Variac woes!

[VintageTest]

I recently bought a defunct Tek 7904 off of fleabay - reported as "won't switch on". I thought to myself "PSU - switch-mode - nobody likes working on them, I'll have a punt" I duly did, and won. Now, any of you who've ever bought gear - especially Tektronix, know that plugging straight in to the mains the second you get it spells disaster. So (I digress) back in the 70's, I came across a rather nice variac in the scrap bin where I worked at the time. I thought "I'll have this". And I did. Back to the present, I stripped the (now 40 idle years in my possesion!) old beast and serviced it and put it all together. However, it was an open-frame model, so I thought I'd mount it on a panel with switches and meters and stuff. So I did, not paying the slighhtest attention to the depth of the mounting holes. I powered up, to be met with that smell you're all familiar with - burning varnish and smoke! I stripped the variac down again, the damage was plain to see . . .so, using a scalpel (don't ask), I gently prised apart the touching wires and inserted little chunks of power transistor mica washer, to prevent them shorting again. Once I'd separated all of the touching wires, I repowered the variac - no smoke - I was over the moon! I covered the mica insulators with hot melt - covering the where the other bolts had made contact too. I reassembled the variac and pot it back in my hurriedly fabricated panel - WITH SHORTER BOLTS - and powered up. All good. The moral of the story? Those little short bolts that the variac came with? - they were short for a reason!

[AlanBeckett]

Well done

Alan

[VintageTest]

All because of my own carelessness Alan, but thanks anyway!

[Refugee]

I have got one of those open frame ones.
I made my own front panel with two 10mm thick steel bars and some laminate flooring. The assembly was then bolted together with long coach bolts so that the dome heads stood high enough to protect the brush on the non panel end. Mine has a square bit on the panel end with 4 holes for 10mm bolts.
The service position is with it standing on the heads of the coach bolts.

[Nick]

Well done on the repair - I had a 7904 for years - lovely scope - was the first scope I had as an undergraduate engineer at STL so had nostalgia value.

Sold it about 10 years ago as it was just too big and clunky - now use a 2465A... Most of the equipment in the photo below has now moved on and been replaced by more modern stuff...

   

EDIT: The little board on the test bench being menaced by the equipment surrounding it is a Jim Williams ultra-low-distortion oscillator ( 0.0003% or 3ppm) as described in Linear Technology Application Note 43f, figure 48. The big amp on the right is a rebuilt Perreaux 2150B (200W min RMS per channel into 8Ω at 0.009% THD)- a glorious class AB amp that was our main HiFi power amp for a few years - when switched on, the top-plate used to flex and the lights dimmed slightly. Next to it is a 150W non-inductive dummy load for soak testing the amp without going deaf

[VintageTest]

What a lovely clean workbench, with some really nice kit on it. I was never able to afford a Tek TM500 rack system, and you seem to have two of them! I limit myself to Tek 7000 series stuff nowadays, although, secretly, I admit to using a 2245A as my daily scope and have a 2430 DIGITAL scope (wash your mouth out!) too, but I like the big, well-made feel of the 7000 series - it's reliability too. Just showing my age now - I remember, back in the war, you young 'uns don't know how lucky you are etc, etc, etc . . .

[ppppenguin]

Again well done on fixing the variac. Also a useful lesson to anyone who's fitting a variac to a panel.

My main 'scope is a 2465B. With a Rigol DS1054 as a hugely capable instrument at a bargain price. When I think about the 'scopes I've used in the past from horrible little Telequipments to unreliable Marconis I realise how lucky we are now to have a wide selection of good instruments at bargain prices both new and secondhand. Whether it's a simple 20MHz Hameg, used, for under £25 at a rally to brand new Rigols with tremendous facilities for under £300.

[Cardigan]

EDIT: The little board on the test bench being menaced by the equipment surrounding it is a Jim Williams ultra-low-distortion oscillator ( 0.0003% or 3ppm) as described in Linear Technology Application Note 43f, figure 48.

I rather like the footnote in that linked piece:
Note 4: History records that Hewlett and his friend David Packard made a number of these type oscillators. Then they built some other kinds of instruments.
How understated...

[ppppenguin]

That was the late Jim Williams, one of the finest analogue engineers that ever lived. He and a few others, notably Barrie Gilbert and Bob Pease, laid the foundations for most of the analogue chips we use today.

[Nick]

When I was working on that board, the circuit was having some stability issues...

Not being a very good engineer myself, I ran out of ideas, so on the off-chance, I emailed Jim, expecting nothing at all from such a busy guy...

Imagine my delight when he replied he same day - he was extremely helpful, great fun, very witty - we exchanged maybe only a half-a-dozen messages in total, but it made a great impression on me - he always had time for folk who were genuinely interested in analogue design.

He died of a stroke 5 years ago - a huge loss to the community - he authored over 350 technical articles, all of which are worth a read...

Bizarrely, Bob Pease died in a car accident (in his famed 1969 VW Beetle) on his way home from a Jim Williams memorial. Pease is another whose every word is to be treasured.

Jim Williams: