Labor savin' devices

I've seen some crimpers with almost polished sides to the crimp dies -- which makes things pull out much more easily. This applies to the QuickSilver radio kit crimper -- but the shape makes things more difficult.
But the crimpers for the Anderson PowerPole have a rather deep notch, sides about 7 to 10 degrees, and for some of them, a bottom sort of like this:
) )
to take two "flag" sides and curl them back to point into the stranded wire being crimped into the terminal.
This in particular is in the 45A slot -- which has a shorter length being crimped. The 15 and 30 A slots are for barrel crimping instead of flag -- different terminals but still fit in the same housings.
The 45A slot is the closest to the hinges, so the best lever action.
There is also a 60A slot (larger connector) which has a slot which looks like an Acme thread with rounded corners, and there is an indentor which is a ridge with a notch in the middle, so it does two punch indents in a line along the axis.
This is next out from the 45A slot, and I think that it requires a bit less force than any of the others.
The actual brand on the crimper from the Quicksilver kit is ""
Enjoy, DoN.

O.K. The inexpensive style. :-) I prefer crimpers which are ratchet controlled -- or the hydraulic ones which have a release at a specific pressure -- 8500 PSI for the ones by AMP -- including electrically pumped as well as foot pumped.
Right. In my case, it was both wires going to the same binding post on the decade box, so you were measuring the resistance between the two binding posts.
Intersting device. But what it is lacking (as supplied) is input power from the line AC source.
What I tend to use for that sort of thing is Power Designs lab power supplies -- which BTW, given the mention of 4-wire measurement -- has a set of sense terminals which are normally back-panel jumpered to the power output terminals. However, if you care about the precision of the voltage at the load instead of the PS output terminals, you run a pair of sense wires to the load along with the power wires.
Anyway -- the Power Designs does have output current regulation, and a momentary contact button to short the output and a meter to let you set the current limit. (Not the degree of resolution that your device panel has, but still pretty good.)
My bench multimeter is the HP 3455A -- which was about $60.00 at a hamfest. (It has been a while, so my memory of the price may be wrong. :-)
Enjoy, DoN.

O.K. The inexpensive style. :-) I prefer crimpers which are ratchet controlled -- or the hydraulic ones which have a release at a specific pressure -- 8500 PSI for the ones by AMP -- including electrically pumped as well as foot pumped.
Right. In my case, it was both wires going to the same binding post on the decade box, so you were measuring the resistance between the two binding posts.
Intersting device. But what it is lacking (as supplied) is input power from the line AC source. ----------------------------- {True, they are components rather than complete solutions. Some people run them from old desktop or laptop power supplies. I built the unregulated Variac-controlled 12V and 24V battery chargers first and then bought the DPS regulators to go with them. They also run from solar and will charge a 12V battery from the 35-40V of two series panels. The 250VA charger has an input tap after the rectifier for up to 50V of solar power, which drops its AC demand to about as low as a KAW can read while the sun is out, then as the light fades it smoothly reverts to grid power. So far the only appliance I have that can take advantage of that free electricity is my Alpicool DC-powered freezer, but it costs me only $2.29 per month on grid AC anyway, according to its KAWez.} -----------------------------
What I tend to use for that sort of thing is Power Designs lab power supplies -- which BTW, given the mention of 4-wire measurement -- has a set of sense terminals which are normally back-panel jumpered to the power output terminals. However, if you care about the precision of the voltage at the load instead of the PS output terminals, you run a pair of sense wires to the load along with the power wires.
Anyway -- the Power Designs does have output current regulation, and a momentary contact button to short the output and a meter to let you set the current limit. (Not the degree of resolution that your device panel has, but still pretty good.) ------------------ {Although electronic surplus is plentiful in New England, good lab power supplies are NOT. I've seen more RF spectrum analyzers than CV/CC lab supplies for sale at ham fests. The best deal I could find was a new PS-305DM (30V, 5A) for about $60, and I NEED MORE POWER! Variacs are pretty scarce here too. } ------------------ My bench multimeter is the HP 3455A -- which was about $60.00 at a hamfest. (It has been a while, so my memory of the price may be wrong. :-)
Enjoy, DoN.
------------------- {That was a good price for a good meter. My best one is an ancient Fluke 8800A, 5-1/2 digits, $25. It matches this calibrated standard to the last digit, after the specified half hour warmup. }

O.K. The inexpensive style. :-) I prefer crimpers which are ratchet controlled -- or the hydraulic ones which have a release at a specific pressure -- 8500 PSI for the ones by AMP -- including electrically pumped as well as foot pumped.
=====================================
I'll gladly order and use the proper crimper if the company will pay for it; the hand tools can cost hundreds of dollars. For home use I collect what I find at flea markets, otherwise I rely on pull and resistance testing to determine the suitability of a crimp with alternate tooling. At Mitre we had a pull force tester to check wiring for aircraft.

