Vintage Geiger Counter with Deketron DK24 displays

[TheFloyd]

Hi folks,
I have recently added a lovely old device to my vintage collection.   It is a Timer-Counter (Geiger Counter) made by Watson-Victor LTD, in what I assume to be the 50's/60's.
After some initial inspection and measurements, I brought it up carefully on my variac and was delighted to see that it worked perfectly.

I set to reverse engineering the circuit, taking multiple measurements along the way, with the firm assumption that all of those old electrolytic's would be dry and in need of replacement.

this is where it got interesting, and I ended up with a growing list of questions that I would love some expert guidance on.

The circuit is driven by two Toshiba 12AT7 and one 12AU7A valves.   From what I can initially tell, one of the 12AT7's is for amplifying the audio crackles and the other for amplifying the output from the detector tube.   I have not fully investigated the function of the low gain 12AU7A in the middle yet.
The front of the device includes three decade counters, that are using Okaya Musen DK24 Deketron counting tubes.    These wonderful things are the first I have ever seen in person, and their operation is fascinating to me.

Where I need help:

The power transformer, is an A&R 6069, that has a primary winding arrangement that I have not see before.    I have been unable to find any documentation on the 6069 transformer unfortunately to better explain it.
- the primary has the normal 240v black and red wires, and it also includes a single green wire.    I initially assumed that this was some kind of primary winding tap, but it is not.   The green wire is open circuit from any other wire on the transformer (both primary and secondary) and is NOT connected internally to the transformer chassis.    This had me puzzled, especially after I saw that when it connected to the PSU board, it immediatly joined to one lead of that I assume is the secondary heater filament winding.    I didnt even know how to draw this on my circuit diagram, until I found an ancient schematic for a circuit using an A&R 5776, which showed the green wire like a dashed line next to the transformer core.   I am assuming this is the same as what my 6069 has?

- - can anyone tell me what that primary wire is called, and why this power transformer specifically depicts the connection to chassis, and to one wire of the secondary, like this?


The rectification part of the circuit uses two ancient "Top Hat" style diodes that are made by Toshiba and have a part number of 1S94v.    I can't find any datasheet for these either.    They measure reasonably for silicon at 0.5v and 0.52v.
- I am not sure if it is just the way that I have laid out my schematic as i traced and measured the circuit, but I am unfamiliar with the topology of the rectification stage, and frankly, even the way to reference voltage measurements through the circuit...    for the sake of consistency, all measurements I made, are with respect to the earth/chassis. (green wire)


So far, most of the values are measuring ok, and the system operates.   Other than the poor dual-electrolytic where one of its 16uF 500V halves, has 565v across it.    I will need to get that under control before i pop it.    the Cap itself is measuring 17uF on both sides, which is far better than i would have anticipated!



I have not had the time to reverse engineer and draw the schematic of the main board yet, as I would like to get the PSU voltages all corrected first, so that I dont skew my measurements.
I would love to hear thoughts on the transformer, the diodes and the circuit design in general, as it is a bit of a mystery to me so far and I cant find any service manual or technical documentation on the device at all.   it is called "Counter-Timer" "Model-4DSPT10".

Thanks In Advance,
Floyd

[TheFloyd]

Here are some pictures of the device, and the first draft of my schematic for the PSU.

[TheFloyd]

I have just found some information that relates to the use of "electrostatic shielding" within a transformer connected to a wire that you hook to ground.   The purpose is to reduce common-mode noise of specific frequencies of 140dB+.
Looking at the schematic symbol for these transformers, and the description of their external wiring, leads me to thinking that this is what the A&R 6069 transformer is doing.

Fair assumption ?

[Analog Kid]

Don't see what else it could be for.

[gbaddeley]

Yes, it’s an electrostatic shield, a layer of foil between the primary & secondary windings that has a gap or insulation so it’s not a shorted turn. Can you see if the green wire comes out of the bobbin between the windings?  It normally connected to chassis / mains earth.

[floobydust]

An interesting mix of Australian and Japanese components. PCB marking "IU3 KEMP" but no leads.
I would add a mains fuse.

The voltages seem high, as if mains voltage is way up there +10% than design for the 1960's would have. It's a voltage doubler? Who is getting 638VDC?!
Have to look at the Rodan DK24 datasheets, it seems to require 360V. Maybe it's for an OG-3?

I could find no data on the A&R PT6069 transformer but assume the extra wire is for an electrostatic shield, usually see that for CRT scopes and sensitive instrumentation.

