[ARC5] 70+ year old errors found and corrected.

[J. Forster]

First, the red inspection marking paint is called Glyptal.

It is possible that things like that slip through inspection, obviously.
It is possible the radio even worked during final acceptance, because
there was a tiny break in the insulation where it wrapped around the
terminal.

I've seen at least four instances of assembly errors.

First, I bought a bad HP Communications Sweep Oscillator cheap. No output.
After a lot of work, I concluded the final amplifier module was bad.
Having nothing to lose, I took the cover off it. Inside there was a #2-56
Allen screw rattling around. It was to hold down an MMIC inside the amp,
but it had quite obviously never been tightened. I contacted HP- nothing.

For a while, I repaired Data General mninicomputers. A customer brought me
a Nova 3 memory card that was intermittant. In those days, a 32kWord
memory card was a big deal, costing well over $5000. After looking at it,
I noticed an entire row of DIPS had never been soldered. I soldered them,
and it ran perfectly!

Third, I bought a brassboard for a prototype of the AMRAAM stand-off
missile. This was from a contractor who did not get the job. While
stripping the thing, I found several wires, wrapped around switch
terminals, that had never been soldered. For all I know, those wires could
have caused the thing to fail, losing the BIG contract.

Fourth, I bought an HP LASER Interferometer System, very cheap on eBay. It
worked for a bit, then quit and would not lock up. These are very, very
expensive- think a luxury car. There was no way I could afford to have HP
fix it, so, thinking it was a doorstop and having nothing to lose, I
opened it up. A few minutes with a hand lens, showed a photo-diode in the
optical feedback loop had been inserted through the PCB holes and the
leads bent over. The leads were never trimmed or soldered. Two minutes
with dykes and a soldering iron, and all is well.

Early on, when I was working on spacecraft payloads, a guy told me:"You
CANNOT inspect in quality"

Words for the ages.

-John

=====================





> You know that 1.5 - 3.0 MHz coil box I was having troubles with? The one
> out of the receiver
> that had been hacked for crystal control back in the 1970s?
>
> Well, today, not having anything much else to do (yeah, right), I decided
> to take a closer look
> at it to see if I could figure out exactly what was wrong with it.
>
> First, since the HFO coil had been the one that had been messed with (I
> figured this out
> when I looked at the screws holding the coil in the box: there were 2
> different kinds, and the
> paint had been "broken").
>
> Although I had restored the two resistors, R-3 and R-6 to their original
> values, the HFO still
> didn't work. I looked more closely at the connections to the coils and
> after taking my glasses
> off and looking VERY closely, I discovered that one lead of the plate coil
> was no longer
> soldered to its connection. What looked at first like solder was flux, and
> all 6 of the Litz wires
> were not soldered. When I had tested the coils for continuity, I had
> clip-leaded one lead of the
> ohm-meter to one end of the coil, and had used the probe on the other end.
> The clip-leaded
> end turned out to be the open one....of course.
>
> Have any of you 70+ year old folks ever attempted to 1) clean the enamel
> off of 6 pieces of
> #60 Litz wire, and then 2) solder them in place with shaky hands? Well, I
> managed...finally.
>
> Putting it all back together and installing in my working receiver, I
> found that the HFO was
> now working, but the signal through the mixer was way down. However, the
> RF amp did peak
> by the Align Input control, so that meant that the RF amp coil was OK.
>
> Back to the coil box. After swapping the good coils, one at a time from my
> working coil box
> into the non-working coil box, I zeroed in on the mixer coil as the
> culprit. Obviously, however,
> the coil had never been tampered with as the screws still had their red
> paint on them and
> they were very tight.
>
> Even so, I removed them with some effort, and removed the coil. Using my
> clip-lead and
> ohm-meter I again determined that both coils had continuity. Furthermore,
> there were drops
> of the original ruby-red enamel on all the connections.
>
> Then, remembering my experience with the HFO coil, I used both of the
> ohm-meter probes,
> without any clip-lead, to check continuity through the connections at the
> bottom of the coil
> from outside it.
>
> Sure enough: one coil no longer had continuity.
>
> Again removing my glasses, I looked VERY closely at the two connections of
> the now open
> coil, both of which looked "factory" and both of which had that ruby-red
> enamel on them.
>
> One of those connections had never been soldered: the Litz wire had been
> wrapped around
> the connection, never soldered, and then had been covered with ruby-red
> quality-control
> enamel.
>
> A 70+ year old construction error, followed by a quality-control error
> uncovered.
>
> A few seconds with the soldering iron and solder fixed the problem.
>
> Re-installing the now repaired coil box proved that the receiver was now
> working properly.
>
> Alignment followed, and I am now listening to it.
>
> Sheesh!
>
> Ken W7EKB
> ______________________________________________________________
> ARC5 mailing list
> Home: http://mailman.qth.net/mailman/listinfo/arc5
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:ARC5 at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>