My HF carbon pile load tester allows a voltage drop check of crimps at 100A. Even 1A is enough to measure the voltage drop and resistance in an inch of 12 AWG wire. For my solar wiring I short the far end and measure the total loop resistance to see if it matches what that length and gauge of wire should show. As a result I've upgraded most of my solar house wiring backbone to 10 AWG and reassigned the older, thinner wire.
Ox-Gard appears to protect exposed outdoor connections such as SAE (flat trailer) connectors for several years.

As should be for a decade resistance box.
O.K. Sounds useful, anyway.
Here (Washington DC vicinity) benefits from surplus sales from government labs which keep upgrading.
The best deal for power supplies which I have gotten was at a hamfest which was in the same town where I live (it later moved to a community college a few miles away).
Anyway -- It was a rack-mount frame holding two Power Designs 2005 supplies -- which were described as not working -- for about $5.00.
A glance at the back panel showed that the sense line jumpers were missing (one or both), and the screws in the terminal strip were loose.
Got them home, added the needed jumpers, and both worked as intended.
Interesting supplies -- 20V max 500Ma max, with four ten position knobs allowing selection from 0.000V to 10.000V, and a switch which added 10V to that. There was also a pot for tweaking over about a 1mV range, as well as a pushbutton to short the output and a pot for adjusting the current limit. (The meter had two scales -- 0-20V and 0-500mA and a switch to select which was in operation at the moment.
You can see one on eBay at auction # 303605152278
Another interesting Power Designs (Model 3650S) power supply is 0-36V 0-5A with a perforated cage on the back enclosing the heat sink, and the voltage pot has a Variac on the same shaft, so the raw DC into the regulator is a specified number of volts above the desired output voltage, minimizing the heat in the heat sink. Quite heavy, of course.
Example on eBay: # 383528772589
The most recent Power Designs I picked up about two years ago at the Manassas (VA) hamfest is 6050A (0-60V 0-5A) with a table on the front:
0-7V 5A 0-15V 3A 0-25V 2A 0-60V 1A
With switches concentric with the pots for (6V/60V) and (0.5A/5A) which both switch the range of the meter, and the range of the pot concentric with them.
And, there is a crowbar with a screwdriver adjust pot in the center of the front panel. A nice feature when you might have a load which is easily damaged by overvoltage, and a switch concentric to the voltage knob which can multiply the voltage by an order of magnitude. (There is a "fault" light and the power switch is marked:
=================== On
Reset ===================
That one cost me $25.00
All of the Power Designs supplies have remote sensing capability, and remote programming (resistor in series with two terminals) as well as current limiting. (Not truly constant current, since when the load resistance goes above a certain value, you *can't* pump that much current through the load with the available voltage. :-)
It helps that I was used to using the first two above at work, so I grabbed them when I saw them. The third was easy to recognize as Power Designs from the color and front panel style -- and I was sort of tempted to go for the extra one which was for sale. (Actually, I spotted one of those being carried by someone else, so I was keeping my eyes open for one for me. :-)
Example on eBay: # 174391995170
[ ... ]
Sounds good.
I check mine against a "standard cell" -- which is now decades old. I'm trying to remember the output voltage -- something like 1.0193V or so. Of course, I don't have the temperature control which should be around it. :-)
The first Fluke that I encountered has manual knobs for each digit of the voltage, and a meter to show how far above or below your setting was compared to the input voltage -- plus another knob to scale the meter's sensitivity.
The first digital voltmeter was a NLS (Non-Liner Systems) which used a Kelvin-Varley divider implemented in stepper switches, and a digital readout consisting of sheets of Lexan with the digits engraved in them, and lit from little aircraft lamps (327 if 28V, 328 if 6V, other numbers for other voltages.)
The big nuisance was when the voltage you were measuring was drifting up and down a little. As it goes up, you would hear click, click, click, but if it went down you would get rrrrrr, rrrrrr, rrrrrr, rrrrr, click, click, click. At least it had sound deadening in it, but when I had to work on them in the shop, the noise was a real pain. And, the vibration tended to crack the epoxy on the rectifier diodes, and they would short out after a while. It got so when they came in *new*, we would open them and replace the diodes with the glass-metal top-hat cased diodes of similar voltage and current rating. (We made the diodes as one of our products, so it did not cost us much. :-)
Enjoy, DoN.

