Nuclear Measurements Corp. PC‑4 Proportional Counter

This time I am work­ing on anoth­er radi­a­tion scaler/counter which is quite a bit dif­fer­ent than the last one. This one is a decade new­er from 1975 and was built by NMC (Nuclear Measurements Corp.). It has a built-in pro­por­tion­al counter, but has exter­nal inputs for Geiger-Mueller, Scintillation, and oth­er sensors.

This unit has eight RCA DR2000 Numitron tubes that are used for count and time dis­play sec­tions. This is the first piece of elec­tron­ics that uses Numitron tubes that I have per­son­al­ly seen or owned, so was quite exit­ed when I dis­cov­ered them.
The sales pic­tures did­n’t direct­ly show the dis­play, but with some pho­to enhance­ments I could tell that there were tube dis­plays which I assumed would be nix­ie’s.
My orig­i­nal thought in acquir­ing this unit was that it would be inter­est­ing to also have a Nixie tube ver­sion of a scaler with sev­en decades of dis­play vers­es the five count decades of the Picker Dekatron scaler.

Numitron dis­play tubes are now obso­lete but were used in the first half of the 1970’s until LED sev­en seg­ment dis­plays became pop­u­lar and afford­able. The Numitron dis­play uses incan­des­cent fil­a­ments that oper­ate from 3.5 to 5 volts arranged in a sev­en seg­ment 8 pat­tern, and some tubes includ­ed a dec­i­mal point. The fil­a­ment cur­rent for each of the seg­ments is 24 mA @ 4.5 volts, with a total cur­rent 168 mA with all seg­ments ener­gized dis­play­ing the num­ber 8.
The RCA spec­i­fied “Mean Life Expectancy” at 95% con­fi­dence for the dis­plays is 100,000 hours.

RCA mar­ket­ed a BCD-to-7-seg­ment dis­play dri­ver CD2501E that could direct­ly dri­ve the fil­a­ments of the Numitron dis­play. Then by adding a Texas Instruments SN7490 decade counter IC you have a decade counter.
NMC decid­ed to use a 7446 BCD to 7 seg­ment decoder/driver for some rea­son, and then cas­cad­ed them togeth­er to build the 7 decade “Count” display.

The NMC PC‑4 includes a high-volt­age adjustable pow­er sup­ply with a volt­age range of 250 to 1,750 volts DC for pow­er­ing the built in, or exter­nal detec­tors. There is a sin­gle adjust­ment knob on the front pan­el with a scale num­bered 0 thru 9.
There are also three knobs for set­ting the pre­set time for the count mea­sure­ment with a range of 3 sec­onds to 1,920 min­utes (32 hours). Each time selec­tor mul­ti­plies the oth­er two time selec­tors.
There is a main pow­er switch and Neon indi­ca­tor, and a sep­a­rate H/V (High-Voltage) pow­er switch and indi­ca­tor.
The momen­tary L/T (Light Test) push­but­ton ener­gizes all seg­ments of all eight Numitron dis­play tubes to ver­i­fy prop­er oper­a­tion of seg­ments.
The momen­tary Purge push­but­ton ini­ti­ates a gas purge cycle and then enables the count and time cir­cuit­ry.
The length of the gas purge is con­trolled by the “Purge Time” rotary knob and has set­tings for Off, 10, 30, 100 sec­onds, and con­stant On.
The M — A switch selects either Manu­al or Auto­mat­ic start of the count cycle.

The N — C — W three-posi­tion switch is a bit of a mys­tery.
The N (Normal?) set­ting blanks the time dur­ing the purge cycle, then starts the time and count after purge com­ple­tion.
The C (Continue Continuous?) set­ting con­tin­ues count­ing time dur­ing the purge cycle, then resets time after the purge ends and begins event count­ing.
The W (Wait?) set­ting paus­es both the time and event counting.

Another front pan­el con­trol is the “Input Mode”:
T — incre­ments the mea­sure­ment count from the sys­tem clock pulse.
P — selects the inter­nal pro­por­tion­al counter detec­tor.
A — selects the rear pan­el MS3106A18-8S eight pin con­nec­tor.
GM — selects the rear pan­el PL-259 coax­i­al con­nec­tor for use with a Geiger–Müller or Scintillation detector.

There are no high-volt­age switch­es involved with the mode selec­tor, only sig­nal switch­ing, so care needs to be tak­en as high-volt­age pow­er is present on all con­nect­ed devices and unused con­nec­tors when enabled.
There is a small shaft with a slot labeled B/R locat­ed just above and right of the “Input Mode” switch, this adjusts the bub­ble-rate of the purge cycle which can be mea­sured in the bub­ble-counter win­dow just above the pro­por­tion­al detec­tor draw­er. There is also anoth­er flow adjust­ment screw locat­ed on the rear of the sole­noid valve.

That pret­ty well sums up the tour of oper­a­tional con­trols and fea­tures of this unit, I guess it’s time to see what actu­al­ly works and what needs to be repaired. None of the sales pho­tos showed the unit actu­al­ly pow­ered on. This can be either good or bad in that they real­ized it might not be safe to plug it in just for a pho­to, or they did plug it in and the unit was non-work­ing, blew a fuse, or released some mag­ic smoke.
For me as always a thor­ough visu­al inspec­tion was per­formed, then a check of the AC cord, fuse, and trans­former pri­ma­ry. For old­er equip­ment an impor­tant thing to check before apply­ing pow­er are the capacitors.

The unit had one slight­ly bulging elec­trolyt­ic capac­i­tor on the elec­tron­ics board, which when removed test­ed at half the list­ed capac­i­tance. There were three more radi­al capac­i­tors of the same val­ue and type, so all four were replaced. All the tan­ta­lum and even the chas­sis mount alu­minum can capac­i­tors were in good con­di­tion for being 50 years old.

