Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

It's actually a very clever design Not only do the neon lamps turn on the LDRs thay act as switches / level detectors. Turning on a Neon in a series circuit clamps the voltage at that point and stops neons lower down the chain lighting. The LDRs have different areas giving different on resistances.

So... like making a Dekatron out of discrete components so you can change the visual format...? Neat.

mnem
 

[factory]

...How wrong I was, unfortunately damp has got inside the block & eaten away some of the traces on the photo-resistor plate and this plate is the main difference in this display board between the 5248 & the 5245 models.

Before & after cleaning with IPA,         Looks like I'm stuck with only 7 usable digits unless/until I can find a replacement.     (SNIP)

David

Aww, man... that truly blows. Think there's any chance of resurrecting with a silverized conductive pen? I can't think of a better candidate to at least give it a try.

mnem
 

It could be, although we had some of those silver pens at work and I never remember being able to draw anything as fine as the traces on this plate. The traces almost look like a gold coloured version of the stuff they coat glass with to make mirrors, they tend to tarnish & deteriorate with damp too.

I've inquired about some of the boards that Sphere have, the option 002/3 ones and some of the earlier boards that are used in a lot of the counters & some of the voltmeters too, all pretty much impossible to find over here as no-one ever offers them for sale (I think they pinch the number readout bulbs & bin the rest  ).

David

[mnementh]

Americans can't handle more than two Si Units at once without suffering mental overload. So capacitors are 1000000uF and 15000pF

Americans can't handle more than two syllables at once without suffering mental overload.
 
is a crap-ass-i-tore...?

Sounds like you tore someone a new asshole, but... Why? You keep leaving out the juicy bits... 

mnem
fifteen thousand puffs...? Well, it could be 4:20 somewhere... 

[med6753]

Smart draftsman uses all the prefixes

And don't use ".".

Current gudance is use the multiplier letter as the decimal eg   0u1 4n7 2M2  0R1 etc.

Sorry, I'm old school. The decimal point is in play. Along with K = 1000. So a 2200 ohm resistor will be 2.2K.

 

No, thrice no. Upper case K is Kelvin  so 2.2K is very cold. 2k2 is OK.

Sorry, but nope. Go look at the schematics I posted today. The 'K' is capital.

[mnementh]

Smart draftsman uses all the prefixes

And don't use ".".

Current gudance is use the multiplier letter as the decimal eg   0u1 4n7 2M2  0R1 etc.

Sorry, I'm old school. The decimal point is in play. Along with K = 1000. So a 2200 ohm resistor will be 2.2K.

 

That's ok until you get a shitty Mauritron scan to work off that looks like a pepper mill exploded near it

Yeah... this is what I was able to find for my 4070A after scouring the intardnet for a whole weekend. Started out looking like this.




A fellow eevblog member posted this "cleaned up" version on IMGUR. 




Not surprisingly, it was still full of scribbled notations and bunk that were not at all relevant to the 4070A (a rebadged BNC 630) so I spent a few hours cleaning it up and de-notating it for my own use.

The BS  we put up with for our boat anchors...

mnem
 

[0culus]

Well, I had a nice break away from the internet for my Thanksgiving. That was most refreshing, no offense to you folks. 

I got a few real germanium diodes from a friend. Figured that was better than risking just getting ripped off for relabeled schottky diodes.  Plan is to use them to build another iteration of w2aew's station monitor with AM detector thing he describes in one of his early videos. I made one with mystery diodes (sorta looked like schottky on the curve tracer, but the linear IV curve after the square law region was really depressed compared to most diodes), and it works quite well out to 300+ MHz but it's not as sensitive as it could be made with real germanium diodes.

[mnementh]

Now you need to find you some crystal earphones and build one of the zero-transistor ElectroCraft "AM Detector" kits with tuning coil rolled on a cardboard Quaker Oats carton...   

mnem

[mnementh]

...How wrong I was, unfortunately damp has got inside the block & eaten away some of the traces on the photo-resistor plate and this plate is the main difference in this display board between the 5248 & the 5245 models.

Before & after cleaning with IPA,         Looks like I'm stuck with only 7 usable digits unless/until I can find a replacement.     (SNIP)

David

Aww, man... that truly blows. Think there's any chance of resurrecting with a silverized conductive pen? I can't think of a better candidate to at least give it a try.

mnem
 

It could be, although we had some of those silver pens at work and I never remember being able to draw anything as fine as the traces on this plate. The traces almost look like a gold coloured version of the stuff they coat glass with to make mirrors, they tend to tarnish & deteriorate with damp too.