It helps that I used to work with the crimpers at Melpar, where we built (among other things) flight simulators to mil-specs. So I could recognize the crimpers at hamfests.
O.K. For that, you really do want to measure the resistance.
Good thing to know. This is the compound used to protect aluminum wiring connections, or something similar?
Enjoy, DoN.
P.S. I may have just answered another of your articles by 'r' (Reply) instead of 'f' (Followup). At least on gmail it will get to you.

.......
=====================
I've disposed of millions of dollars worth of older USAF-property government lab equipment but never was able to acquire a bit of it. In theory it was to be offered to schools and similar non-profits, in practice scrap dealers won the bids. The one I tried hardest to buy was a lightly used 14" long bed South Bend lathe. I would have grabbed an RF spectrum or vector network analyzer too. What I have is a 10" South Bend and a 1970's non-computerized HP spectrum analyzer with a tracking generator and frequency counter to make it a scalar network analyzer.
----------------------------------
0-36V 0-5A with a perforated cage on the back enclosing the heat sink, and the voltage pot has a Variac on the same shaft, so the raw DC into the regulator is a specified number of volts above the desired output voltage, minimizing the heat in the heat sink. Quite heavy, of course.
---------------------------------- My homebrew power supplies can operate that way, though with independent variac and voltage regulator controls. A ganged variac and voltage pot would work pretty well for the supply with a control transformer but not at all for the one made from an arc welder, whose constant-current (50A) output characteristic makes the voltage drop fast as the current rises. The transformer's no-load magnetizing current is around 3A at line voltage so reducing the input voltage with the variac is helpful, particularly if running it from a small generator during an outage. ----------------------------------
--------------------------------- Shorting the output while charging a battery would not be good, and batteries tend to rise above a do-not-exceed voltage. --------------------------------
digit of the voltage, and a meter to show how far above or below your setting was compared to the input voltage -- plus another knob to scale the meter's sensitivity.
used a Kelvin-Varley divider implemented in stepper switches, and a digital readout consisting of sheets of Lexan with the digits engraved in them, and lit from little aircraft lamps (327 if 28V, 328 if 6V, other numbers for other voltages.)
drifting up and down a little. As it goes up, you would hear click, click, click, but if it went down you would get rrrrrr, rrrrrr, rrrrrr, rrrrr, click, click, click. At least it had sound deadening in it, but when I had to work on them in the shop, the noise was a real pain. And, the vibration tended to crack the epoxy on the rectifier diodes, and they would short out after a while. It got so when they came in *new*, we would open them and replace the diodes with the glass-metal top-hat cased diodes of similar voltage and current rating. (We made the diodes as one of our products, so it did not cost us much. :-)
Enjoy, DoN.
------------------------------------
I remember those clunky voltmeters. At least they were better than measuring with a Wheatstone bridge. Along with programming via punchcards they dissuaded me from electrical engineering. However the Army decided that was where I belonged, and by the time I got out in 1973 the field looked much more promising.
The simplest voltmeter I've built was a DAC chip wired to the data bits of a PC's printer port, feeding a comparator that came back in on a status bit. An R-2R ladder would also work if the port's outputs are CMOS. A short QBasic routine made it a successive-approximation voltmeter.
Point contact diodes in DTL logic caused many rocket failures during the 1950's. They'd open from the vibration.

