-
Very nice construction, the PCB home made as well I suspect,
-
Beautifully implemented Enigma-man! I particularly like the use of smaller displays for the annunciators. And my type of 'fun' engineering too.
Yes, all the pre 106x Datrons used socketed Sperry / Beckman / Babcock/ SP3xx displays. The mV / V etc annunciators were done with dim lamps behind transparencies. They are also a 'micro free', no batteries, no eproms, just discrete analogue and socketed logic, making it possible to maintain them more or less indefinitely. They still achieve 0.01% class accuracy and stable 1uV resolution. Well worth grabbing if you see one cheap (avoid the 105x 5 1/2 digit series, it uses an unobtainium custom PMOS counter logic chip and a very touchy triple slope conversion circuit).
I think it was sad that adopted a custom panaplex display on the later models, but every major manufacturer did the same, but it is a definite weakness when it comes to restoring older units. Remember though that a lot of the 106x meters on the market have been left on 24/7 for decades.
@bitseeker:QuoteWhat's the sepia-colored staining on the PCB?
No idea, it's on the surface and definitely not heat related, my guess is either flux residue or lacquer. It's not sticky and I've never tried removing it. BTW did you notice the big beehive trimmer cap up next to the glass encapsulated crystal in the top right corner? That's the frequency trim.
QuoteI really like the look of the Panaplex displays, but don't have any yet. What's a reasonable expectation for lifetime? I've seen others with, for example, Datron equipment and the displays have gone bad. Are they better, worse, same as VFD?
Life is certainly way better than VFD - particularly due to absence of uneven brightness loss. The Sperry / Beckman spec states a blanket 10 years with a note that operating at Maximum rather than Typical ratings can reduce life. In practice it's difficult to get them dim enough, even at Minimum specified current (They were specified at 200ft lamberts (
) and a viewing distance of 40ft). I suspect that minimum current drive increases their life even further. Like all cold cathode tubes, the cathodes will splutter if overdriven.
As you say Cubdriver, they are more mechanically sensitive than anything, the front glass is bonded as are the pins, but I think they use glass 'frit' seal (as cerdip packages). There are also internal getters (the little ring structure at the bottom) to mop up gasses. Given that many of the ones still in use are 40+ years old, long term leakage doesn't seem to be an issue.
@Arhammon: Yes the Philips one is particularly nice, it's a multiplexed display (multiplexing does seem to look better, maybe because it drops the brightness below what can be achieved at minimum current in static drive). It looks a lot nicer than the LEDs they used on later counters.
P.S. I should probably qualify that multiplexed display comment. What sets the minimum current spec of the display is having sufficient current density to light up the entire surface of a segment. When multiplexed, the segments get driven at higher current, but for lower duty cycle. This seems to make them look dimmer and 'sharper'. -
Thanks for all the info, Gyro. The only regret will be my lighter wallet after having learned more about Panaplex displays.
-
At least you're in the US - you get much better pickings than we do!
http://www.ebay.co.uk/itm/Beckman-3-Digit-Display-Panel-Panaplex-Model-SP-353-/162079434878?hash=item25bcafe87e:g:LtEAAOSwkEVXGSDF#shpCntId -
That's true.
The pins are going every which way on the back of that display. Poor thing.
And that's how it starts, right Pat? (Well, that's one way of many.
)
-
Thanks for all the info, Gyro. The only regret will be my lighter wallet after having learned more about Panaplex displays.
Yup. I have a half-assed search for Datron stuff on the 'bay, I will doubtless now be looking more carefully.
-Pat -
One is glad to be of service.
Maybe I'll do a teardown sometime, if pressed (better let this thread get back to Nixies).
Chris. -
BUILDING A THERMIONIC TUBE TESTER (V , V , I curve tracer)
Whilst not a Nixie tube it is close, I have built and tested the uTracer3+ from foreverDOS. If you have a need for a thermionic tube tester I would recommend this as an interesting and high quality project, ultimately not cheap though.
It comes as a complete PCB and parts kit with an great manual and an excellent website. You need to provide the box, a 19V laptop power supply, the 'pegboard' sockets and the tube sockets (these can get quite expensive).
