开云体育

Inside the S6 Sampling Head there are 3 PC boards. One connects to the
socket inside the 7S12, and nothing else. One is the strobe board, and the
third is the preamp board. They are interconnected in a 3D stack using long
pins. But that is not what my question is about.



At the heart of the S6 is the sampling bridge. This diode bridge is
contained on an extremely thin ceramic substrate inside a very substantial
steel clamshell "cavity". Four SMA barrels and 5 pins enter this "cavity". I
was troubleshooting one of my S6 heads that had a lot of unusual noise. This
was not sampling noise which would look like a fat cloud of dots instead of
a thin cloud of dots, Instead the sampling was actually very tight but the
trace was jumping up and down randomly as if there was low frequency
"Popcorn" noise affecting it.



After a few hours of troubleshooting by replacing all three boards of the
noisy head with boards from a good head (and vice versa) I knew the problem
was inside the cavity. I thought there might be a loose bond wire or debris
on the substrate that was slightly conductive so I opened up the cavity to
take a look. I found nothing which might be causing the noise.



But when it came time to put the cavity back together I discovered something
I have never encountered before. Here is where some mechanical genius at Tek
had to come up with a solution to a an impossible packaging problem. The
barrels cannot be soldered to the substrate since they will be torqued
during the assembly of the sampling head and this would break any connection
to the delicate ceramic substrate. They have to be free to turn a bit.



At first glance the SMA barrels look like typical Female SMA to Female SMA
Adapters. But on closer inspection one end has a tiny protruding "male" pin.
The "male end must be screwed into the cavity after it the top and bottom
halves of the cavity have been screwed together. So there is no way to see
what is going on inside the cavity as you screw it in. Somehow (I haven't
got a clue how) you must stop just as the male tip touches the ceramic
substrate and makes contact with it. If you go any further you can damage
the ceramic substrate.



Since both of the barrels coming out to the front of the sampling head are
connected together (the S6 is a loop thru TDR head) you can carefully turn
the barrels until you get continuity from center pin to center pin of each
barrel since the loop thru connection is made on the substrate.



I have absolutely no idea how you tell when either of the two output barrels
are touching the substrate. The only way I think Tek did it was by applying
strobe signals and monitoring the output on the center conductor of the
barrels. As I turned them I kept hoping for some change in resistance
between the center conductor of the barrel and the 5 pins that go into the
cavity. I never got a reading of any kind. In the end I failed to get any
output at all from the head when I was done. It was dead.



It did seem to eliminate the noise but it eliminated everything else as well
:(



Does anyone know how Tek did it?



Dennis Tillman W7PF

I have absolutely no idea how you tell when either of the two output barrels are touching the
substrate. The only way I think Tek did it was by applying strobe signals and monitoring the

output on

the center conductor of the barrels. As I turned them I kept hoping for some change in resistance
between the center conductor of the barrel and the 5 pins that go into the cavity. I never got a

reading

of any kind. In the end I failed to get any output at all from the head when I was done. It was

dead.

Dennis Tillman W7PF

How did you check for a change in resistance, Dennis?

Craig

The input (loop thru) signals J10 and J12 are 10k from another pin, "O".
The output signals at 16k and a diode drop away from another pin ("J" ->-
J15 or J16 ->- "M"). I feel like this should be measurable with an ohm
meter to see when you're making good contact, but I'm not going to find out
the hard way that it doesn't work that way.

All the manual says is "The assembly female 3 mm coaxial connectors may be
replaced by using a small wrench on the flat portion of the connector to
remove and replace a connector from the assembly." (Page 5-1, second to
last paragraph)

Dave Casey

On Fri, Jun 15, 2018 at 1:42 AM, Dennis Tillman W7PF <dennis@...>
wrote:

Inside the S6 Sampling Head there are 3 PC boards. One connects to the
socket inside the 7S12, and nothing else. One is the strobe board, and the
third is the preamp board. They are interconnected in a 3D stack using long
pins. But that is not what my question is about.



At the heart of the S6 is the sampling bridge. This diode bridge is
contained on an extremely thin ceramic substrate inside a very substantial
steel clamshell "cavity". Four SMA barrels and 5 pins enter this "cavity".
I
was troubleshooting one of my S6 heads that had a lot of unusual noise.
This
was not sampling noise which would look like a fat cloud of dots instead of
a thin cloud of dots, Instead the sampling was actually very tight but the
trace was jumping up and down randomly as if there was low frequency
"Popcorn" noise affecting it.



After a few hours of troubleshooting by replacing all three boards of the
noisy head with boards from a good head (and vice versa) I knew the problem
was inside the cavity. I thought there might be a loose bond wire or debris
on the substrate that was slightly conductive so I opened up the cavity to
take a look. I found nothing which might be causing the noise.



But when it came time to put the cavity back together I discovered
something
I have never encountered before. Here is where some mechanical genius at
Tek
had to come up with a solution to a an impossible packaging problem. The
barrels cannot be soldered to the substrate since they will be torqued
during the assembly of the sampling head and this would break any
connection
to the delicate ceramic substrate. They have to be free to turn a bit.



At first glance the SMA barrels look like typical Female SMA to Female SMA
Adapters. But on closer inspection one end has a tiny protruding "male"
pin.
The "male end must be screwed into the cavity after it the top and bottom
halves of the cavity have been screwed together. So there is no way to see
what is going on inside the cavity as you screw it in. Somehow (I haven't
got a clue how) you must stop just as the male tip touches the ceramic
substrate and makes contact with it. If you go any further you can damage
the ceramic substrate.