... The first digital voltmeter was a NLS (Non-Liner Systems) which used a Kelvin-Varley divider implemented in stepper switches, and a digital readout consisting of sheets of Lexan with the digits engraved in them, and lit from little aircraft lamps (327 if 28V, 328 if 6V, other numbers for other voltages.)
The big nuisance was when the voltage you were measuring was drifting up and down a little. As it goes up, you would hear click, click, click, but if it went down you would get rrrrrr, rrrrrr, rrrrrr, rrrrr, click, click, click. At least it had sound deadening in it, but when I had to work on them in the shop, the noise was a real pain. ... Enjoy, DoN.
======================
Stepping switches and relays arose after an undertaker became upset that the local telephone operator, the wife of his competitor, was diverting phone calls away from him, so he invented what became the dial system to eliminate her. "Early advertising called the new invention the "girl-less, cuss-less, out-of-order-less, wait-less telephone".
In 1940 US cryptographic researchers figured out that the Japanese cipher machine used telephone stepping relays to implement a complex automatic version of the Crackerjack secret decoder ring. The researchers happened to choose the same model of relay to build their decryption device.
It didn't help predict Pearl Harbor because the Japanese hid their real plan among a flood of plausible fake ones such as destroying the lock gates of the Panama canal. We did the same to disguise the target of D-Day.

Where I worked, an Army lab at Ft. Belvoir, VA, things wound up in surplus sales (DRMO -- Defense Reutilization Management Office) open to anyone, though dealers got lots of things, I got things of interest, and recognized a lot of things which I had been forced to turn in as "uneconomical to repair" (actually quite easy to repair). (The property system had us either put special stickers on with the property number and the property ID number of who was supposed to inventory it about twice a year. Some from before then had the property id number or the initials of the person written on with a felt-tip marker.
The trick was taking enough leave off to participate in the auction(s).
Machine tools were a lost cause. They kept being stored outdoors in the rain -- including interesting CNC ones. The people who ran the property disposal office had no idea what machine tools were really worth (just what someone called their "current value" when they were turned in.) Electronics stuff got stored indoors, at least.
And I regularly saw items which I turned in, or which were turned in by friends/co-workers at hamfests. Lots of the dealers got things to re-sell, and some of them were re-sold at hamfests. Including some things from a contract which I had to hand-carry back from Texas by commercial airplane -- and they were classified, with paperwork to show that -- so they could not be inspected -- back before TSA at least. :-)
Yes -- it is more convenient when a manufacturer made it in a nice package, like Power Designs.
Right -- so you can set the crowbar up above the max output voltage for those conditions.
Speaking of batteries -- I used to use the 36V 5A supplies at work to un-short the NiCad cells in the HP 35/45/57/67 calculators. They would get to the point where they could accept no charge, until I would set the supply to 36V, and reach in with sharp points to connect to both poles of an individual cell. This would burn out the short in the cell, and it would accept a reasonable (no longer full) charge after that.
[ ... ]
:-)
Transitron used the NLS voltmeters to measure forward voltage drop in diodes, drop in Zeners, and forward voltage drop in base-emitter on transistors -- at three temperatures -- -50C, 20C, and +150C, reading the voltages in from those into a minicomputer, where they were punched on cards and carried to a mainframe where they were sorted for best match to get close to a 0 temperature coefficient. (So the NLS voltmeters must have had a digital output -- likely extra taps on the steppers which implemented the Kelvin-Varley divider.)
These were assembled in potted circuits called "ref-amps". I once had to test several returned ones from a customer, starting at -50 (in a mixture of two silicone oils -- one to be still liquid at the -50C, and one to avoid boiling at 150C) measuring them every 10 degrees all the way up to the +150C. Yes, they were out of spec at some intermediate temperatures. :-)
And -- someone passing by tossed a chunk of dry ice into the large beaker in which the tests were being performed while it was at 150C and still sitting on the lab hotplate. The outgassing dry ice carried silicone oil vapor over the lip of the beaker, down the side, and into the still hot coils, where the silicone oil vapor caught fire. There was a CO2 fire extinguisher ready to hand, and the fire was extinguished. The lab bench was covered with sand from the burning silicone oil vapor. :-)
Interesting.
Transitron made diodes of all kinds, including both germanium and silicon ones in glass tubing with a spring-loaded contact point, which was fused to the chip by a (relatively) high current zap. They had a machine which took the halves, assembled them, fused the glass closed, and zapped them. They decided they needed another one, and someone I worked with was part of the team which assembled it -- with some debate as to the polarity of the electrolytic which delivered the zap. I would like to tell you about it, but it would be longer typed out than the whole of what we already have here. (I'm at line 168 so far. :-) It is more fun to tell via voice, face to face. :-)
Enjoy, DoN.

