart and check the tracks and solder joints around this socket. However I am wondering how this problem could have occurred in the first place. It's not as though the device is chucked into my bag of bits that come to each gig and then kicked about on stage. The DTI Box has lived permanently in this rack from new where it is surrounded by flight case foam to hold it firmly in position. The leads going to and from it are cable tied into place and under normal circumstances there is no access to the area of the rack where the device sits.

Other than to check (and repair if necessary) the PCB is there anything else I should be looking for when I have the DTI box apart? At least I now know for sure where the problem is and if necessary I can take the back rack lid off to gain direct access to the interface connections before the DTI box if I have problems at tomorrow's gig.

[James Perrett]

Sounds like it is probably dodgy soldering on the PCB. When you have the PCB out, I would suggest reflowing all the joints just to be sure - assuming that you are reasonably competent at soldering.

[BigRedX]

That sounds like a plan. A bit miffed that it's taken me so long to track down, but I was working on the assumption that it must be something that I had soldered up that was causing the problem and not a ready made piece of equipment. Just spent all of this morning listening to the backing track playback and everything seems to be fine for now.

[Hugh Robjohns]

Hmmm. For what its worth, I have four DTI boxes, two of which are more than 5 years old, and none have given the slightest trouble.

I also have a T8 unit (four DTI boxes in a rack mount format for 8 channels), which I use to balance a bunch of synths. That's been permanently connected in the back of a rack for five years and has caused no issues either.

James' idea of reflowing the solder joints around the wobbly socket is a good one — movement could have fractured the solder joint. But I'd also investigate why its moving at all. It should be secured solidly to the end plate.

Another possibility is tarnishing of the connector pins. Some deoxit or similar would be a good idea. I try and use it on permanent connections about once a year.

[BigRedX]

James' idea of reflowing the solder joints around the wobbly socket is a good one — movement could have fractured the solder joint. But I'd also investigate why its moving at all. It should be secured solidly to the end plate.

I have a feeling that the screw that attaches the end plate to the casing next to LH XLR input socket wasn't quite as tight as it could have been, as it was a lot easier to unscrew than the others. There was definite movement of the socket when assembled. Interestingly having taken the input side end plate off and reattached it the socket no longer moves, and I don't have any audibly dodgy connections, so that will do me as a temporary fix until I have the right tools to get the other end plate off and the PCB out to check the tracks and solder joints.

The way the DTI Box is put together it shouldn't have been possible for the socket to move at all, but it could, albeit very slightly. I'll get some Deoxit on everything too.

[Rob Kirkwood]

I have a DTI you could borrow if you like? (I'm in Nottingham).

It's currently sitting on the aux input to our hifi, think it's fully working (and wobble-free) - but I'll check it thoroughly if you want to borrow.

Rob

[BigRedX]

I have a DTI you could borrow if you like? (I'm in Nottingham).

It's currently sitting on the aux input to our hifi, think it's fully working (and wobble-free) - but I'll check it thoroughly if you want to borrow.

Rob

Thank you. However taking it apart and reassembling appears to have sorted it enough for me to be confident about tomorrow's gig, and if it starts playing up when we set up at the venue I know what is happening and will either by-pass it, or disconnect the offending LH channel. We don't have another gig or rehearsal the following week which will give me plenty of time to take it apart completely and look at the solder connections and PCB tracks.

[Rob Kirkwood]

OK, no problem - hope the gig goes well!

Rob

[tacitus]

Mine’s been in irregular use for over a decade with no issues.

Reading this thread reminded me of a friend, only about 4 years older than me, who used to maintain that cleaning jack plugs with brown paper was the thing. Haven’t heard that mentioned for at least 30 years …

[shufflebeat]

With (if I’m understanding correctly) both the XLR and the plate retaining screw showing signs of trouble, I’m wondering if the XLR hasn’t taken a sneaky knock by someone who might have scarpered without reporting.

[ef37a]

This problem can arise when the PCB is solidly mounted to the case and, as has been suggested, the plug gets a whack. The XLR moves relative to the PCB. Having the PCB "floating" i.e. just secured to the sockets might look like poor design but is in fact the better way. I have had the problem with MIDI DINs and USB ports.