Those five blue electrolytics (Philips?) I would outright replace, the bung shrinks and dries out and they short or go low value.
A pic of the power board solder side and we could verify the PSU schematic.

[TheFloyd]

Yes, it’s an electrostatic shield, a layer of foil between the primary & secondary windings that has a gap or insulation so it’s not a shorted turn. Can you see if the green wire comes out of the bobbin between the windings?  It normally connected to chassis / mains earth.

Thanks for this, I suspected it was.  I don't know how i have never come across a transformer like this before.   I guess the sensitivity of the detector in the geiger counter, is reliant on reliable and clean power, to prevent false readings etc.

[TheFloyd]

Unfortunately I cant clearly make out where the green wire goes inside.  (attached image of the underside)

[TheFloyd]

Thanks for your thoughts.

An interesting mix of Australian and Japanese components. PCB marking "IU3 KEMP" but no leads.
I would add a mains fuse.

Oh Wow!   I read this, and it just dawned on me that this device doesnt have any fuse protection at all!    I will absolutely be adding one of those, well spotted and good call!

The voltages seem high, as if mains voltage is way up there +10% than design for the 1960's would have.

I suspect that voltage is so high due to some wildly out of spec components!

Have to look at the Rodan DK24 datasheets, it seems to require 360V.

I had spotted this webpage, and the image of the DK24 shows the exact device that I have, rather than the taller DK23 in the datasheet.   They are probably electrically similar for the most part, other than the differences listed in that DK23.pdf

I could find no data on the A&R PT6069 transformer but assume the extra wire is for an electrostatic shield, usually see that for CRT scopes and sensitive instrumentation.

Good to know, and it probably explains why I have not come across one of these before, as while i have pulled apart mountains of old valve gear over the years, I have never worked on an old CRO or other sensitive equipment.

Those five blue electrolytics (Philips?) I would outright replace, the bung shrinks and dries out and they short or go low value.

Totally agree!   Just need to source some appropriate replacements and get it done.

A pic of the power board solder side and we could verify the PSU schematic.

Thanks for this, I have attached some more images, including one of the underside of the PSU.    I have to say, that arrangement of rectifier diodes and electrolytics has me scratching my head...
unfortunately I might be a tricky task for you to follow the schematic from the images, because they have packed the components so closely together, that the legs on the capacitors often dont go directly through the board below them.

I cant wait to hear your thoughts on the operation of this circuit!

[TheFloyd]

Oh WOW!   

I have found it!    https://dalmura.com.au/static/Kemp%20Counter-Timer%20Model%204DS10.pdf
This was what let me to finding this.  thanks!

An interesting mix of Australian and Japanese components. PCB marking "IU3 KEMP" but no leads.

Other than the fact that their device has a fourth DK24 in it, and mine has the motorised preset timer, they look extremely similar.

A couple of observations and questions:

- I noticed that I had drawn one of my rectifier diodes back to front in my schematic.    I am glad about this, as the way that I thought it was hooked up, had me completely bamboozled!  haha
- - That said, I am still less familiar with this approach to a rectifier design.   I would have expected to see either a bridge or a full-wave centre-tapped arrangement.    Is there a reason that this design was used for sensitive old valve gear like this ?

- There are one or two other wonky bits in my schematic, that I need to double check too.    while I am assuming that his is correct and my draft has initial errors, I will make sure.

- The big one is the fact that his picture shows that his 12AU7 is in the right most socket and not in the centre socket like mine.
- - I am not sure if it is his or mine that has the tube in the wrong socket.   It could have happened years ago during a service i guess.
- - I am under the impression that 12(AX/AT/AU)7 tubes are all electrically compatible, but have different gain and bias characteristics, which explains why both devices work (even if not ideally).
- - - Is there a clear way that I can identify which socket the 12AU7 should be in?    it is not marked on the board unfortunately.    I think the AU7 is the low gain tube, so wonder if that is typically used to drive the DK24's, and the AT7's used for the audio amplification and signal amplification from the detector tube ?

[TheFloyd]

Just spotted that the author of this amazing document is a member of this forum too.

a big thanks to trobbins

[TimFox]

The two diodes rectifier circuit is called a “full-wave voltage doubler” (q.v.)

[trobbins]

Good on ya Floyd for taking the time and effort to restore your timer. 