The uTracer is basically 3 power supplies (and Voltage samplers) in the one unit controlled by a serial port and a program on your computer,
One low voltage variable , modest current continuous supply for the heater circuit
Two high voltage (boost SMPSU) relatively lower current supplies for the Anode/Grid,
The tester does quite a clever trick in that in only actively tests for some milli-seconds at a time , so the SMPSU can supply up to a 200mA or so briefly up to 400V(from -50V if needed) !
Wiring the multiple sockets is a bit challenging but made a lot easier by a wire stripper.
Below are photos of the complete PCB, the inside wiring of the sockets and testing a 12BYA , note the hot filament and while HV is active a warning light is displayed.
http://dos4ever.com/uTracer3/uTracer3_pag0.html
Note No association other than happy customer!
-
Maybe I'll do a teardown sometime, if pressed (better let this thread get back to Nixies).
Could always start a "Let's see your Panaplex display equipment" thread. (Hint, hint, wink, wink, know what I mean, know what I mean?) -
Boys and girls, today's show and tell is the following:
Here is an HP 5221A four digit Nixie counter. It is in near pristene condition for a piece of equipment with date codes
of 1967 and 1968. Some IC's have unusual codes. My guess is that this unit is from 1968 or 1969. It is absolutely spotless
inside and out. All components are original. Not bad for 48 years of age. The logic board has the upside down Nixies and there
is provision for two more digits on the gold plated pc board. The IC's are non-standard decade counters, storage and drivers.
7400 series 7490, 7475 and 7441 or 74141 could be subbed but are not direct replacements and would require a piggyback board
if and when repair is needed.
Judging by the condition, this 5221A didn't see a lot of use and could sit in HP's museum if they already didn't have some.
The big and heavy piece of equipment which I thought was a frequency counter turned out to be a Gould K115 Logic Analyzer from around 1990.
It has a 5.25 inch floppy with a diskette with that supposedly has a Z80 disassembler on it if it can still be read.
One day I'll fire it up. No real use for it but it's in really good shape complete with cables.
From Dr.Dobbs concerning logic analyzers
"There's two problems with these big logic analyzers. First, the Gould machines weigh something like 80 pounds and take up roughly the
space of a midsize PC tower. You have to be pretty desperate — I mean motivated — to go drag one of these out and clear off a spot for it."
I'm not desperate or motivated... I got it for nuthin' and he ain't kiddin' 'bout the weight...
Back to the cutesy counter. Here's the pix.
Mr. Sparky provides some comic relief.
-
Very nice Enigma-man. Nixies and HP have got to be a winning (and collectable) combination.
I agree, it would be nice to see the other two digits populated, but I'd probably only be tempted to do it if you could find original pin compatible parts, it would lose some of its charm with a piggyback board. Hopefully a scrap donor unit might come along.
Upside down Nixies are a bit of a rarity but they are shown on Dieter's Nixie World site here...
http://www.tube-tester.com/sites/nixie/trade01-nixie-tubes.htm
It's a really nice site to go and browse, even if you don't have any of your own to play with. It also has useful Panaplex data too.
It looks as if your displays are Burroughs B-5560 from the description.
Oh, @bitseeker:Could always start a "Let's see your Panaplex display equipment" thread. (Hint, hint, wink, wink, know what I mean, know what I mean?)
Not a bad idea, give me a few days to take some pictures.
-
What type are normally those small nixies?
I like to use a few for a projects (hopefully there is a cheap Russian equivalent with a real five) -
@b_force:
The answer to your question is go to Dieter's Nixie World link above and you will see there is nothing really "normal" about Nixie tubes.
You will see different sizes and styles.
If I were you, I would decide on a style after seeing what is available for purchase and budget if you have one.
You may have to settle for what you can get because some in the list are rare. There appears to be a fairly good supply of Russian tubes.
@Gyro:
A piggyback board would only be temporary. I would use two machine sockets where the Nixie drivers would go and use a DIP header to attach the piggyback board.
The upside down Nixies may take a while to find. Most likely things will remain as they are for now. The service manual showed the extra digits as options.
