On Sunday, January 2, 2022 at 3:51:27 PM UTC-5, Trevor Wilson wrote:
Well, it's been a few years. My 1700B has decided to cease operation, so
it really is time for a repair and full rebuild. Here's what I am planning:
* Replace all electrolytic caps in the unit. One of the main filter caps
is dead, so it's time to do the lot.
Probably not a bad idea.
* Replace all the carbon composition resistors. A quick check on half a
dozen has shown a 15 ~ 25% upward drift. Why did ST use CC resistors?
I've been using cracked carbon since the 1960s.
Do remember that the ST1700B is a product of the late '60s and
early '70s to begin with. A LOT of these resistors are used as the frequency-determining elements on the oscillator, so if they drift,
one result is frequencies are going to be off. If that's the only problem,
it's not a killer. I don't recall the eaxt topology but it may require replacing two resistors per frequency.
And this is where you might run into problems. ST talked about how
components were carefully "hand selected" for optimum operation.
It may or may not be the case that those resistors that seem to have
drifte upwards by 15~20% may have been that way the day the
unit left the factory, so proceed with caution.
* I was going to replace the OP amps with LME49710 OP amps, but, well,
you know. NLA. AD797s look like the best, available option. Any other
suggestions?
Unless they have measurable gone noisy or otherwise are known to be
bad, this is probably the LAST thing I would do, simply because the
amount of feedback around them tends to make differences less relevant.
* Replace all the opto couplers.
I'd be VERY careful on this onje. Unless, as above, you can show they are a problem AND if so, you can find suitable replacements (and their specs
are likely to be very specific, I'd leave them alone. Typically, in things like the ST, they are used as part of the quadrature detectors and bridge balance circuits, so unless they have similar transfer functions, newer, "better" ones may, in fact, just not work.
Overall, my strategy would be to replace the obvious bad things first, put the unit back together, test it thoroughly and then only proceed if there are still issues.
Also, note that ST was very specific NOT to use vacuum desolderers (e.g. "Soldapullit") on the unit because of their tendency to spray debris around will-nilly. They recommend using capillary desoldering (solder wick or similar).
The ST1700B was, for its time, a truly qwonderful, almost miraculously good unit. For about 5 years, I used one every day, measuring probably thousands
of amplifiers, preamps and so on. I had a lab where I built a very high-quality passive RIAA pre-compensator so I could measure phone preamps with it,
I also had a switchable 50/75 uS preemphasis filter so I could use it with
my ST 1000 FM generator and do all sorts of measurements of FM tuners,
and so on.
A while ago, I acquired an Audio Sciences 32/44/48/88/96 kS/s 24 bit professional broadcast sound card and got a licensed copy of the ARTA measurement suite from artalabs.hr. It will do everything an ST1700B will
do, with THD residuals at least as low as the BEST 1700B, and will do
MUCH more (e.g., IHF IM, multi-tone IM, frequency response, FFT, RTA,
MLS, ...), and it will do it faster and more reliably. My ONLY complaint
about it is that it's Windows based, and Windows sucks.
Since then, I have been offered a couple of ST1700's and a ST1710,
one for small money, the other two free but for shipping. I turned
them down.
They were great units, 50 years ago...
+-----------------------------------+
Dick Pierce
Diverse Pursuits
Technical Engineering/Development
cartchunk.org
Boston - Spruce Head
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)