Bonjour a tous:

Working on 7L5 with L3 input module and OPT 25 tracking generator.

FREQ knob has several defective positions, 1k 2k 5kHz : These intermittently fault to 100k, 200k and 500k respectively.

It contains a shaft encoder under the knob housing

As you jiggle the knob you can get them to read properly for a moment.

Obviously a dirty contact inside the encoder.

To access the encoder guts, seems the entire front panel must come off.

1/ Any way to spray contact cleaner inside ? What recommended cleaner to use? How to get it in?

2/ Any easy way to remove the knob /encoder without complete panel removal?


Many thanks!

Jon

On Fri, Nov 13, 2020 at 02:30 PM, Jean-Paul wrote:

2/ Any easy way to remove the knob /encoder without complete panel removal?

Yes. The freq knob/encoder assembly is attached to the panel from the back by two
Pozidriv screws. These screws should be accessible through the circular
cut-outs in the front panel PCB (they are just to the right of the edge connector).
Because of limited space, I used a Pozidriv bit and a 1/4" open ended wrench
to rotate the bit. Once the screws are removed, you should be able to gently pull
the assembly away from the front panel.

Note: I have never disassembled the knob/encoder, but I though that I
read somewhere that it was an optical encoder. If so, spraying contact cleaners
into the assembly might not be the best idea. Perhaps someone on this list
knows more about how the assembly is constructed.


dan

Rebonjour, many thanks for the tip on knob/encoder removal.

I have a second unit that is much worse, on both shaft encoders, I shall try the right angle PosiDrive to remove the encoders.

The original TEK service manual reveals the knobs are just switches, circular PCB and wipers. No shaft encoder!
Pressing the knob in/out or pushing the front panel can change it 5k><500K. Probably a single contact on the x1 x100 switch.
I washed with 99% iso alcohol, no change. Suspect oxidized connector on shaft encoder to front panel PCB.

Any other experienced TEK spec analyzer plugin experts out there?

Besides this fault, it tests 100%. Amazing for such a complex box that is ~40 years old!

Bon Journée a tous,

Jon

PS: With the nice large CRT of 7603, and a 7L12 + manual, I am in spectrum city.




.


Jon

TekWiki solves!

http://w140.com/tekwiki/wiki/7L5/Repairs

Jon

On Sat, Nov 14, 2020 at 04:42 AM, Jean-Paul wrote:

The original TEK service manual reveals the knobs are just switches, circular
PCB and wipers. No shaft encoder!
Pressing the knob in/out or pushing the front panel can change it 5k><500K.
Probably a single contact on the x1 x100 switch.
I washed with 99% iso alcohol, no change. Suspect oxidized connector on shaft
encoder to front panel PCB.

Interesting. I've found the article that I had vaguely remembered at first:

"A 5MHz digitally controlled spectrum analyzer" published in TekScope Volume 7
Number 3. Towards the end of the article, it describes the optical encoder innovation,
and states that it is used for TIME/DIV and FREQUENY SPAN/DIV controls.

Looking at the Front Panel <2> schematic in my service manual, only the resolution
selector seems to be mechanical. Is it possible that Tek had at some point switched
to an all-mechanical solution, and your unit happens to be one of those?


dan

On Fri, Nov 13, 2020 at 8:30 PM Jean-Paul <jonpaul@ix.netcom.com> wrote:

FREQ knob has several defective positions, 1k 2k 5kHz : These

intermittently fault to 100k, 200k and 500k respectively.

I have a 7L5 with a similar problem and it's driving me nuts. Restoration
of front panel PCB sockets made no difference, so it all points to the knob
(switch) being the problem. I found a few old posts on the group dealing
with this issue but (someone please correct me if I missed it) no one yet
dared disassemble the knob. IIRC I read somewhere else that it may crumble
into zillion little pieces which are a nightmare to put back properly.

I've seen a fair number of 7L5's on eBay with broken knobs; would be great
if someone from our group had such an unusable one and would be willing to
dissect it and photograph the process. I admit I'm a chicken to disassemble
mine and possibly end up with a dead 7L5 (far from Tek homeland in Eastern
Europe, where I live).

Best regards,
Nenad

Today I attempted disassembly of my Resolution/Freq. Span knob and got halfway there, photos (album name "7L5 Knob Repair Attempt"):
https://groups.io/g/TekScopes/album?id=256833&;p=Name,,,20,1,0,0

The resolution section of the switch is mechanical, no challenge to disassemble and easy to restore. The span section appears to feature an optical encoder, and I could not find a non-destructive way to open it up. I'm currently stuck at the third photo in the album, the white plastic base of the frequency span knob still holds tight even if the two screws are removed. The metal L shaped insert? (circled in red on the photo) appears glued to the central shaft and won't budge. Wonder if that's by design, or somebody had glued it on purpose after attempted repair? I'm suspicious as I discovered two of the span knob pins heavily bent during inept assembly, and I doubt it came like that from Tek.

Straightening the pins, tightening the sockets and cleaning everything did not make much improvement. Bit marked TJ on the schematic is persistently acting up, even the lightest touch of the span knob makes it jump from 500k->5k, 200k->2k... etc.At this point I'm out of ideas.

Best regards,
Nenad

Nenad, exactly the same symptoms, as mine has. Slightly touching the knob in 1,2,5 K res will make the 100k 200k 500k fault disappear momentarily. All other resolutions are fine.

Any TEK Spectrum Analyzer veterans out there who can assist?

Many thanks


Jon

On Tue, Nov 17, 2020 at 05:25 PM, Nenad Filipovic wrote:

Today I attempted disassembly of my Resolution/Freq. Span knob and got halfway
there, photos (album name "7L5 Knob Repair Attempt"):
https://groups.io/g/TekScopes/album?id=256833&;p=Name,,,20,1,0,0

Hi Nenad,

Your photos encouraged me to attempt to disassemble my own res/freq knob assembly.
I was hoping to be able to tell you whether my L-shaped insert also appears glued, but
interestingly, I could not even get that far. The best I could do was to remove the black
plastic resolution knob/ring.

My 7L5 s/n is >B090000, and therefore has the 263-0064-00 encoder. In this version,
the hollow cylindrical centre of the (die cast?) metal base seems to extend all the way
through the frequency span knob as a single piece, rather than being made up of
two fluted interlocking parts. Hence, the resolution and frequency span sections do
not come apart when the resolution knob/ring is removed.

Peering down the length of this cylindrical opening toward the underside of the
frequency span knob, I believe I see the bottom of a machine screw, but I cannot be
certain.

It would appear that different disassembly procedures are needed for
260-1693-0x and 263-0064-00 encoders. I don't think that I will attempt any further
disassembly, for fear of damaging or marring a perfectly working pristine unit.

Perhaps someone with an older 260-1693-0x assembly (and more courage) will be
able to make further progress.


Good luck,
dan