Looking at options for test boxes.

Discussion in 'Armory - Q&A' started by ktinoue3, Jan 16, 2017.

Tags:
  1. Mergs

    Mergs Podium

    Joined:
    Nov 8, 2000
    Messages:
    3,608
    Likes Received:
    249
    Actually, I find the A - C short, along with the B - C (foil/saber test), useful in detecting grounding issues with epee's.
     
  2. neevel

    neevel Armorer

    Joined:
    Jan 17, 2000
    Messages:
    3,463
    Likes Received:
    345
    It is possible to get Nixie tubes for much cheaper if you are willing to have them arrive in a box with a faded label in Russian that says 'CCCP' :). However, expensive artisinal hand-made Czech ones are far more in keeping with the hipster aesthetic. And you can use an Arduino to drive them.
     
  3. monitorlizerd

    monitorlizerd DE Bracket

    Joined:
    Aug 5, 2007
    Messages:
    414
    Likes Received:
    36
    *feverishly reading Wikipedia*-oh, THAT'S what a nixie tube is...
    I may be the only armorer in existence with no electronics background.....
     
  4. dcchew

    dcchew Podium

    Joined:
    Jan 19, 2016
    Messages:
    784
    Likes Received:
    183
    Nope, you're not alone. I'm a mechanical engineer and anything other a relay or mechanical switch is beyond me. I'm thinking I need to catch up the younger generation and start playing with Ardunio and software programming. The last software packages I used was MS Basic, Fortran, and Cobol. Yes, I'm an old geezer.
     
  5. ktinoue3

    ktinoue3 Podium

    Joined:
    Jan 14, 2014
    Messages:
    217
    Likes Received:
    20
    Thank you guys for all the information and feedback. Taking a bit of time to digest it all. I will be talking to our regional Armour next weekend at the Battle in Seattle and will look at building my own. I am a programmer so I might take some time to play with an Arduino or Raspberry pi. Should be an interesting learning experience.
     
  6. dcchew

    dcchew Podium

    Joined:
    Jan 19, 2016
    Messages:
    784
    Likes Received:
    183
    A bit of trivia for the younger generation, TV's used to have vacuum bulbs inside of them which is similar to the Nixie tubes. They require a high voltage to operate them. When my parent's TV would crap out, I'd have to remove the back cover of the TV and take all of the bulbs out. Then, I would go down to Radioshack or some TV repair place to check each tube out to find out which one was burnt out.

    I used to wear leather gloves to remove and install the tubes. You didn't want to touch a charged capacitor or the big tube. You would get a real jolt if you shorted yourself to one of them. The tubes also got very hot to the touch. An old electrical technician taught me to clean the tube surface with alcohol before installing them to remove fingerprints.

    Those old TVs would heat a small room. Don't ask me how I know.
     
    ktinoue3 likes this.
  7. St.Meow

    St.Meow Made the Cut

    Joined:
    May 22, 2015
    Messages:
    74
    Likes Received:
    7
    The Arduino one is a topic me, Brian, and Mergs have been thinking about and worked a little on while at Columbus. It works, but it may encounter noise problems. I don't know if Mergs finished his prototype for it.
     
  8. monitorlizerd

    monitorlizerd DE Bracket

    Joined:
    Aug 5, 2007
    Messages:
    414
    Likes Received:
    36
    And rotary dial telephones, rabbit ear antennas for TVs, and party lines!!!
    You clean off the fingerprints because they caused "hot spots" on the glass when the tube was in use, and caused early burnout. Saw a lamp in a theater spotlight blow up once- they found several pieces of glass with fingerprints burned into the glass. The lighting crew pitched holy fits, the head tech said he would be the only one to change the lamps from then on. He wore white cotton gloves to do it, and they were all polished to a high shine before installation.
     
  9. mtwieg2

    mtwieg2 DE Bracket

    Joined:
    Nov 24, 2015
    Messages:
    346
    Likes Received:
    71
    I'd like to see a test box with Kelvin sense connections to compensate for resistance in its connectors. You'd probably save decent amount of body cords from failures that way.
     
  10. brtech

    brtech Podium

    Joined:
    Nov 4, 2003
    Messages:
    2,425
    Likes Received:
    202
    You must be a young-un if you had a Radioshack with a tube tester when you were a teen ager. Tubes were sold in lots of odd stores, and many of the places that sold tubes had tube testers. The drug store we went to had one. They were free standing kiosk-like affairs, with rows of sockets and lights. I had a TV repair shop a bike ride away from my house who would test tubes for me. You could often tell the bad ones by looking for the blown filaments (no light at all), or leaky glass or blown plates (glowing red or blue).

