dbx VCA chips and sockets?

[Ike Zimbel]

Hi All,
I have a vague recollection of dbx (or maybe THAT) NOT recommending sockets be used for their SIL package VCA chips, but so far I haven't come across that note in any of my data sheets. Anyone recall that? Reason I ask, I have two 160-A's on the bench for service. One of them is giving me distortion measurements in the 0.0x% range and the other one is reading in the 0.x0% range, along with a few dB higher noise floor. The main difference between the two units is that the one with higher distortion has sockets for the VCA IC's and the other one doesn't (they are clearly different builds). I have tried a new dbx 2150A as well as a THAT 2180B and neither changed the distortion and noise figures. I'm at the point where I can't think of anything else to test short of removing the socket for the VCA in the signal path and soldering the VCA directly to the board. Any insights will be much appreciated.
One more thing, the distortion trim, V7, is working, both with the original IC and the two replacements that I tried.
Thanks!

 

[JohnRoberts]

if it's in a socket does removing and reinserting it, or wiggling it make a difference?

Sockets get dodgy over time, but percussive troubleshooting might reveal a weak connection.

JR

 

[Ike Zimbel]

Hi John,
So far, that doesn't appear to be the case. I have also cleaned the pins on the original IC to no effect.
Thanks!
IZ

 

[CJ]

you might try re-flowing the solder connections for the socket

sometimes they solder mask those holes during wave soldering and then solder in the sockets later. this keeps the light weight sockets from boiling out of the holes,

it is fun to watch wave soldering, all those components dancing around like a skeleton on a hot tin roof while they move across the solder tank, there are usually a lot of components at the bottom of the tank when they clean that 400 pound monster out

you might even want to pull the sockets and see if there is any white goop in there, that mask is very heat resistant for obvious reasons, and it will deny a good solder joint as long as it is in the through hole.

if we are talkin surface mount then simply ignore all of the above.

 

[Ike Zimbel]

Thanks for the suggestion. De-soldering the socket is definitely my next move. I will report back.

 

[Ike Zimbel]

...and that did 42 different kinds of nothing . Current status: The original VCA is soldered in place, no change to the THD and N readings.
And, while I had it out, something went squiffy in the LED meter board so now I have to look into that, too.

 

[Matt Syson]

There are a number of quite subtle comments in the application notes for DBX (THAT) chips which some manufacturers of equipment seem to miss. Routing of circuit traces and component placement being pretty vital although if the 2 units of the OP are the same copper layout exactly that may not be the issue. Noisy control port voltages, either noisy op amps or crosstalk picked up 'anywhere basically' can be detrimental. The control ports are as sensitive as the main 'forward' audio port so need careful consideration. As swapping the VCAs has been tried with no reported change it suggests that the VCA itself isn't the root of the problem. Swapping the VCA between units cOULD be a posibility BUT the act of soldering and desoldering has a fair chance of disturbing the VCA distortion characteristics just by doing it. Not suggesting the soldering capability is poor by any means but any stress on the chip can alter things. Is the distortion trim voltage actually 'clean'? is the 'no signal' noise level exactly the same between units? Have you looked on a 'scope' at what the distortion actually 'looks' like between the 2 units? Distortion is of course related to the signal level and amount of attenuation. All in all a whole bunch of variables that could be going on.