Stirring the pot on BBD testing...

[David]

(Ladies and) Gentlemen:

The keyboard player I work with at church told me yesterday that he has a DOD analog delay which one of his bands used in the 80's for vocals.  He said he would find it and let me use it(!)  If everything works properly on it, I am going to offer to buy it from him.  I should get it at a reasonable price.  That would be cool indeed!

In preparation for this possible acquisition, I did some digging around here for some info on DOD delays.  I don't know his model number, but he mentioned it was a rackmount unit.  Based on the info here, I'm thinking the thing uses a SAD-4096.  After some further trolling here, I uncovered the fact that the "MXR Time Delay" is regarded as a kind of "diamond in the rough" because it sounds somehow less than ideal, but is equipped with the platinum SAD-4096.  Much was made of its low cost.  It seemed like it might be a good idea to snag one as a future BBD donor.  Well...

I may have gotten my cart ahead of my horses, but I just happened to prowl around on E-bay last night on the off chance that someone might be selling a MXR Time Delay.  Someone was.  The price was good.  Unfortunately, someone had already bid on it.  To make a long story short, I managed to outbid him and got it cheap.  Of course, the guy is going to ding me on shipping, but c'est la vie.  Anyway, I am a firm believer in caveat emptor, and seller boy mentioned that he wasn't "sure that the pedal is working.  It might need a new delay chip".  I'm a little concerned about that, but it sounds like seller boy is unaware of the esteemed Mr. Hammer's commentary on "broken BBD's" and how the problem is usually somewhere else.

Anyway, I say all that not to make anyone jealous or possibly not make myself SAD (pun definitely intended), but to inquire about the following:

When I put this Time Delay on the operating table, what would be the most effective diagnostic process if it should happen not to work?  I recall that there was an entirely-too-short discussion not long ago about a BBD test rig.  Has anyone come up with anything on one?  Failing that, could a scope and a function generator somehow be pressed into service during the trimpot-twiddling troubles to come?

I appreciate any and all assistance!

[R.G.]

When I put this Time Delay on the operating table, what would be the most effective diagnostic process if it should happen not to work?

Put your scope probe on the input pin of the SAD4096, verify that it's getting audio signal and has the right bias voltage. Probe the clock pins and verify that both are getting full-power-supply clocks and that the clock signals are mirror images (on's off when the other's on) and that they look like good sharp square waves. Then probe the output pin. Is there a stairstepped approximation of the audio signal there?

If all these check out, the chip's good.  If it's not getting audio input, the signal path before the SAD is bad and the SAD may be good. If there's not a good pair of clocks, fix the clock generator, then try again. If there's good signal into the input pin and good bias voltage, and also good clocks but no audio output, the SAD is toast.

[toneman]

a scope would B almost a necessity!!
check 4 clocks.
check for input to the SAD.
check for output from the SAD.
the input check, checks all previous opamp/tranny stages.
output from the SAD checks to see if "Buckets" are working.
there might B a trimpot for bias adj on the BBD output.
B4 U plug it in, eyball it for obvious broken connections,
and/or bent/broken parts on the pcb.
Oh, one more, if U can get the unit 2 test B4 U pay,  then
U might have more "negotiation" ammunition.....
Don't B SAD,  play it!!
good luck.
T

[David]

You guys are GOLD!!!!!!

Thank you!  Maybe I can pull this off!

Oh, one more thing:  what is the protocol for doing a "BBD-ectomy" if the 4096 is not socketed?

[Mark Hammer]

If the chip is not socketed, it will still be possible to remove it without any risk of damage IF:

a) It's not sitting in a double-sided board (these are a real bugger to unsolder - I have my fingers crossed for you)

b) You make a point of using a grounded soldering pencil/iron

c) You take at least some steps to heat sink the chip itself; what I've done in past is just bunched up some toilet paper with cold water on it and pressed it up against the chip to absorb heat

d) You use solder wick instead of a solder sucker; I absolutely *love* my solder sucker, but there are some jobs that solder wick just does better, and removing the solder around an IC pin is one of them.  If you have some liquid flux, applying some to the solder wick before you press it onto the IC pin can also speed up desoldering time and heat exposure considerably.

Incidentally, the SAD-4096 is a decent chip, but nothing special in comparison to the MN3005 (i.e., no better or worse). The SAD1024 is often found in better flangers, but that difference stems from how easy it is to clock that chip really fast.  In the case of analog delay lines, the intent is to clock slower, rather than faster, so there is not big advantage to the Reticon vs Matsushita chip.

[David]

If the chip is not socketed, it will still be possible to remove it without any risk of damage IF:

a) It's not sitting in a double-sided board (these are a real bugger to unsolder - I have my fingers crossed for you)

Me too.

b) You make a point of using a grounded soldering pencil/iron.

Yup.

c) You take at least some steps to heat sink the chip itself; what I've done in past is just bunched up some toilet paper with cold water on it and pressed it up against the chip to absorb heat.

Can do.

d) You use solder wick instead of a solder sucker; I absolutely *love* my solder sucker, but there are some jobs that solder wick just does better, and removing the solder around an IC pin is one of them.  If you have some liquid flux, applying some to the solder wick before you press it onto the IC pin can also speed up desoldering time and heat exposure considerably.

Got the copper braid.  Is that what you meant?

Incidentally, the SAD-4096 is a decent chip, but nothing special in comparison to the MN3005 (i.e., no better or worse). The SAD1024 is often found in better flangers, but that difference stems from how easy it is to clock that chip really fast.  In the case of analog delay lines, the intent is to clock slower, rather than faster, so there is not big advantage to the Reticon vs Matsushita chip.

I'm only guessing about the DOD having the 4096 chip.  Since I don't know what model he has, I can't do much background research.  I do know, however, that Anderton's chorus / flange / ADT / delay unit hosted on your site DOES use the 4096, so I always have that option in my back pocket.