On 7/18/24 05:19, Chuck Harris wrote:

All I really need is one tiny little drop, about 0.04mm.

And no flow additives can lower the conductivity when dried/cured.

you've stumped us all as far as keeping it liquid by changing it.

So, what's left is keeping the surrounding very full of the same volatiles in the conductive paste/paint,
or humid with water.

I think a humid atmosphere around the work area is more feasible than anything else dreamed up, so maybe worth testing on non-critical similar parts.

Frank,
You have a serious problem if you can only get the Log/Lin amplitudes to match at a ref level of -70dBm.
Let’s start from scratch. Operators check pages 3-10 through 3-14 of the 8557A manual.
In Log mode, Center Freq (CF) set to 250MHz Ref Level set to set to -30 dBm, Span set to 10 MHz, Resoluton Bandwith to 1MHz (RBW) Ref level fine set to 0 dBm, Cal signal connected to the input, what do you see on the display? If you have a displayed signal, can you gradually reduce the Span and RBW (30KHz/div. and 10KHz RBW) and keep the signal centered on the display? Has the amplitude of the displayed signal changed much as you gradually stepped the span down? What should happen is the displayed cal signal (250MHz @ -30 dBm), should be at the top graticule line, centered on the display, and as you reduce the Span/RBW switch (coupled) from widest to narrowest should stay at the center frequency of 250MHz and stay at the top graticule line whether you are in 10dB Log / 1dBLog / Lin mode. There are tolerances so you might see 1 vertical division change in amplitude as you switch to. 1dB Log or Lin mode. You might have to slightly adjust the frequency the keep the signal centered, however if you see abrupt changes in frequency and amplitude - a couple of graticule major divisions or more that points to another problem - resolution bandwidth adjustments (for centering and bandwidths). Also as you switch to narrower spans/RBW (mainly at the narrower span sans RBW), the displayed signal may pull/stretch to the right side of the display, if so increase the sweep time/div (SWP) until it looks normal - a bell shaped signal.
Don Bitters

This is all getting way too complicated for the task. If I
knew how to take a picture with my scope, I should have, but
I don't, and didn't. For that I am truly sorry.

Let me describe it in terms of the macro world we usually work
in:

It is a dual row PC card edge connector where the rows are
staggered, and the conductors connecting to the pads on the outer
row pass in between the pads on the inner row.

The flaw is in one of the conductors that passes in between two
pads on the inner row.

Shrink that model down to sub mm pad widths...

I can't simply cover the gold plated traces, willy-nilly with any
sort of solder paste and blast it with hot air. It will cease to
be a reliable connector if I do.

I can't leave any coating on the surrounding pads to protect my
repair, they are a connector.

I was once a skilled amateur watchmaker, and did lots of work on
ladies watches. This job would have taxed my much younger, skilled
hands back then. Now, at almost 70, I can tell you, things haven't
improved!

As I said originally in the heading, I need a conductive paint
for micro circuit repair.

I need a paint that will allow me to place a single dot over the
top of a flaw that is about 0.04mm in diameter... about 0.0015"
of an inch, or the thickness of one of my hairs.

Thanks for all the help. I think I have enough new ideas to finish
this job.

-Chuck Harris




On Thu, 18 Jul 2024 05:19:55 -0700 "SCMenasian" <scm@...>
wrote:

What I would try (first on a test sample piece) is:
1) Chemically (or thermally) remove the insulation from a piece of
#46 magnet wire. 2) Paint the wire with some Chipquick (T7 or T6
particle size) solder paste. 3) With some tension on the wire (to
hold it straight), heat it to melt the paste and tin the wire. 4) Cut
the tinned wire to length. 5) Paint the entire area of the defective
flexible part with solder resist. 6) With the tip of a scalpel,
scratch a groove (down to the copper conductor but not through it)
where the repair should be made. 7) Apply solder paste so it fills
the groove. 8) Lay the prepared tinned wire in the groove.
9) With a hot air gun, melt the solder paste to complete the
electrical connection. I might experiment with restricting the size
of the hot air stream by using some teflon spaghetti tubing to make a
tiny nozzle.

Solder would have been my first choice, but this is too
small for my smallest soldering iron, and the substrate
isn't meant for solder temperatures. My smallest iron is
a Weller MT1000, which is meant for soldering surface mount
IC's with about a 1mm pitch.

Everything I do at this level is under an Olympus stereo zoom
scope. Ladies wrist watches are easy compared to this stuff.
(I used ladies wrist watches to train myself for SMD work some
30 years ago.)

I tried using watch oilers, the problem is the viscosity of
watch oils and greases is much higher than this silver paint
that I have been using. No drop is formed when I dip into
the paint. All that happens is I put a thin coat on the
oiler, and it dries while I move from the paint to the flex
board.

All I really need is one tiny little drop, about 0.04mm.

-Chuck Harris


On Wed, 17 Jul 2024 16:30:49 -0700 "Scott Vance via groups.io"
<scott_l_vance@...> wrote:

For flexible connections, I've had the best luck with really thin
wire.? You can use solder on a soldering iron at lowish heat to very
carefully burn off the coating and prep it.? Next cut it to length
and then solder in place (or conductive 2-part silver epoxy or
whatever).

For putting down fine lines, in the watchmaking trade we use small
oilers.? They have a tiny tip which holds just a bit of liquid.
Cousins UK is generally the best source for this sort of thing.? ?If
I was putting down a conductive adhesive and it had to be precise it
would be this under a good scope.? Good luck!

The object of the adjustment is to get no relative change in displayed signal amplitude when you change between Log mode and Lin mode. In the adjustment you are trying to do one adjust affects the DC offset more and the other affects the overall gain of the Log amp. Start the procedure with the fine adjust centered in the adjustment range, then use the coarse adjust to get no shift in amplitude when switching between Log and Lin modes on the display. Then use the overall gain to bring the signal amplitude to the top graticule of the display while decreasing the coarse adjust to 0. Adjust the fine to maintain no relative change when switching between Log mode and Lin mode, if you still run out of fine adjustment range, then note where the coarse adjustment is set (the knob is scaled 0 to -11). If you are off by a small amount on the coarse knob (less than 1dB) then you may have a slight DC offset when switching into Lin mode, if you are off by about -2.5 dB or more on the coarse knob, you have a gain issue from the front end through the log amp section. I have seen this caused by “wounded” (leaky) input mixer diodes. Another symptom of this will be marginal or out-of-tolerance frequency response. More typically, you may have excessive DC offset in the Lin mode (caused typically by a leaky transistor in the Log amp Lin mode).
You generally will not not notice a DC offset in Log mode, but it will definitely be noticed in Lin mode.
For whatever reason the 8557A procedure is a little harder to follow than the same procedure in the 8558A manual. Download that portion from the 8558A manual and compare notes with the 8557A procedure that you have. This may help you understand the adjustment procedure.
I am recalling this from memory so there may be a glitch in what I am citing.
Feel free to post more questions.
Don Bitters

开云体育