Testing for very intermittent issues

[pinebox]

I was recently called to a studio to find a quiet pop/tick that happens only every hour or so, sometimes once a day. I would try different things, then have to wait around and see if the pop recurred and this is extremely time intensive and impracticable. I have a Fluke 8808A that the manual says will hold the value it changes, but the change has to last a certain amount of time which is unspecified. I need something to monitor the voltage rails, as well as the logic rails and find out if it is from power or a logic spike that causes a status change for a fraction of a second or something. It would also be nice to monitor the power coming into the building. Would a digital scope have a function like this? Anyone have any advice for best tools or approaches to do this sort of thing?

 

[sodderboy]

Do they have a recorded example of it? Is there any idea where it is, like mix bus or a send?
I am working on a (search "data logging dc voltage") logger that I can use to certify repaired power supplies. Keep all voltages under load and log for 24 hours. I like Omega products for everything industrial. There are other meters that will hold a highest/lowest reading, but the weakness in all loggers is the sample rate. 4 Hz might be cool for me, but to catch a click? Also not cheap for a one-time fix.
In the old days we would keep a DAT running on whatever, monitor out R, mix out L, etc., document the sound, get it on tape, and try to narrow-down the source. Of course it is "on the monitors" at first. Now you can have an old laptop sit there and record, or even film an oscilloscope screen.
The seldom, random clicks you describe was almost always static for me. Very dry air with electronics and HVAC in a space. Is there proper humidity in the room? I have patients keep a humidistat for measuring, 40-50% is good especially if a piano is present. A humble humidifier was the solution to most random clicks.
Mike

 

[kags]

Yea QuickTime on a Mac will record - audio, screen capture, or camera- for 24 hours or until drive is full. Maybe run REW or ? and and screen capture. The hard part will be locating the time of the incident if nobody witnessed it for a time stamp.

I’ve used QT in the past to record audio for 12 hour periods to log an intermittent noise.

 

[JohnRoberts]

Sounds like some kind of interference. Look for some time correlation with some external device/event.

JR

 

[Voyager10]

If you can leave a laptop & audio interface lying around in the studio, I'd start by recording just some noise from an open input (e.g. plug in a length of mic cable, nothing else). This will pick up environmental noise (from EM interference or mains wiring) - if you pick up the click/pop independently it will eliminate faults in the console/monitor/outboard electronics themselves.

I'd use Audacity (Audacity ® | Free Audio editor, recorder, music making and more!) to do the recording. You can then normalise the audio and use the "Sound Finder" function (on the Analyse menu) to label areas where the sound is over a given threshold.

This was me turning the phantom power on & off on my audio interface, the '1' and '2' at the bottom were added by Sound Finder:

 

[mrclunk]

I was recently called to a studio to find a quiet pop/tick that happens only every hour or so, sometimes once a day. I would try different things, then have to wait around and see if the pop recurred and this is extremely time intensive and impracticable. I have a Fluke 8808A that the manual says will hold the value it changes, but the change has to last a certain amount of time which is unspecified. I need something to monitor the voltage rails, as well as the logic rails and find out if it is from power or a logic spike that causes a status change for a fraction of a second or something. It would also be nice to monitor the power coming into the building. Would a digital scope have a function like this? Anyone have any advice for best tools or approaches to do this sort of thing?

I've had a few of these here over the years.
Some were inductive load(?) power spikes causes by HVAC fans turning on or off. Adding a big snubber cap got rid of them. Also fridge compressors a can cause it too.
A trickier one was a random full scale pop across all channel inputs. Turned out to be a dodgy ribbon cable connector linking the digital to analog boards on SSL alpha link interface. SSL service pointed me in the direction of that problem.

The seldom, random clicks you describe was almost always static for me.

Can someone explain how that works? I've heard people talk about it but never understood...

 

[ruffrecords]

I've had a few of these here over the years.
Some were inductive load(?) power spikes causes by HVAC fans turning on or off. Adding a big snubber cap got rid of them. Also fridge compressors a can cause it too.
A trickier one was a random full scale pop across all channel inputs. Turned out to be a dodgy ribbon cable connector linking the digital to analog boards on SSL alpha link interface. SSL service pointed me in the direction of that problem.


Can someone explain how that works? I've heard people talk about it but never understood...

I cannot speak to recording studios but, back last century, I was involved in solving a similar problem for a well know chain of petrol stations in the UK. It was wintertime and occasionally the entire POS system would crash. Apart from the down time while all the tills and pumps rebooted, they lost their totals for the say. This meant they not only did not kn ow how much money they had take but, more importantly they did not know how much fuel was left in each tank. Big headache.

To cut a long story short, it was wintertime, the kiosks were cold places and the tills were operated almost exclusively by women. Often they would have a fan heater beneath their till to keep them warm. It turned out warm dry nylon tights rubbed together often enough, combined with poor microprocessor decoupling and software not designed to cope with exceptions meant the equipment was very susceptible to small electrostatic discharges.

Cheers

Ian

 

[pinebox]

Very interesting responses. I did not get notification of replies here or I would have responded sooner. I am seeming to get them on all the group outputs. Turning off buckets, removing channels, turning off the center section, all this has not been fruitful. We have replaced the PSU, disconnected all the patchbay (besides to Pro Tools) changed outlets for the converters. I think it may be logic switching related with the outputs, because I have not gotten the same noises/pops from the insert sends. Here are some photos of the noises.