I've inquired about some of the boards that Sphere have, the option 002/3 ones and some of the earlier boards that are used in a lot of the counters & some of the voltmeters too, all pretty much impossible to find over here as no-one ever offers them for sale (I think they pinch the number readout bulbs & bin the rest  ).

David

Oh, it is for sure a silver-based deposit; one of the first reliable methods developed for electroplating plastics and ceramic. I've run into this construction method many times; very popular in remote controls from the 80s, along with printed resistors. I was thinking maybe mask off with Kapton tape to make the traces...?

mnem
Time for a "Hail Mary"...!

[med6753]

Well it's been running now for over 2 hours mocking me.  So I'm going to button it up and put it on the torture rack, which I had planned on anyway since I re-capped the PSU. We'll take from there.

The 7904 in the background is picking up the HV Oscillator even though the probe is hanging in free space several inches from the TO-3 output transistors mounted on the top of the HV cage. It demonstrates how important good bonding of the HV cage to chassis and covers is required to prevent that pulse from getting into the vertical circuits.

 

[McBryce]

Smart draftsman uses all the prefixes

Yup, I have a schematic in front of me right now, where the transistors all have a "VR" prefix?? For me that's a variable resistor.

McBryce.

[Robert763]

It's actually a very clever design Not only do the neon lamps turn on the LDRs thay act as switches / level detectors. Turning on a Neon in a series circuit clamps the voltage at that point and stops neons lower down the chain lighting. The LDRs have different areas giving different on resistances.

So... like making a Dekatron out of discrete components so you can change the visual format...? Neat.

mnem
 

Not really,
A Dekatron is a counter. The HP decoder does not count It provides BCD to decimal decoding and high voltage driver.

[Ero-Shan]

Next one: Burster Digistant Typ 6705.

They invariably have one big problem: the keys, each and every one of them!, bounce like they get paid for it. That makes it impossible to use these otherwise nice precision DC sources.
Not to mention the sticker-strewn face of it:



The front comes off easily. The PCB with the display and keyboard also are easy to remove - the ribbon cable can be unplugged at the other side. The keycaps are also no problem, but the actual keys (white nylon) need some force.



Finally, we're as close to the core as we get.



Although one can't see the actual contact point, spraying some deoxit into the keys and hitting them repeatedly did the trick. I saw not a single bounce after that procedure.

Since it was easy to bare the front plate of the PCB and the binding posts, I did that gave it a thorough cleaning. Even the remains of the tamper-proof asset sticker finally came off. It really looks like new now (but I forgot to take a picture ).

And since we all want to see the innards, here they are (except the PSU). A GPIB controller card, a digital board and an analog board (with an ICL7134 DAC). Between the digital and analog boards is a metal shield.



Finally, how does it fare?



 

[Robert763]

Smart draftsman uses all the prefixes

And don't use ".".

Current gudance is use the multiplier letter as the decimal eg   0u1 4n7 2M2  0R1 etc.

Sorry, I'm old school. The decimal point is in play. Along with K = 1000. So a 2200 ohm resistor will be 2.2K.

 

No, thrice no. Upper case K is Kelvin  so 2.2K is very cold. 2k2 is OK.

Sorry, but nope. Go look at the schematics I posted today. The 'K' is capital.

That does not make it right.

[andy3055]

Here is a good site for this question:

https://www.electronics-tutorials.ws/dccircuits/dcp_3.html

[mansaxel]

Sorry, but nope. Go look at the schematics I posted today. The 'K' is capital.

That does not make it right.

It is about as right as using "M" for "milli". Or slightly worse.

[shakalnokturn]

Oh, it is for sure a silver-based deposit; one of the first reliable methods developed for electroplating plastics and ceramic. I've run into this construction method many times; very popular in remote controls from the 80s, along with printed resistors. I was thinking maybe mask off with Kapton tape to make the traces...?

mnem
Time for a "Hail Mary"...!

I've had a couple fine reworks to do and seem to remember masking adjacent tracks wasn't the best solution removing mask didn't leave a clean cut and disrupted the added layer.
What worked best for me was removing the tip from the pen (clockwise unscrew for the CircuitWorks pens) and dipping in a sewing needle tip, move fast and repeat regularly.

[med6753]

Smart draftsman uses all the prefixes

And don't use ".".

Current gudance is use the multiplier letter as the decimal eg   0u1 4n7 2M2  0R1 etc.

Sorry, I'm old school. The decimal point is in play. Along with K = 1000. So a 2200 ohm resistor will be 2.2K.

 

No, thrice no. Upper case K is Kelvin  so 2.2K is very cold. 2k2 is OK.

Sorry, but nope. Go look at the schematics I posted today. The 'K' is capital.

That does not make it right.

So.....all the Tektronix schematics published since the company was founded in 1946 are wrong? Surely you jest. 

