Test Equipment Anonymous

Obtained a Tektronix 2215A for a slightly borderline £58 delivered. 60MHz 2 channel crate with delay timebase. Slightly more than hamfest fees but my test gear was costing a car on top of previous purchases so that is a good deal. It looks like it works absolutely fine but has the usual problem with focus which I will describe below...

Firstly money shot. Quite tidy
Guts
Our first problem. As always the HV divider chain is drifted thanks to the rancid old carbon comps and some clever person has tried to fix it and failed. Actually failed quite badly at fixing it. There are supposed to be 4x 510K resistors there and we have 1x 656k, 1x 503k, 1x bodged 340k (the replaced one which has failed terribly) and 1x 636k. Doh!
And of course the usual cracked RIFAs. This time only the Y-class capacitors so not good not great. The main X-class is not a RIFA which is good news.

Mouser order will be fired off once all the gear I ordered turns up and has been evaluated.

Ok so this one finally got some attention today after my Mouser order was finally fulfilled.

The mains filter and X2 capacitors previously indicated were swapped out with exactly the same parts but considerably less than 40 years old. This was trouble free as expected. The HV divider resistors were however a bit of a mess. I found two "substitute networks" in there which were completely wrong. Also all the carbon comp resistors were completely all over the place.

The following bodies were extracted:


The resistors were substituted with somewhat more correct Vishay VR37 510K units. Much better:


Initial power up was uneventful (thank goodness as I hate the power supplies in these) and it appears to work as a scope now.


After letting it warm up for a bit it's obvious the calibration is shot by the half wit monkey who was in it before me. Also the trigger is slightly jittery and occasionally misfires which suggests that some of the capacitors will need replacing. For another day that. Happy with today's progress.

I had the same issue (twice) with the HV divider resistors in a B&K 2120 20MHz scope.

Ditto in my PM3400. The brightness fell over 15 minutes, eventually becoming unusable.

Stress leads to failure. Applies to humans too

MED6753 wrote: ↑Mon Jul 01, 2024 8:03 pm I had the same issue (twice) with the HV divider resistors in a B&K 2120 20MHz scope.

HV resistors were a terrible problem in valve (tube) scopes. Back in the day they were expensive and hard to find and not at all reliable. Very often some hero would have attempted a repair using 1/4 watt resistors, with a max working voltage of 500V and they would die quickly. The Vishay VR types are damned good.

I haven't noticed high voltage resistors playing up as much in solid state scopes. I've had EHT capacitors play up.

bd139 wrote: ↑Mon Jul 01, 2024 8:41 pm Stress leads to failure. Applies to humans too

Odd you should mention that. On my recent visit to TNMoC, I was looking at their Colossus rebuild and admiring the impressive rows of valves. They had a big desk fan to blow a soothing breeze over the bank of thyratons. The guide said it dissipated 8kW. 8kW amounts to around £2 an hour to run. 2x24xx365 = £17,520, the better part of a not inconsiderable sum for something of a wing and prayer operation. They may be taking advantage of their charitable status and off peak rates, so it may be lower, but still a lot of loot. When Tommy Flowers was presented with the problem it was expected he would produce a design based on relays, but he said relays were too slow and valves were the way to go. But valves are unreliable. Not if they are never switched off said Flowers. And so it was.

In my view Flowers was never properly recognised for his efforts.

I asked the guide about powering it, and he said he knew the original Colossus was never switched off, but they couldn't afford to do that. They turned it off and powered it up slowly using a variac which avoided thermal shock. They had 10 dead valves a year, which was manageable. The total number of valves is 2,400.

The thyratron bank. Side view including fan. The scaffolding is because they are having to have the roof replaced.

I actually went to TNMOC and saw that first hand back in 2019. Doesn't seem like that long ago but it's half a decade already. Very impressive bit of equipment that and on the day it was actually operational.

Certainly agree with the recognition of Tommy Flowers

On that, one of his...

bd139 wrote: ↑Wed Jul 03, 2024 10:11 pm I actually went to TNMOC and saw that first hand back in 2019. Doesn't seem like that long ago but it's half a decade already. Very impressive bit of equipment that and on the day it was actually operational.

Certainly agree with the recognition of Tommy Flowers

On that, one of his...


Screenshot 2024-07-03 at 23.03.51.jpeg

He's got a couple of tiny roads named after him.

Strange to think they used a CV188 voltage stabilising neon tube as the source of random numbers.

One of my Solartron DVMs also used a neon as it's voltage reference, calibrated against an internal Weston standard cell.

Well technically he's got a whole estate named after him (old PO research station converted to flats)

Well the CV188 was a voltage reference yes, but a crap one. The datasheet specifies the noise limits which are at 100V output with no more than 100mV measured with a selective level meter between 50Hz and 5000Hz. Someone went "hey lets just make that noise bigger rather than smaller". Also they specified the pulse output at around 18KHz which was a lot higher than the specification. I wonder if there was much more wideband noise than quoted.

I literally built the same circuit with a zener at the wrong current and a bunch of 2n2222a's. Flat at 1dB to 30MHz at least. Sure it'd have been better with some diode selection and some slightly less lethargic transistors. Storage normaliser would have been better but it was a bit outside the scope of a 7L12 for measurement purposes that.

I see a "Chartwell Ct" with Flowers Close. Since that building is the old Dollis Hill GPO Research Station, it seems to tie together.

tggzzz wrote: ↑Fri Jul 05, 2024 11:19 pm I see a "Chartwell Ct" with Flowers Close. Since that building is the old Dollis Hill GPO Research Station, it seems to tie together.

It's known as the Flowers Estate informally. I suspect many people don't know what the capitalised F represents.

bd139 wrote: ↑Sun Jul 07, 2024 2:19 pm

tggzzz wrote: ↑Fri Jul 05, 2024 11:19 pm I see a "Chartwell Ct" with Flowers Close. Since that building is the old Dollis Hill GPO Research Station, it seems to tie together.

It's known as the Flowers Estate informally. I suspect many people don't know what the capitalised F represents.

There was an informal flowers estate between my hall of residence and university. Roads were primrose, violet, pansy, ...

For someone who had such a significant effect, it shouldn't be merely informal.

Then again there are many places not named on maps. In Cambridge everybody knows "Mitchams corner", after a earl-mid century shop. Near me there are similar places named after farm outbuildings, pubs (allegedly), confectionery. The only way gurgle knows about them is if a local shop is appropriately named.

Sometimes road are renamed. I remember "Nursery Close" (near you !) being built in the early 60s because it stomped on a nursery. After a few years the Hyacinth Bucket types didn't like the name due to its association with small children, so it was changed to the completely overblown "Regency Close".