Tektronix 7A29 teardown

[muvideo]

Hello,
Today teardown of a unlucky 7A29 plugin.

I was troubleshooting it and started to take pics as
reference for the reassembly process, so here are them.
The plugin is a single channel vertical amplifier for the 7000
series oscilloscopes, if used with the 7104 or R7103
the bandwidth is a nice round 1GHz.
This unit has a broken position knob, and some
ranges that dont work.


The internal construction is nice, with the signal path
built more like water piping than electrical circuit
On the right side there is the switching assemblies ,
the readout switch top rear, and amplifer chain.


The amplifers are the square ceramic hybrid ics.
On the left side there is the delay line, and bottom
side of the main board.


The signal path is simple: the signal enters from
the front bnc, goes trough the delay line and
into the bottom switching assembly that contains
the AC/DC selection and input protection relay.
From the first assembly it goes to the second
switching assembly on top, where it is attenuated
and then to the amplifiers.

Here is the problem, in all it's glory, it is the attenuator.
Contains three group of switches and resistors.
The stages provide attenuation factors of 2x 5x and 10x,
while they keep the input impedance 50ohm.


Each stage is be combined with the others to
give a total of 7 ranges in 1-2-5 sequence,
from 10mV to 1V per division.
This is the camshaft that actuates the switches
in sequence, through metallic rollers.


This is the other side of the actuators.


And finally the actuators removed from their
sliding vane, here can be seen that the
actuators push on 4 pins each: two on
top and two on bottom, and each
pin has it's trimming screw.


Time to remove the input and output connectors.


And remove one of the covers. On the bottom
plate there are the gold plated switches, and
on the top there are the resistive attenuators,
printed on ceramic boards.


The metallic strings, are pushed down to short,
or up to touch the attenuators.


Here the hybrid attenuators, on the left the 5x,
center 10x and right 2x.


The problem is the attenuators are
fried, in particular the 2x has the resistors
to ground opened. The left and right
pads go to ground, while top and bottom
are inserted in the signal path when
the 2x is selected:


The 10x has part of the resistors overheated
and off-spec, bottom of the picture:


the 5x is fine, the black lines are
the cuts used to trim resistors to
the correct value (laser?):


I hope I will find some spares...
That's all for now.

Fabio.

[robrenz]

Nice Pics muvideo! hope you find some parts

[nctnico]

If you know the values, you could replace the resistors with small (0402 or 0201) surface mounted resistors. I'd use kapton tape to keep the solder from getting onto the contact areas.

[muvideo]

nctnico, thank you for the tip.

I know the values, they are on the manual.
The 2x attenuator probably could be modded,
I have to check contacts clearance for the 10x one.

It can ba a good solution for checking the
rest of the plugin.

Naturally I hope to find some original spares,
the parasitic capacitance of the discrete resistors
and the fact that I will not have the exact resistance
values could screw up plugin response a bit

Fabio.

[PA0PBZ]

Not very helpful to you, but interesting to see that it clearly failed at the weakest point (see the x2 resistors to ground, they failed at the laser cut).

[mazurov]

Nice teardown! Now I have an inspiration to fix one of mine with a mechanical issue with attenuator - it it hard to turn in the clockwise direction.

[nctnico]

nctnico, thank you for the tip.

Naturally I hope to find some original spares,
the parasitic capacitance of the discrete resistors
and the fact that I will not have the exact resistance
values could screw up plugin response a bit

Fabio.

You can get 0402 resistors in the craziest values these days and their parasitic capacitance only starts to create problems at several GHz. I recently made some filters with 0402 components and those filters still behaved very well at 3GHz.

[JoeyP]

If you use chip resistors to repair it, keep in mind that the max input specification will be severely derated. The 7A29's max specified input of 10Vrms is 2W into 50 ohms. The 2x attenuator is the first in the path, so 1.5W of that power will be dissipated in the 2x attenuator. Standard 402 resistors are rated at 0.063W, 201's are rated at 0.05W. You'd have to use a bunch in parallel to get anywhere near 1.5W. The parasitics for that many in parallel (if they would even fit) would totally screw the response.

[muvideo]

JoeyP you are right, I was thinking to this aspect.
If I'm not mistaking I will have about 3.7V rms maximum
input voltage, before overheating the 133ohm resistors
(2x33ohm from input + 2x133ohms to ground) of the
2x attenuator.

It's a good comprimise while I wait for the right
combination for the spares.

Fabio.