Audio Test Equipment - Computer Control

[davada]

Thanks, yes I agree on your reasoning there.
That is an impressive design, do you know which relay they used?
The capacitance neutralization is interesting, not sure why they needed
another non-inverting OP amp for the guard. They drive it with 150 ohms
perhaps that's why, but I'd think that 1K would work.

I should probably just buy an 1121.

What do you consider SOTA in sine wave generators, have a schematic?

"not sure why they needed
another non-inverting OP amp for the guard."

To avoid board parasitic capacitance loading. Also why the 150 ohm series resistance on the driver.

 

[Audio1Man]

Hi Pete
I worked for Audio Precision and retaired 2005 and returned to Southern California. I repair / sevice AP System that are not supported. I have a serviced a few hundred system during the last 10+ years. S/n 10002 (1984)through s/n 45xxx.
Many are still in production tests. The RELAYS are the least of the problems. The relays are NEC MR602-5SR for the most part. The ones with most usage are in the ANALYZER fontend (3 each). I have replaced only a handfull in my repairs. The relays with the most failure are in the GENERATOR outputs (7 each). They fail because some one connects the power amplifier output into the generator. The output attenuator resistors and relays burn up (This is a big problem, System Two's we installed a protection circuit to reduce this problem).
We use FET'S in many circuits, however many circuits need relays for best preformance.
You & I are on LinkedIn, join "Classic AP Audio Systems LA" or look @ SOUND LOGIC LA ...
I hope this may help in your effort to construct your test system.
Duke duke.aguiar@ieee.org

 

[1audio]

HI Dale,

The Shibasoku 725 series analyzers are entirely relay switched, three notch filters and three gain stages. The two units I have here date back to the 80s and early 90s. Both are working fine.

I did lose one relay on my 725. One of the 1 KHz tuning relays. I found a spare in china. These are shielded reed relays, almost the mot expensive type. Mine is a first gen 725 that dates from the late 1970's or early 1980's.

A QA401 would be the easiest and quickest way to get up to speed servicing and even designing audio products. Its simple, comes with decent software and has a distortion floor low enough to know your device is working properly.

An AP is great and very versatile but more expensive and more of a commitment to get running.

 

[BlacK_Chicken]

@1audio: Do you really think a QA401 is worth a try? At the Moment I am fed up with unpredictable soundcard-based measurements and I am looking for professional audio analyzers with the capability to perform automated sweeps/measurements via software. My personal favorite, a dScope, is much much more expensive than a QA401 whilst the devices are not so far away from each other. I am not sure, if the better analog frontend and the sophisticated software-interface justifies 20x the price. BTW I have already some practical experience with the dscope, love it.

Better analog frontends with high voltage capability (this is actually what i am looking for) you get with the "old" stuff like the HP and the System 1, but the software-interface kind-of "lacks"

To be honest, I am absolutely lost ^^

 

[1audio]

I still use soundcard based measurements mostly, however thats because I'm measuring headphones and speakers mostly. The right soundcard + software + OS can be quite stable. Change anything and you will need to check everything like its a new install. My recent microphone measurements too a full day with 1/2 of it troubleshooting incorrect connections.

For what you are describing a QA401 may be a good choice. There is a learning curve and none of the PC based stuff gives the instant feel of an analog analyzer but you can get a lot of info quickly and run complex tests. The QA401 software is good, but has some limitations you will need to learn how to work around. You can write scripts for it (please share if you do). For an analog front end you can make interface cables and probes to address whatever you are trying to measure. High voltages like a tube circuit can be done with care and an X100 probe. High power amps with an X10 divider integrated into a load resistor bank.

Audio1man could support you in an older AP but its a bigger commitment (and 30X the volume). There is lots more support and automation already available.

Possibly the most important aspect is to understand what you are doing well. Its really easy to get this stuff to lie to you. Check everything when any measurement looks odd. When the system gets more complex you have more places to make mistakes.

 

[BlacK_Chicken]

Thanks a lot! From my own experience I can confirm every single point you made. Any measurement system (audio or others) has a lot of pitfalls. Things get worse the more complicated they are. Even a up-to-date 4ch scope, like a leCroy HRO, can deliver dubious results if the operator does not know what he is doing.

For Speaker related stuff, I use a soundcard too. Thats fine, measuring acoustic systems I do not care about absolutely correct levels. Its more a relative thing. But cable conducted stuff is much more demanding in terms of acurate and predictable readings.

I checked the QA401 software, I like the fast and responsive interface but its indeed very limited in functionality. Looks pretty straight forward compared to others. If the scripting stuff works (the docu is really bad, next to not existing) it may be a powerful tool.

Actually, has anyone ever tried making phase-measurements with the second channel of the QA401? (In my opinion a frequency response consists of amplitude and phase behaviour.)

 

[PB2]

Hi Pete
I worked for Audio Precision and retaired 2005 and returned to Southern California. I repair / sevice AP System that are not supported. I have a serviced a few hundred system during the last 10+ years. S/n 10002 (1984)through s/n 45xxx.
Many are still in production tests. The RELAYS are the least of the problems. The relays are NEC MR602-5SR for the most part. The ones with most usage are in the ANALYZER fontend (3 each). I have replaced only a handfull in my repairs. The relays with the most failure are in the GENERATOR outputs (7 each). They fail because some one connects the power amplifier output into the generator. The output attenuator resistors and relays burn up (This is a big problem, System Two's we installed a protection circuit to reduce this problem).
We use FET'S in many circuits, however many circuits need relays for best preformance.
You & I are on LinkedIn, join "Classic AP Audio Systems LA" or look @ SOUND LOGIC LA ...
I hope this may help in your effort to construct your test system.
Duke duke.aguiar@ieee.org

Hi Duke, thanks for the input, we are already connected on Linkedin, nice to see you here also.