I found it an interesting instrument as it was my first foray into dekatrons and geiger counters.  I didn't have much in the way of emitters to test, and no nuclear wars recently, so not having an on-going application was a bit of a damper.  As such I didn't try and reverse engineer the design of valve stages, or find somewhat similar designs to work through, and performance (related to a triode model) may partly relate to the actual geiger tube in use, and to the dekatron model and associated passives for waveshape forming.  The triode for SE transformer driving may benefit from some design engineering if you have the same output transformer and can find a datasheet or measure its inductance at different dc currents to get a handle on what idle current to set, and what the default idle would be for say an AT7 or an AU7.

Perhaps worthwhile using a waveform generator to confirm dekatron operation, and perhaps give the beast some use for parties!

Ciao, Tim

[floobydust]

I tried to post but the website was choking today. That's a good find of Trobbins schematic and teardown.

The Dekatron carry-output needs amplification, 4 counter tubes so I would think three triodes to do that? And there are six. But why they are different I'm not sure. AU7 is usually driving a speaker but that output transformer is so tiny. How loud did it need to be.
Caution some components are only grounded through the PCB standoffs. I saw your jumpers there but the front panel I'm not sure how it gets circuit ground when disassembled. Careful it does not bite.
I think I see bead tantalum caps as well to check on the counter board.

Star Trek "Dagger of the Mind" it sorta looked like a Dekatron brainwashing Captain Kirk, with added sound effects and a white light.

[TimFox]

See pp 698-700 of Millman and Taub for Dekatron tubes.
Pirate copy: https://archive.org/details/MillmanTaub-PulseDigitalSwitchingWaveforms

[TheFloyd]

The two diodes rectifier circuit is called a “full-wave voltage doubler” (q.v.)

Awesome.  thanks for that and sorry if it was a silly question

I have experience with voltage multiplier circuits, but typically they have been limited to the ones that are a series multiple cascading stages of caps and diodes, never as part of the rectification step.

[TheFloyd]

Good on ya Floyd for taking the time and effort to restore your timer. 

I found it an interesting instrument as it was my first foray into dekatrons and geiger counters.  I didn't have much in the way of emitters to test, and no nuclear wars recently, so not having an on-going application was a bit of a damper.  As such I didn't try and reverse engineer the design of valve stages, or find somewhat similar designs to work through, and performance (related to a triode model) may partly relate to the actual geiger tube in use, and to the dekatron model and associated passives for waveshape forming.  The triode for SE transformer driving may benefit from some design engineering if you have the same output transformer and can find a datasheet or measure its inductance at different dc currents to get a handle on what idle current to set, and what the default idle would be for say an AT7 or an AU7.

Perhaps worthwhile using a waveform generator to confirm dekatron operation, and perhaps give the beast some use for parties!

Ciao, Tim

Thanks again for the work you did on your schematic and the way that you documented the process.
In the image below, if you counted the tubes from left to right: 1-2-3

on mine,
1 - is a 12AT7 that has one plate wired into the audio output transformer with its grid hooked to the 1M pot, just like how you have V3B in your schematic.
2 - on mine, this is the 12AU7A and it is the one that is driving the DK24's
3 - on mine this is the second 12AT7, and the pin1 plate is hooked to the outside lug of that 1M volume pot, via a 470p and 1M resistor.  This is wired exactly like your V1A, which is a 12AU7 in yours, only there are three 'afterthought' resistors in parallel that read as 120K, connecting the plate to VS3 in place of the 150K resistor in your schematic (that I cant even find on my board).

So, my theory is :
At some point, my device was serviced by someone, who also had no service manual or schematic.   
- When they put it back together, they mistakenly put the 12AU7 in spot 2, rather than spot three.
- They then did some measurements and found things to be a little bit off, which is why there are a number of 'newer' resistors in there, than you have in yours.
  - the 3 in parallel on the on the plate of tube 3, dropping from 150K to 120K
  - the 82K resistors that yours has in parallel with 47K resistors from K0 on the DK24's in your diagram, have been changed to 47K resistors, so the total resistance between K0 and GND is only 23K

nothing else immediately obvious stands out, so I am of the view that they modified the circuit values, to accommodate the fact that they put the tubes in the wrong spots, because my device seems to work perfectly (albeit with some dry old caps that are operating at far to high voltage!)

When i re-cap this, I think that I will put the tubes back in the correct spots and put the resistors back to how they should be. 

[TimFox]

The two diodes rectifier circuit is called a “full-wave voltage doubler” (q.v.)

Awesome.  thanks for that and sorry if it was a silly question

I have experience with voltage multiplier circuits, but typically they have been limited to the ones that are a series multiple cascading stages of caps and diodes, never as part of the rectification step.

The cascade circuit is a “half-wave” multiplier, and can expanded to triplers, quadruplers, etc.