HP 59307A VHF Switch refurbish (New LEDs for old bulbs)

[TerraHertz]

A couple of very sorry looking HP 59307A VHF Switch units arrived. They were cheap due to the poor condition.
Dirty, so many stickers, baulky buttons, and supposedly one of the units didn't even power up.


Seller's pic.

The HP 59307A VHF Switch is one of the very early HP-IB devices. It's very simple. No CPU intelligence at all, just small scale TTL ICs. A couple of refs:
  https://bukosek.si/hardware/collection/hp-59307a.html
  https://www.holzleitner.com/el/hp-59xxx/hp-59307-en.html



Considering how poorly packed they were, this was one of those rare shipping miracles when things are NOT damaged despite the sender's best efforts to ensure they would be smashed. Thin layer of bubble wrap around each one, but plenty of void space for them to bang around in.


As arrived. No more attractive but undamaged.


Hey, what's this! First nice surprise. So this is a NASA relic. Too bad the sticker is damaged.


JPL too!  (Jet Propulsion Labs.) I suppose this is the far low end of rocket science,


Then a really nice surprise. Underneath one sticker, a prize. Normally I remove all stickers but this NASA one is a keeper.


Virtually all the old bulbs in both units are blown.  Including the Power light and all the other bulbs in one, which is why the seller thought it didn't power up. Should I just replace the bulbs? (And have to keep replacing them in future.)

Of course not. LEDs it is. I didn't have a manual yet, but could see from the traces they are all run from +5V with switching of the low side leg to ground. LEDs will need a series resistor. The Power and Remote lights a resistor each. But the buttons not, because the mechanism has an 'only one at a time' interlock. No not the cheap kind where you can actually push multiple switches down at once. These really won't allow more than one. So there can be just one resistor per group of 4 buttons.

I thought about using dimly lit orange LEDs, to pretend they still used bulbs. Or something spacey like violet. Maybe annoyingly intense blue, to blend in with every other bit of modern consumer electronics? Hmm, so tempting. Or traditional red?
But no, let's just go with a nice green.


The dead button bulbs from the 1st unit. Who can guess what happens next?


Yep.


Skipping all the messy scraping out the remnant glass and crumbly white cement with an X-ACTO knife and stabbing myself with the blade point a few times, then solder sucking and removing the wire remnants, these are the cleaned up bulb bases.  The black parts are solid glass, with a grommet of solder adhesion at the tip. (I'd really like to know how they do that.)

The LEDs are standard 5mm dia round top green LEDs, modified.  To fit in the lamp bases, the bottom flange had to be removed. Very easily done in the lathe, holding the body of the LED lightly in the chuck and using a sharp cutter. (A parting tool actually, but only because that's what happened to be in the tool post then.)  Then the round tops of the LEDs were ground off (very easily!) on a bench grinder. Two reasons:
1. To produce an even illumination of the button. Removing the LED 'lens', and replacing with a flat, frosted surface closer to the die worked well.
2. Height. This resulted in a LED-bulb size close to the original.

The bits of teflon tubing on the long LED legs, are more to act as a height spacer than to really insulate.

In the 59307A the bulb center contact is negative, the bulb body is positive. And of the 16 LED-bulbs I made, I only got the LED wrong way round in one. No idea how that happened, I was being careful to observe LED orientation. For these LEDs the internal 'substrate' (larger) lead form is negative. It's easy to see it...


A LED-bulbs set made. (7 here, one more in the unit already.)

[TerraHertz]

Using a bit of tubing as a bulb holder/extractor. (And a small screwdriver to push the bulb sideways in or out of the clips, but would have needed three hands.)
It's useful because if you drop the bulb here, it falls into the unit.


Current limit resistors on the button LEDs. One 220R resistor per group of four.
Originally that bare wire ran along the whole row, and at the left was wired directly to to +5V on the PCB. (Where the orange wire connects to the board.)

The switch interlocking mechanics were a bit 'sticky' too. Some light lithium grease fixed that.

The structure at top right in this photo used to hold two more small bare-wire bulbs, for the Power and Remote indicators. The green wire to it carried +5V. Interesting color choice.


The inevitable RIFA mains caps, replacing them. The ones I had in stock were short leaded,



Replacing the indicator bulbs with LEDs, and fitting the resistors in the housing with them.
Used 1K resistors for these due to the small indicators not needing as much backlighting as the buttons.


Cleaned up and working. Buttons switch the co-ax relays correctly. Still to verify HP-IB operation.


The button switch assembly is mounted on the one PCB. If the PCB isn't located exactly right and any buttons are touching the edges of the front panel holes, it really messes up the button operation. Because if the depressed switch can't pop out easily, the interlock prevents fully depressing any other switch in the group of four.
One of the units was doing this.

[TerraHertz]

The photo doesn't do them justice. To the eye the green looks gorgeous. I'm very happy with this.

 I guess my photography 'skills' are crap, because I could not capture the niceness of it, despite several attempts. 

As for 'Kubrick'...   The Moon landings were real (for eg, other nation's lunar orbiter programs have produced video and photos of the landing sites, landers, rover tracks, etc.) But there is a related conspiracy - an organized effort to convince as many people as possible that those landings never happened. As for _why_ that's being done, look up Yuri Bezmenov and 'demoralization.'
The Moon landings are a great test of anyone's intelligence and general grasp on reality - both physics and politics. Can they tell the difference between truth and targetted propaganda? Or are they perhaps acting as a dishonest agent, which is far worse than being a fool.