Since both of the barrels coming out to the front of the sampling head are
connected together (the S6 is a loop thru TDR head) you can carefully turn
the barrels until you get continuity from center pin to center pin of each
barrel since the loop thru connection is made on the substrate.



I have absolutely no idea how you tell when either of the two output
barrels
are touching the substrate. The only way I think Tek did it was by applying
strobe signals and monitoring the output on the center conductor of the
barrels. As I turned them I kept hoping for some change in resistance
between the center conductor of the barrel and the 5 pins that go into the
cavity. I never got a reading of any kind. In the end I failed to get any
output at all from the head when I was done. It was dead.



It did seem to eliminate the noise but it eliminated everything else as
well
:(



Does anyone know how Tek did it?



Dennis Tillman W7PF

On Thu, Jun 14, 2018 at 11:42 pm, Dennis Tillman W7PF wrote:

I have absolutely no idea how you tell when either of the two output barrels
are touching the substrate. The only way I think Tek did it was by applying
strobe signals and monitoring the output on the center conductor of the
barrels. As I turned them I kept hoping for some change in resistance
between the center conductor of the barrel and the 5 pins that go into the
cavity. I never got a reading of any kind. In the end I failed to get any
output at all from the head when I was done. It was dead.

Dennis,
This from a Maintenance Note from June 74:
//S-6 SAMPLING HEAD INTERMITTENT OUTPUT
In the past the S-6 Sampling Head has had an Intermittent problem in the Hybrid Bridge Circuit.
The center conductor of the 3mm connector on the front panel connects directly to the bridge sub-strate
via a mechanical spring loaded plunger. The Intermittent output Is caused by the plunger. Norm Farmer of
the Rockvllle Service Center has suggested a field fix which entails cleaning the plunger. When you re-Install
the spring and plunger make sure the plunger is free enough not to bind up when pushed all the way Inside
the center connector, yet not so free that it will fall out when you turn the connector upside down. This
practice insures that the center connector makes good contact with the plunger. The front 3mm connector
will be the only one you probably will have to be concerned about. You will need a wrench to remove the
3mm connector. You can do this by increasing the width slightly of a 3/16th open-ended wrench
The 3mm connector spring and plunger are not shown In the S-6 manual. The part number for the spring
is 214-1072-00, and the plunger Is PN 131-0632-00.//

/H?kan

Hi Craig, Dave, and H?kan,
I failed to follow my own advice and RTFM before I took the sampling bridge cavity apart. I knew when I was debating to do this that I might be in over my head but all the previous things I tried up to this point convinced me that there was nothing unusual to a sampling head after all. So confidence gave way to cockiness when I should have stopped and read the manual to at least know what to expect. If I look back on my life it has been filled with similar behavior. There are two possible outcomes: 1) I admit that I broke it (whatever it was) and I have to face the consequences, or 2) I have an "Oh Shit" moment followed by a desperate scramble to first figure out what happened and then I start a desperate search for a solution.

On balance I usually succeed when I am fortunate that option 2 is available to me. In the process of searching for a solution I learn a huge number of things that, until that moment, didn't seem very important to know. Many of those become very useful later on for avoiding the same problem twice. But there is a curious side effect of learning all the things which is to increase my overall confidence level. This makes it more likely that I will fail the next time to RTFM before I take the next thing apart.

I see now that I have been stuck in an endless spiral of screw-ups of my own making. Fixing them has certainly kept me busy most of my life. But every once in a while I surprise myself by doing something really astounding (at least to me) and I think all this stuff I learned had a purpose.

Dennis Tillman W7PF

I see now that I have been stuck in an endless spiral of screw-ups of my own making. Fixing them has
certainly kept me busy most of my life. But every once in a while I surprise myself by doing
something
really astounding (at least to me) and I think all this stuff I learned had a purpose.

Dennis Tillman W7PF

Hey - don't beat yourself up Dennis. The head might not be bust, there might still just be a duff
connection.

You didn't do something really stupid, like me when I plugged a harmonica connector one pin out on the
power supply of my 577 curve tracer (unplugged originally to find a dead tant). Blew up most of the
silicon in the thing.

I've now got it pretty much sorted, other than one remaining problem with either the step generator or
amplifier. Real nightmare to sort out. Rather fortunately every piece of silicon was easily available,
and socketted - even the dual FETs, which are still made.

Craig

Hi Craig,

Funny you mention it. I have been thinking that there may be a tiny
possibility if I actually read the maintenance directions on the sampling
bridge cavity, which Hakan wrote out so nicely for me, that I might get it
to work again. It would only be an academic exercise since the noise was too
large to want to ever use the head.

The 577 amplifier is a work of genius. I had to read the explanation 4 times
before I understood how it worked. Why they did it that way is beyond me but
that is because I don't have a clue what problem they had to solve. But
understanding the theory of operation was not enough insight for me to
actually find out how to fix it. One of my 577s some years ago had hum
pickup on the most sensitive Vert current settings as I seem to recall.
Eventually I had to give up since I couldn't even see the hum because it was
so tiny. But after being amplified it was noticeable on the screen.

Dennis Tillman W7PF

If I look back on my life it has been filled with similar behavior. There are
two possible outcomes: 1) I admit that I broke it (whatever it was) and I have
to face the consequences, or 2) I have an "Oh Shit" moment followed by a
desperate scramble to first figure out what happened and then I start a
desperate search for a solution.

Oh but you are definitely not alone on that ;-)

I had a good laugh when I saw which (apart from the time axis scaling) seemed awfully familiar ...

Cheers
Peter