[mansaxel]

So.....all the Tektronix schematics published since the company was founded in 1946 are wrong? Surely you jest.

No, I'm not joking; the SI system is pretty clear on this.

The practice of ignoring case and never straying from the 26 characters of the english alphabet is a malpractice that plagues a lot of US-sourced technical documentation et c. Some -hp- gear from the 50s also have such faux pas on their stenciled front panels.
-hp- cleared up their act in the late 60s, IIRC. Tek took some more time to get it right.

We all understand what is happening, and can adjust, but like all such accommodation processes, it uses up mental capacity better spent on understanding what the %&(()% the machine is doing.

[Robert763]

So.....all the Tektronix schematics published since the company was founded in 1946 are wrong? Surely you jest.

No, I'm not joking; the SI system is pretty clear on this.

The practice of ignoring case and never straying from the 26 characters of the english alphabet is a malpractice that plagues a lot of US-sourced technical documentation et c. Some -hp- gear from the 50s also have such faux pas on their stenciled front panels.
-hp- cleared up their act in the late 60s, IIRC. Tek took some more time to get it right.

We all understand what is happening, and can adjust, but like all such accommodation processes, it uses up mental capacity better spent on understanding what the %&(()% the machine is doing.

I think Med6753 was asking me, but I agree, he Tek scehmatics are wrong to use UC K. Electronics is odd that a circuit can use multipliers from pico to Mega on a single unremarkable design. 10^18 range of numbers.

[mnementh]

Oh, it is for sure a silver-based deposit; one of the first reliable methods developed for electroplating plastics and ceramic. I've run into this construction method many times; very popular in remote controls from the 80s, along with printed resistors. I was thinking maybe mask off with Kapton tape to make the traces...?

mnem
Time for a "Hail Mary"...!

I've had a couple fine reworks to do and seem to remember masking adjacent tracks wasn't the best solution removing mask didn't leave a clean cut and disrupted the added layer.
What worked best for me was removing the tip from the pen (clockwise unscrew for the CircuitWorks pens) and dipping in a sewing needle tip, move fast and repeat regularly.

Yeah, I've done similar with resist pens back in the day. Maybe something like these: https://www.amazon.com/AIEX-Miniature-Painting-Brushes-Watercolor/dp/B07PJ5GMX7/

with 000 size, or even a better set with 00000 brushes in it.

mnem
 

[BU508A]

[mnementh]

Learn it, live it, love it!

mnem
 

[capt bullshot]

Next one: Burster Digistant Typ 6705.

They invariably have one big problem: the keys, each and every one of them!, bounce like they get paid for it. That makes it impossible to use these otherwise nice precision DC sources.

Oh, I've seen that type of key before. Even designed them into the first real product I've designed. They didn't bounce when they were new (around 1995), but I've got a leftover piece of that product, and today, these keys either don't work at all or bounce like hell. I had a sample box of them too, I've binned the keys but kept the nice sorting box.

[factory]

Oh, it is for sure a silver-based deposit; one of the first reliable methods developed for electroplating plastics and ceramic. I've run into this construction method many times; very popular in remote controls from the 80s, along with printed resistors. I was thinking maybe mask off with Kapton tape to make the traces...?

mnem
Time for a "Hail Mary"...!

I've had a couple fine reworks to do and seem to remember masking adjacent tracks wasn't the best solution removing mask didn't leave a clean cut and disrupted the added layer.
What worked best for me was removing the tip from the pen (clockwise unscrew for the CircuitWorks pens) and dipping in a sewing needle tip, move fast and repeat regularly.

Yeah, I've done similar with resist pens back in the day. Maybe something like these: https://www.amazon.com/AIEX-Miniature-Painting-Brushes-Watercolor/dp/B07PJ5GMX7/

with 000 size, or even a better set with 00000 brushes in it.

mnem
 

This is how bad it was after cleaning, I think if I had a go at trying to paint that it would end up an absolute dog's dinner. It would also be very difficult trying to find out where they need to connect under the photo-resistive material.


I think the seal has helped trap if any moisture that got underneath, picture below is of one the 5245L boards.


I opened up one of the six identical boards from the 5248M, which seems to have no corrosion. 
The decoder plate is the same as used in the faulty board, I've moved it to the high speed board. Sphere have some of the lower speed boards of this type and I've ordered some.


For completeness here is one of the display decoder plates from the standard 5243L, note the difference in layout of the traces & photo-resistor material.


David

[srb1954]

How difficult was it to remove the ceramic substrate from the 5245L decoder module?

I have a bad decoder module which requires a replacement neon but I haven't been brave enough to attempt removing the ceramic substrate for fear of cracking it.