If "touring" kit is taken apart for other reasons a dab of Loktite on such threads is a good idea. In extremis, nail polish will do (but don't be seen nicking it) Keep it well hidden in your tool box lest you get quizzical looks! Few manufacturers go to the vast expense of a star washer these days.

Dave.

[BigRedX]

Following on from this, after 9 months of problem-free running, the intermittent fault re-occurred at last night's rehearsal.

Now that I finally have access to all the correct tools I was able to get the DTI circuit board out of the casing and discovered this:

Unsurprisingly the problem socket is the one connected by the solder "joints" in the top left corner. As this unit was bought new and this is the first time I have had it apart, I am not very impressed. Also weird how all the other solder joints look fine.

If I just remove all the problem solder, clean the socket legs and board and make new joints will this be sufficient?

Tempted to share this image with ART and see what they have to say.

[Drew Stephenson]

If I just remove all the problem solder, clean the socket legs and board and make new joints will this be sufficient?

Tempted to share this image with ART and see what they have to say.

That's what I'd do.

On both fronts.

[Martin Walker]

If I just remove all the problem solder, clean the socket legs and board and make new joints will this be sufficient?

I'd be inclined to solder on a short length of single-core wire between those two circuit traces (where you've scraped away the resist), rather than rely on the solder bridging any gaps.

[Wonks]

I'd be inclined to solder on a short length of single-core wire between those two circuit traces (where you've scraped away the resist), r

It wasn't BigRedX who did that. That was just as he'd found it after opening it up, so it's obviously either not new, or else a fault was discovered during a factory QC inspection and poorly repaired.

[Hugh Robjohns]

I can't see how that broken trace would ever escape even the most rudimentary QC.

I suspect the socket wasn't fitted tightly to the board in the first place and has flexed in use, eventually cracking off the trace. The missing bit is probably floating around in the box. The joint above and top the right looks similarly dodgy, as if it is being pressed away from the PCB.

I can't see anything fundamentally wrong with the design or construction, it's just a poor implementation.

And an easy fix with a jumper wire... but first, desolder, reseat and re-solder the joints on the other XLR pins. The XLR shouldn't be able to wobble...

[BigRedX]

The traces all appear to be fine looking under high magnification, it's just the resist that seems to have been scraped off. I checked in the case and there's nothing floating about. The middle "connection" appears to be the worst as the solder is only partially attached to the circuit board track. Anyway I've removed the old solder and remade all three connections. That particular socket was standing slightly proud of the board. It's not now.

I've run one song while wiggling the leads connecting the DTI Box and it appears to be running fine. I've also re-orientated the box in the rack so the connections are all made at 90° with respect to the back/front so hopefully that will be less likely to be affected by any transit bumps. Going to run the whole set and check that the sound is consistent. We have an important gig tomorrow so it needs to be perfect.

I've also shared the photograph with ART, but I've not had a reply as yet.

[AlecSp]

I'm also of the view that this looks like damage, rather than a manufacturing defect, albeit that the design doesn't look too resilient.

Given that, based on the original description, this had been in operation for 3.5 years without problems, something has introduced the fault, which is most likely to be one or more impacts.

Positively, it looks like a relatively easy fix for a capable solderer.

[BigRedX]

Looks like it's fixed.

When I bought it the DTI Box was supposed to be new, but I can't find an invoice for it and therefore identify retailer I bought it from. I had thought it might have been Thomann who have form for sending out B-stock products and returns as new items, but this was part of a fundamental makeover of my live playback system and while some of the things did come from Thomann the DTI Box is not among them. In fact I have found invoices for everything bought for this revamp apart from the DTI Box.

[ef37a]

There is a slight manufacturing shortcoming in that the jacks are not secured to the front panel by a nut, thus years of "wiggling" can eventually cause the joints to fail. Best practice is to bolt the sockets to the panel then let the PCB "float".

This is a slightly more expensive way to do it and does not suit some forms of construction. Otherwise the box looks of very good build quality and should survive all but the "golly-handed"!

Dave.