Monday, December 14, 2009

Converting an HP Counter into a Nixie Tube Clock

Hmmm...I had an old HP 5233L nixie tube counter with 6 nixie tube digits gathering dust in the corner. What to do with it...?

Let's see, six digits...why not make it into a clock? After all, nixie tube clocks are pretty cool. And the counter came integrated with almost everything I needed: power supply and even a case with nicely milled holes. Mechanical work would be minimal, which defines my ideal project!

And as a nice bonus, there was also a back-panel BNC input for an external frequency reference, so I could use my GPS-locked frequency standard to keep the clock's time from "drifting" over time.

Conceptually, I figured it should look something like this (this includes a simple method for setting time):

Block Diagram


(Click on image to enlarge)


Pretty straight forward. But there was a small complication: I didn't have schematics for the 5233L.

This really didn't prove to be much of a problem. HP used the same nixie "single-digit" decimal counter plug-in module in a variety of its counter products. And I did have a manual for an HP 5232A. A quick glance revealed that it had the all-important schematic for the decimal counter module (HP part number 5212A-4A -- and although my counter used 5212L-4A modules, the 5212L-4A design seemed to be very close (if not identical) to that of the 5212A-4A)).

That manual, plus some scope probing of various signals within my 5233L counter, told me pretty much all I needed to know.

Some things I needed to do were:
  1. Change two of the counter modules to count from 0 to 5 (instead of 0 to 9), so that minutes and seconds would each count from 0 to 59, instead of 0 to 99.
  2. Add a reset circuit to the two "hours" digits such that, if the hour count increments to '13', it immediately resets the two hour digits.
  3. Modify the Hour LS (Least Significant) digit to reset to '1' instead of '0' (so that hours count from '1' to '12'.
  4. Disable the "Storage" feature of the Decimal Counter modules so that we can see the module counting.
So first, to count from 0 to 5, the 5212L-4A assembly(or 5212A-4a, or other variants) can be modified as follows:
  1. Remove R45, R50, and C10
  2. Change C11 from 200 to 470 pF
  3. Move R59 from CR12 to CR11
If you're trying to understand how the decimal counter circuit works from its schematic, note that it counts per the following pattern:
D C B A
0 0 0 0 (0)
0 0 0 1 (1)
0 0 1 0 (2)
0 0 1 1 (3)
0 1 1 0 (4)
0 1 1 1 (5)
1 1 0 0 (6)
1 1 0 1 (7)
1 1 1 0 (8)
1 1 1 1 (9)
So, to count from 0 to 5 instead of from 0 to 9, we can use the transition of B from 1 to 0 to set C to 0. And D must be always kept at 0.

(Important schematic note: Q1/Q2 control bit A, Q3/Q4 control bit B, Q5/Q6 control bit D (not C!), and Q7/Q8 control bit C (not D!)).


For item 2, the Hours Reset circuit needs a few more parts. I mounted them under the counter chassis.

(Click on image to enlarge)

For item 3 -- to modify the HOURS LS digit decimal counter module (HP Assembly 5212(x)-4A) to reset to a count of '1' instead of '0', simply modify that module so that R23 connects to the base of Q2 instead of to the base of Q1.

And finally, for item 4:

The "Transfer Line" signal to the counter modules controls the modules "storage" function (necessary for a counter to maintain a stable display while the modules are counting). But there's no reason to use a storage function when operating as a clock -- we can simply view the count while it's incrementing, and it makes the modification simpler

To disable the "Transfer Line", I cut the wire from the driver that drove pin 5 of the Decimal Counter card-edge connectors. This line will then float at about-12V, and the storage feature of the Decimal Counter modules will be disabled).

So...that's essentially it!

[I did add various switches and buttons to give me some functions that I wanted (such as turning off the display, but continue running the clock, to save on "wear and tear" of the nixies). But the modifications listed above are the main mods to the basic counter function.]

Here's the top view of the modified counter...

...with the modified decimal-counter modules:

And a view under the chassis. The HOURS reset circuit is wired between two of the edge connectors towards the right.


The counter, before I made an overlay for the front panel...


...and after:

(Martini time!)