And, thanks also to everyone else, I need to digest all the new input.

 

[mark_alex]

Hi Pete
I worked for Audio Precision and retaired 2005 and returned to Southern California. I repair / sevice AP System that are not supported. I have a serviced a few hundred system during the last 10+ years. S/n 10002 (1984)through s/n 45xxx.
Many are still in production tests. The RELAYS are the least of the problems. The relays are NEC MR602-5SR for the most part. The ones with most usage are in the ANALYZER fontend (3 each). I have replaced only a handfull in my repairs. The relays with the most failure are in the GENERATOR outputs (7 each). They fail because some one connects the power amplifier output into the generator. The output attenuator resistors and relays burn up (This is a big problem, System Two's we installed a protection circuit to reduce this problem).
We use FET'S in many circuits, however many circuits need relays for best preformance.
You & I are on LinkedIn, join "Classic AP Audio Systems LA" or look @ SOUND LOGIC LA ...
I hope this may help in your effort to construct your test system.
Duke duke.aguiar@ieee.org

Duke,

I have a 1994 322G model (S1 dual domain unit) S/N 34322 (in good operational and cosmetic condition). Purchased used and fully calibrated from Marshall Buck in 2008, sent back for cal in 2014 and received back all perf checks passing.

Unit now shows high distortion in the A analyzer channel on the 12dB attenuation setting (5V and 10V manual input ranges). It is almost 20dB greater than the B channel set to the same 5 & 10V ranges. Analog generator A and B outputs are fine - no problems there (both show low distortion on B channel input). Gen-Mon shows same behavior as the XLR input so it is not the input connectors or external cabling.

The 0dB atten setting on the A channel exhibits very low distortion and matches the B channel exactly (same goes for the 24dB atten settings for 20V and 40V input ranges). I have eliminated the preamp 6dB gain setting as a possible source of THD problems, since it is also used with a 0dB atten setting for the ranges below 2.5V, and there is no excess distortion when selected.

Also, in an attempt to troubleshoot, I have hooked my scope to the A & B monitor outputs to make sure that both the A & B attens are actually attenuating by the correct amount when selected (they are). So it does not appear to be faulty divider resistors.

Things now appear to point to a faulty channel A 12dB attenuator relay (K572 P/N 4530.0002.1 in the schematic). Any advice on further troubleshooting before I pull the LVF-2 analyzer board would be most helpful. I have seen relay contacts go bad over the years (and my S1 is now 23 years old).

Noticed that the K572 part # is different from other relays - what does the ".1" designation suffix indicate? Also idea where I can obtain a replacement NEC MR602-5SR relay (or a sub that matches the P/N)? I noticed that the original red color NEC types are not generally available any longer.

Any help/advice would be gratefully appreciated!

Thanks, Mark Alexander

 

[Audio1Man]

Hi Mark
Email me duke.aguiar@ieee.org and I can send you a file to run and return the data file back to me.
You should join the LinkedIn "Classic AP Audio Systems LA" and post problems like this. I can give you more info after I get your return data
Duke

 

[mark_alex]

Hi Mark
Email me duke.aguiar@ieee.org and I can send you a file to run and return the data file back to me.
You should join the LinkedIn "Classic AP Audio Systems LA" and post problems like this. I can give you more info after I get your return data
Duke

Hi Duke,

Done - joined your Linked-In group. Will send you an email later in the evening...

I suspect my issues could be related to the low noise analyzer section mods that Marshall Buck did to my S1 back in 2014, and not the K572 relay.

Ch B has AD797 x2 in the DIFF pre-amp, but Ch A has an SOIC op-amp on an adapter, ADA4898-1 (2 each). Clearly the input pre-amplifiers in each channel are different (I think there were some stability issues with the AD797 on the CH A input section, bit not on the B channel).

I see that in the 12dB attenuator setting, the DIFF pre-amp op-amps see the maximum input source impedance (~20k ohms). So, the THD signature differences at 12dB atten could well be due to the differences in the non-linear input bias currents of the op-amps in each channel. These, of course, would cause a small (but measurable) non-linear voltage drop across the effective source resistance thus producing a possibly higher level of THD. This is conjecture on my part, and I would have to swap op-amps on the LVF-2 board to verify this (maybe put 2x 5534's back in to test).


Thoughts?

Best regards, Mark A.

 

[Audio1Man]

Hi Mark

Yes I have known Marshall for 40+ years. I do fix Marshall AP dogs. The LOW NOISE MODS using AD797 have different compensation. Don't try to plug in NE5534 unless you recomp the circuit.

Call me as we need to run some specific tests to find what is the real problem.
The LinkedIn old "Classic AP Audio Systems Group" I believe had my posted LOW NOISE MODS. Don't use the old group as it is no longer moderated after Jim's passing. Use LinkedIn "Classic AP Audio Systems Group LA" to post any AP questions. Duke