The main cir­cuit board includ­ed test jacks for the two pow­er rails so I con­nect­ed two mul­ti­me­ters to the red and brown test point jacks and chas­sis ground before the ini­tial pow­er-up. I was fair­ly cer­tain that the brown test point was the +5 volts as it was con­nect­ed to all of the 74XX series IC’s. The red test point volt­age was an unknown volt­age, but was fil­tered with a 35 volt rat­ed capac­i­tor, so I had an edu­cat­ed guess of what the volt­age range would be.
The unit pow­ered-up with no issues and the 5 volt rail was mea­sured at 5.05 volts DC, and the red test point was 23.8 volts DC. I was also pre­sent­ed with my first view of an oper­a­tional Numitron dis­play as the time count­ed up from 0 to 9 which then pre­sent­ed me with a loud bang.
Luckily it was only the gas valve sole­noid slam­ming shut at the end of the count cycle. But that did get my heart rac­ing a lit­tle bit more than just the excite­ment of pow­er­ing the unit up for the first time.

Now that I had ver­i­fied most of the basic func­tions were work­ing and many hours of clean­ing. I decid­ed to take a clos­er look at the 2π pro­por­tion­al detec­tor.
I had nev­er seen one of these before so was very curi­ous as to how it was designed and worked. On this unit the sam­ple under test is placed in the cylin­dri­cal depres­sion of the draw­er, the draw­er is closed, and lever is then rotat­ed 180 degrees to seal the sam­ple con­tain­er to the detector.

The detec­tor is com­prised of a hemi­spher­i­cal cham­ber with an insu­lat­ed anode at it’s top cen­ter. There are two small holes for the purge gas to flow thru the cham­ber and the sam­ple is placed on the flat base of the cham­ber. If you look very close you can see the anode ring wire attached to the cen­ter post.

While wait­ing on some parts to arrive I decid­ed to reverse engi­neer the pre­am­pli­fi­er that is built into the detector.

The cir­cuit­ry did­n’t seem too com­pli­cat­ed and only had four wires attached. Black is ground, Black/green/white is the 10.6 volt sup­ply, Black/yellow/white is sig­nal out, and Purple is the high-volt­age bias. The com­po­nents are mount­ed on a dough­nut shaped cir­cuit board sur­round­ing the cen­ter insu­la­tor and cham­ber anode.
The most inter­est­ing com­po­nent was a 3N203 dual-gate N‑Channel MOSFET with four leads, which I have nev­er come across before.

I did check all the com­po­nents which were well with­in spec­i­fi­ca­tions, but I do not have any way of test­ing the detec­tor. This type of detec­tor requires a very spe­cif­ic mix­ture of purge gas to oper­ate in the pro­por­tion­al detect­ing range. The required gas is P10 which is a mix­ture of 90% Argon and 10% Methane. A quick check of my local sup­pli­er and the cost is way out of my lim­it­ed bud­get.
I do have a home­made scin­til­la­tion detec­tor that works very well using the GM mode and the rear pan­el PL-259 coax­i­al connector.

The entire unit was very dirty and required a con­sid­er­able amount clean­ing. The top cov­er had tak­en the brunt of it’s past abuse and need­ed to be cleaned, sand­ed, and re-paint­ed to be pre­sentable again.
After the ini­tial sand­ing and cleanup I end­ed up apply­ing mul­ti­ple coats of a filler and sand-able primer to fill the deep scratch­es in the alu­minum cov­er. The cov­er was then paint­ed using sev­er­al coats of Krylon col­or­maxx satin Pistachio paint.
The paint was a great match for the even lighter shade of green of the front pan­el and dark­er green knobs in my opinion.

After using the unit for a while I decid­ed that it was worth doing some small upgrades and improve­ments. The reg­u­lat­ed 5 volt sup­ply uses an unreg­u­lat­ed sup­ply of 9.2 volts using a 10,000 uF at 10 volt capac­i­tor, which is just a bit too close to the mar­gins for me. It was replaced with a high qual­i­ty Nichicon long-life capac­i­tor rat­ed at 16 VDC which I was able to mount inside the chas­sis leav­ing the orig­i­nal can in place,
The next item was replac­ing the 10 volt Zener diode and Tantalum capac­i­tor that sup­plies reg­u­lat­ed 10 volts to the pro­por­tion­al counter pre-ampli­fi­er board. And while I was at it all the Tantalum capac­i­tors were replaced, along with an elec­trolyt­ic capac­i­tor in the high-volt­age pow­er supply.

In my research of Numitron tubes I found that RCA offered a tube shield that elim­i­nat­ed reflec­tions from adja­cent tubes. It seems that the high­ly reflec­tive glass tube could pro­duce image arti­facts from adja­cent tube fil­a­ments reduc­ing vis­i­bil­i­ty.
I put togeth­er a 3D CAD draw­ing from mea­sure­ments of the units dis­play, and after a test pro­to­type I was ready to print a shield using some com­pos­ite poly­car­bon­ate / car­bon-fiber fil­a­ment (PC-CF).

Another item that I decid­ed could ben­e­fit from and upgrade were the 7446 BCD to 7 seg­ment decoder/driver IC’s. These IC’s were get­ting quite warm, greater than 50 degrees C when the low­er dis­plays were chang­ing rapid­ly. I had some DIP style heat-sinks and adhe­sive ther­mal mate­r­i­al, so all eight 7446 pack­ages were upgrad­ed which should increase their life expectancy.

The PC‑4 has been a fun project to work on, and I have learned a lot dur­ing the refur­bish­ment, but it is one of the few projects that is for sale to some­one that can uti­lize it bet­ter than me.