From left to right, front panel controls are:
  1. Toggle Switch: Mode, RUN/SET-Time
  2. Push Button: Increment clock when in SET-Time mode
  3. Toggle Switch: FAST/SLOW increment speed when in SET-Time mode
  4. Rotary Switch, 3 Position. Selects Display Format: H:M:S, H:M, Display OFF (but clock still counting)
  5. Push Button, Reset SECONDS count
  6. Pot with Power-On switch. Only the power-on switch is used.

Other Notes:


1. HP Manuals are great. I've always found their "Principles of Operation" section to be an excellent description of how their equipment operates -- very useful and well worth checking out.

2. To use a 5212(x)-4A module in slot A19 of the 5233L counter, clip wire going to pin 8 of that slot's card-edge connector. (I no longer recall which module was originally in this slot).

3. You can use the "decimal point" neon lamps in the modules as separators to differentiate between hours, minutes, and seconds digits (refer to photo of my counter).

[Note: the 5212L-4A modules in the 5233L counter do not have neon lamp decimal points within the modules themselves (unlike the 5212A-4A module). Instead, the neon lamps are mounted on a separate small PCB that runs underneath the modules, next to the front panel.]

4. You could probably make this a 24-hour clock as follows:
  • Do not modify the Hours LS digit to reset to 1.
  • Change the Hours Reset Circuit to use the following 2 bits instead of the 3 bits shown in my schematic above:
  • Hours LS Digit pin 13
  • Hours MS Digit pin 9
(I haven't tried this, so take this recommendation with a grain of salt.)

5. Pin 7 is the Clock input pin for a 5212(x)-4A module. It is to this pin of the first module (the least-significant digit of the "seconds" modules) that you connect the 1 Hz time reference (and the faster clocks, when in SET-Time mode).

6. In an HP counter you should find a number of identical divide-by-10 modules that are used to divide down the time-base reference frequency (HP p/n 5212A-65C). It is on the card-edge connectors of these modules that you can find the 1Hz, 100Hz, and 1KHz signals (as well as others, should you need them).

Note 1: These boards have pin 7 as their input and pin 5 as their divide-by-10 outputs.

Note 2: Although my block-diagram above has 1 KHz as the "fast" set-time frequency, I might have actually 10 KHz.

7. After I'd mucked around with making an overlay and installing it on the front panel, the counter started "double-counting" (it would increment twice in one second -- on the rising and the falling edges of the 1 Hz clock). It turns out that CR9 on the first nixie tube module had opened up. HP's p/n for this part is 1910-0015, but unfortunately it didn't list the manufacturer's part number. I replaced it with a 1N4148 -- and although this new part is silicon, rather than germanium of the original diode, it fixed the problem.

8. A note regarding my "Reset Seconds" pushbutton. This button simply zeroes out the MS and LS seconds digit (and it also keeps the other digits from counting). But if one does this when the seconds count is greater than '19', the minutes count will increment by one. This isn't a big deal for me, and actually, it's kind of useful when setting time. Here's what I do:
  1. I'll run up the time to the current time (per WWV), making sure that there are at least 19 seconds on the clock.
  2. Then I'll flip the RUN switch to RUN and depress and hold the Reset Seconds button. This will set the clock to the next minute and zero the seconds.
  3. I then wait, with button depressed, until WWV hits the minute mark before releasing the button.
  4. The clock is now within a fraction of a second of WWV!

9. The +20VDC rail within my counter was way off (it read +31 VDC). Some probing revealed that a couple of transistors, as well as a zener diode, in the voltage regulator circuit had blown. The two transistors were both germanium PNP parts (one was an HP 1850-0062, which crosses to a 2N404A, and the other was an 1850-0105, for which I cannot find a cross reference). I replaced these both with 2N3905 transistors (PNP silicon). I replaced the blown zener with a 6.8 volt one, and, when finished, the voltage read +19.98 VDC. The voltages across the 2N3905 transistors are well within their range, and they aren't getting warm, so I believe everything should be copacetic.

Final note
...this posting was written not to give detailed instructions for modifying an HP counter to be a clock, but rather to generate ideas and inspiration. If you have an old HP counter kicking around somewhere, consider giving it a try!

- Jeff, K6JCA

1 comment:

Ross Miller said...

I just got one of these in the mail today.. And it is in nearly mint condition. If it wasn't so nice, and if I wasn't going to try to use it, I would try this... What a great idea!