If someday I do find a donor, the options will be done in a proper fashion.
pmanning at smugmug.com shows his nearly beat to death six digit 5221A as received and after cleaning etc. I like the one I have better...
https://pmanning.smugmug.com/Electronics/HP-5221A-Electronic-Counter/
-
Ah yes, I see. I was surprised that the footprints aren't industry standard but looking at Wikipedia, 54 series parts weren't around until 1964 and 74 series until 1966. They may be based on Sylvania (Edit: SUHL) parts (1963). That's quite a teardown pmanning did... 55 photos.
It's a bit surprised that anyone bothered to order the counter without the 6 digit option. The cost difference should have been minimal unless HP loaded the pricing. Datron offered an extra digit resolution on the 1030 RMS voltmeter as a no cost option but that was just an extra driver chip as the Panaplex display already included the digit (actually according to the assembly drawing, the chip got unplugged after factory test!). -
@Gyro:
All of pmanning`s restorations are like that, which is good for comparison. I suppose the best gauge for any attempt is opening up whatever it is and
see what`s there or missing, be brave and fire it up. If it doesn`t blow fuses or burst into flames you have reached the first step. How far one gets
depends on the "unobtanium factor".
In the 5221A there are decade counters that are in TO-5 cans listed as IC2, IC3 and IC8. They are TTL with HP numbering.
According to the Hewlett-Packard Journal 1967-08, the 5221A was introduced and was said to have readily available off-the-shelf Texas Instruments 74 series as
replacements. TI had their pinout, Motorola and others had their own. Not sure who won out but it was probably Motorola. TI went with the standard as I do have several
74 series TI chips.
The 5221A, four digit counter is listed at $475 US. There was also a time base upgrade option. I think I saw a price of $675 US for the six digit model with time base option, somewhere...
Looking at schematics from HP equipment in the early 70's and comparing with a Motorola TTL Databook one can determine the following:
1820-0055 is a 7490 decade counter
1820-0100 is a 7441 decoder/driver
1820-0707 is a 74141 decoder/driver (improved 7441)
1820-0301 is a 7475 four-bit latch
-
@ Enigma-man
Thanks.
I am already familiar with nixes. I have a bunch of clocks (developed) myself.
I was only wondering if people would directly know a specific type or so -
Wow, E-man -
That counter is in beautiful shape - looks like it came from us of the box yesterday. Mine looks like it went through the war in comparison (as you saw from the SmugMug pictures). I've gotten a replacement tube since the photos were taken, but haven't found the time to get it back on the bench to install it and for further surgery. It has some other issues, and I plan to try to clean it up more as well.
For what it's worth, I looked up the prices in the 1970 HP catalog:
5221A base model: $425.00
Option 001 - 1 add'l digit: $75.00
Option 002 - 2 add'l digits: $125.00
Option 003 - Xtal time base: $100.00
This all seems cheap to us now, but on a relatively inexpensive counter I suppose it was a pretty big nut at the time. It's also quite possible that many were using it literally as a counter (I think we're predisposed to always think 'frequency' here) and had applications that would never go over 9999 counts, thus not needing the extra digits at all.
-Pat -
Sorry Pat, I didn't realise it was your teardown!
They really did load the pricing for those options didn't they. -
Sorry Pat, I didn't realise it was your teardown!
They really did load the pricing for those options didn't they.
No worries! Between forum screen names, first names in signatures and other account names, the connections aren't always obvious. (And, I don't recall posting that counter in this thread.)
HP's stuff wasn't cheap, but if they manually reworked the boards to add the digits I could see it costing a pretty penny to retrofit the additional digits, even at the factory. The 'B' model of the counter, which as a base unit included the crystal time base, seemed to have been aimed more at the cheap (frugal?) frequency measurement crowd. It had 5 digits with an optional 6th, and IIRC more gating positions on the switch.
-Pat -
@Cubdriver:
I almost (not really) felt bad seeing the condition your 5221A was in when you received it and for that matter some other pieces you have.
Sometimes one gets a really good one, like me for instance...
Seriously, though I did feel bad and we learn something new everyday.