    Fingerprints on the TV tubes wasn't really much of a problem. Now, on projection lamps, it's huge, and even today, you don't want to leave fingerprints on projector lamps or even the smaller halogen bulbs. But getting zapped by the anode on the Cathode Ray Tube would really hurt, and could kill you. You learn to discharge the CRT anode circuit before working on a TV, as well as the big capacitors that were the filters for the main power supply, which usually would pump out 200V or more for the anode supply on the other tubes.

    Now, fixing the radio or TV by replacing the tubes was kinda normal stuff we all did. Building Heathkits with tubes was a whole nother level; I did the "Vacuum Tube Volt Meter" (IM-13 VTVM). Ha! It even had an Rx1 with 10 ohm center scale!!! Some of us built ham radio stuff out of WW2 surplus tube electronics. I really did get into transistors before I left high school, but was VERY impressed with the kid who became my best friend in college, who had built the Heathkit Color TV set with his dad.

    Dennis, what version of FORTRAN did you learn first? FORTRAN-II on an IBM1620 for me. Card reader/punch only, no disc. Full box of FORTRAN pass 1 + your source code. Punched out an intermediate deck. Then load the FORTRAN pass 2 box and your intermediate. Punched out a binary. The loader was only a couple inches of cards, plus your binary. Your error was printed on the chain printer, and then back to the keypunch to fix it. Then FORTRAN pass 1 ... But I got to do that in High School, which was pretty much unheard of at the time. Computers were only in Universities, but a teacher in high school had a friend in the University computer center and we got access to that 1620 for an hour or two a week.

    Now, Arduino's ROCK, but making a box with one isn't trivial. Much easier to build a one meter test box. The next step up is to build your own ohmeter into that box. If you can get that to work, THEN maybe take a whirl at an Arduino based box, or step up to a 3 meter box with shorts testing.
     
  11. Mergs

    Mergs Podium

    Joined:
    Nov 8, 2000
    Messages:
    3,608
    Likes Received:
    249
    Well, I was going to spend some time on it last weekend, but the parts didn't make it back from DC in time. Still on my to do list - along with a million other things. Hope to have something we can play with in Cleveland.
     
  12. brtech

    brtech Podium

    Joined:
    Nov 4, 2003
    Messages:
    2,425
    Likes Received:
    202
    I've spent some time on this. It's not worth it.

    To get the basic 10 ohm center scale with a 1.5V battery, you run 75 ma through the circuit, with a total of 10 ohms plus the cord. That's 20 ohms total (for 10 ohm center scale), I=E/R, E=1.5, R=20, 1.5/20 = .075. With zero ohms in the cord, you would only have that intrinsic 10 ohms in the circuit, and you would be pushing 150 ma through the circuit (1.5/10). Those are not random values, You want to be able to have 10 ohms center scale and zero ohms full scale, and 150 ma with 1.5V battery does that. You can make a 5 ohm center scale, with a 5 ohm total circuit, 10 ohm total resistance at center scale, 150 ma center scale, 300 ma full scale (1.5/5). 300 ma will empty a battery pretty quick.

    The circuit to do 10 ohm center scale has a 1 or 2 ohm fixed resistor, across which you have a fixed or variable resistor and your panel meter in a current divider. Then you have another fixed or variable resistor, and a zeroing pot; typically 5 ohms, in series with the current divider. The sum of the 1 or 2 ohm resistor, plus the fixed or variable resistor, plus the zeroing pot, plus the connector resistance, adds up to nominal 10 ohms. It's nominal because it depends on the actual battery voltage and the zeroing pot setting. You adjust it with a short across the measuring points to get 150 ma through the main circuit. With a new, 1.6V battery, it will be more than 10 ohms and with a 1.4V, close to replacement time battery, it will be less than 10 ohms.

    Suppose you had .5 ohms of connector and wire resistance. That's high; a good tester should be under .1 ohm, but suppose it was .5 ohms. What would happen is that you have to turn up the zeroing pot a bit to get 150 ma through the main circuit. The zero pot cancels out any resistance due to the connectors. You need a good deal more than that .5 ohms for the circuit to work, and it doesn't matter where it comes from. You need the zero pot because the battery voltage isn't constant, but it does double duty as a way to cancel out any resistance in the rest of the circuit. Or, if you like, that fixed or variable resistor in the middle (after the current divider and before the zero pot) can be smaller by the amount of connector resistance.