-
E-man,
That's a beautiful 5221A! It's the first time I've seen the A version with the default four digits. I started my Nixie adventure hunting for either a 5221B or its sibling the 5321B. The B version has five digits standard with an option for the sixth. The 5321 is a half-height (or thereabouts) and slightly longer/wider variation of the 5221. Both are nice desktop units.
During my hunt, I discovered and acquired a couple of HP 5216A counters. This model looks like a wider version of the 5221 with seven Nixie tubes. I guess I should post a few pics here since I haven't yet gotten around to tearing it down for its own thread. Both units are a bit flaky. I just hope it's not the ICs that are bad since HP built in additional decoder states to blank the left-most zero digits.
Pat,
It's nice to see your fully-populated 5221. I guess I'll be picking your brain when it comes time to work on my 5216. As for the dead tube, I have yet to run across any available replacements. If I do, I'll be sure to let you know.
Does the SmugMug gallery represent the current state of your 5221? Are you looking for other parts for it such as the logo, knobs, etc. to fully restore it? -
Bitseeker -
I've been on the lookout for a B model as well (it has the better time base, among other things), and would like to find a 5321 of either flavor just for the variety. No luck so far, but I did snag another 5221 from the 'bay a few months back. It is a base model with only four digits, and hasn't yet been wrung out or photographed. It was modified by a previous owner and sports dual banana jacks for the input rather than the BNC it initially came with, so it's not entirely original. It's also not nearly as abused as the poor one shown in the SmugMug pics was...
I did eventually find a replacement tube on the auction site; just haven't gotten to putting it in thus far. My best suggestion for finding a replacement is to set up a couple of ebay searches, one with its HP part number and another with its manufacturer's part number at least. Then be patient - they seem to turn up once in a blue moon. The Electro Instruments meter is presently occupying the bench, so it may be a bit before I get back to the 5221. If I should stumble on any replacement tubes in my travels I will point you towards them, or grab one for you if it's at a swap meet. It kills me when equipment like this is ruined so that people can make clocks when there are gazillions of NOS Russian nixies available that can be had inexpensively and without the murder of fine old test gear.
I'll be happy to help however I can when you get to the 5216. I'm on the lookout for one of those as well; eventually one will turn up.
Yes, the pictures in the gallery do show the present state of the counter. I'm going to revisit it and see what I can do to make it prettier, and would love to stumble upon a box of logo badges - they seem to be missing on a lot of instruments. I'm not sure if they fall/break off or if people are prying them off as keepsakes or something, but regardless of the reason it's annoying to me when they're missing! I do have some spare knobs from a swap meet buy a year or two ago, but need to excavate them from wherever they've gone off to to hide. My house eats things - this should come as no surprise to anyone who's seen the pictures of the bench area and the metro cart (wire shelving rack) with the test gear on it.
It's nice to see this thread still alive and kicking!
-Pat -
@Cubdriver:
I almost (not really) felt bad seeing the condition your 5221A was in when you received it and for that matter some other pieces you have.
Sometimes one gets a really good one, like me for instance...
Seriously, though I did feel bad and we learn something new everyday.
I only just now expanded that thumbnail. Somehow I doubt that little tid-bit of knowledge is even in the same GALAXY as E=mc^2!
Well, the condition winds up being whatever it is, and there are some things I've passed on due to the shape they're in, but in other cases depending on the rarity or cost, it's worth the roll of the dice to try to snag and restore something.
-Pat -
Pat,
I've seen a few 5321 go by a couple-three months ago. This B model was a bit more than I wanted to pay and was missing the logo (yeah, what's up with so much HP gear missing that logo?). Earlier, this A model I saw was complete and cosmetically good. The "blows fuses" issue didn't bother me, but it had the basic four digits and one or two tubes looked like they might be blown. So, I passed on it too.
I'm not actively looking for one, so there's no worry about bidding against each other if you do see one. Once you get one, though, anything's possible.
-
Before getting off the topic of the 5221A, I thought it might be a good idea to extend the range to above 100MHz by connecting a homebrew
ECL divide by ten prescaler. I have ten 10138 counters that can be wired for divide by 2, five or ten. It would function as an inline device so the
5221A would remain unmodified. A crystal oscillator as a timebase would be better than the AC line frequency divided by sixty as is now the case.