    I usually make my boxes with a variable 10 ohm resistor. I set the zero pot to about 4/5ths with a new battery, put a low-series-resistance current meter across the test terminals and adjust the variable resistor for 150 ma. Then I adjust the current divider to get full scale on the panel meter. There is some complexity here - the panel meter does have some series resistance and adjusting the current divider changes the resistance seen by the main circuit. The current meter has real series resistance. I null all that out by switching between an accurate 10 ohm resistor and a short circuit to get full scale with the short circuit and half scale with 10 ohms - without the current meter. The current meter test, and the full scale adjust gets you close, and then playing with zero and 10 ohms gets it dialed in right.

    But doing a Whetstone bridge with Kelvin sense wouldn't help.
     
  13. dcchew

    dcchew Podium

    Joined:
    Jan 19, 2016
    Messages:
    784
    Likes Received:
    183
    I grew up in what I called a "1/2 horse town" which means it was very small. I do remember the old tube testers in stores. I used to use my Dad's old cigar boxes to carry the tubes in. Good God, that was a long time ago!

    My 1st computer course was in my sophomore year of college (1972). We used punch cards and was only allowed to submit the card deck once in the morning and once in late evening. God help you if you couldn't type error free.

    Another piece of trivia! You could carry exactly 100 computer cards in each breast pocket of a Vietnam era fatigue jacket. Don't ask me how I know.
     
  14. Purple Fencer

    Purple Fencer Podium

    Joined:
    Jan 19, 2001
    Messages:
    16,577
    Likes Received:
    621

    Aaaaaaand since no one else has said it...

    Which half??
     
  15. dcchew

    dcchew Podium

    Joined:
    Jan 19, 2016
    Messages:
    784
    Likes Received:
    183
    It depends on your perspective. Some people love living in a small dusty little town in New Mexico. But for me at the ripe old age of 17, I couldn't wait to get out of there.

    So when my parents gave me the option of going to college at the University of New Mexico in Albuquerque or going to San Francisco State, I was gone in a flash. Remember this was 1970 and things were really happening in the City by the Bay.
     
  16. mtwieg2

    mtwieg2 DE Bracket

    Joined:
    Nov 24, 2015
    Messages:
    346
    Likes Received:
    71
    I'm not sure how you come to this conclusion, since none of what you described corresponds to kelvin sensing. It would completely eliminate any need for all that manual calibration and ranging.
     
  17. brtech

    brtech Podium

    Joined:
    Nov 4, 2003
    Messages:
    2,425
    Likes Received:
    202
    No, it wouldn't. You need all that manual calibration and ranging because the battery doesn't maintain a constant voltage, not all meters have the same internal resistance, and it's hard to measure current at these voltages and resistances without the meter series resistance affecting the reading. Contact resistance is one of several variables that have the net effect of needing a variable resistance in more than one place. Resistance is resistance. If you get rid of one source, you still have all the rest of the sources.

    Suppose I used all gold contacts, 12 gauge wire and excellent soldering, making the sum of all the wiring and contacts down in the .05 ohm range. What of my above design or process would change? I think nothing. Now make the contact resistance 1.2 ohms. What changes? Only the settings, not the schematic or the calibration procedure.
     
  18. mtwieg2

    mtwieg2 DE Bracket

    Joined:
    Nov 24, 2015
    Messages:
    346
    Likes Received:
    71
    I guess if you limit yourself to batteries and standard meters then you have a point. I was envisioning regulated current sources with simple voltage readouts.

    Sure the internal wiring resistance could be very negligible, but contact resistance is the real issue addressed by kelvin sensing (gold plating wouldn't even last one NAC's worth of testing). On the other hand it would also mostly ignore the contact resistance of the circuit under test, which you may actually want to measure (debatable).
     
  19. brtech

    brtech Podium

    Joined:
    Nov 4, 2003
    Messages:
    2,425
    Likes Received:
    202
    If you are going to use a constant current supply, and voltage measurements, then I guess, sure. Actually, if you built a regulated voltage source, you could eliminate the external zero adjust for current measurements. I've considered that. It would be good to go down to maybe .5 V because you could lower the current you need to get a 10 ohm center scale. But then having an internal trimmer to set the full scale current would still be much more attractive to me then Kelvin sense, and building a regulated .5 V voltage source is a PITA. Both those get into the complexity range of a microprocessor based design, so I'd go there first. But the basic battery voltage source and measure current circuit that powers most existing test boxes is very simple for anyone to build.
     
  20. St.Meow

    St.Meow Made the Cut

    Joined:
    May 22, 2015
    Messages:
    74
    Likes Received:
    7
    I was considering this when designing mine, because there are a few easily found DC voltage regulators, but I also wasn't sure if they would decrease battery life unless I put in a switch to disconnect the battery. I just relegated to using a variable resistor to compensate.
     

Share This Page