I'm very familiar with those. I built a small dial exchange with original versions of the Strowger switches -- one line finder, one selector (one dial digit), and one connector (two dial digits). (Later AT&T cloned them, and called them 10x10s.)
And I think that, based on the complexity of the Strowger switches, he was mis-employed as an undertaker. :-) He really had to be a good machinist to make the prototypes, and good with relay logic to design the circuitry.
Those steppers were two-axis -- first up to one of ten levels, and then rotate to one of ten positions. And they could reset directly with a single pulse to a solenoid.
The ones in the NLS were unidirectional, 11 positions, with three sets of wipers at 120 degrees, so it took 33 steps to do a full rotation of the wipers. And there was no reset, so you had to go at least 10 steps to reach a zero point again.
There were other steppers, ten position after stepping off the at rest positions, and a single pulse to a solenoid to let the stepper spring back to the rest position. I used one of these as a line finder before I got the proper Strowger two-directional one. With the right relays, the line finders, or the connectors could really jump through hoops.
And the associated telephone set, was mis-called "The world's most beautiful telephone". :-)
Interesting. Probably made the task easier.
Of course.
Enjoy, DoN.

outdoors in the rain -- including interesting CNC ones. The people who ran the property disposal office had no idea what machine tools were really worth (just what someone called their "current value" when they were turned in.) Electronics stuff got stored indoors, at least.
{333333333333333333333333 I bought a dusty former Transitron oscilloscope that had been stored in a cave in New Mexico. I wondered if the white stuff was bat guano (or fallout?)
My $100 surface grinder was outdoors under a ripped tarp. Fortunately it hadn't rusted much yet. 333333333333333333333333]
And I regularly saw items which I turned in, or which were turned in by friends/co-workers at hamfests. Lots of the dealers got things to re-sell, and some of them were re-sold at hamfests. Including some things from a contract which I had to hand-carry back from Texas by commercial airplane -- and they were classified, with paperwork to show that -- so they could not be inspected -- back before TSA at least. :-)
{333333333333333333333333 I couldn't even transport classified material in my car. I had to lock it in the safe of an M109 shop van or strap it into the back seat of an Army Beech Queen Air. Once we got lost while carrying crypto gear under -very- low overcast a few km from the Iron Curtain. I called out town name signs while the pilot dodged power lines. We found our destination by IFR, I Follow Railroads.
Later I mentioned the incident to a Czech emigree who told me their radars were powerful enough to control air traffic over Paris. There was one in Russia that reached around the world.
I suspect we looked like just another car speeding down the road, except that we were 50' above it at 150 kts. I've watched cars on the Autobahn overtake our aircraft. 333333333333333333333333] Speaking of batteries -- I used to use the 36V 5A supplies at work to un-short the NiCad cells in the HP 35/45/57/67 calculators. They would get to the point where they could accept no charge, until I would set the supply to 36V, and reach in with sharp points to connect to both poles of an individual cell. This would burn out the short in the cell, and it would accept a reasonable (no longer full) charge after that.
{333333333333333333333333 I was taught to use a beer can sized electrolytic charged to 20V to blow out the dendrites, as you did with the diodes. The energy impulse was greater than from the smaller output cap in a power supply, and there was less chance of overcharging the cell if the leads welded stuck. Even an AA NiCad is equivalent to over 1000 Farads.
The same cap sometimes opened internal shorts in circuit boards. We found the location of the short by passing a few Amps through it and mapping voltage drops. Shorts can happen when you ask the board maker to push the limits of their spacing tolerances. When I went back to PCB design a few years later their tolerances were considerably tighter and Mil Spec was the normal standard.
The biggest zapper I've built was for GM, to duplicate the Load Dump when an alternator at full charge is suddenly disconnected by vibration of a corroded battery terminal. They had only been able to test their prototype surge suppressor with lab equipment that cycled too slowly. My fast-cycling machine melted their circuit in a few seconds. 333333333333333333333333}
Transitron used the NLS voltmeters to measure forward voltage drop in diodes, drop in Zeners, and forward voltage drop in base-emitter on transistors -- at three temperatures -- -50C, 20C, and +150C, reading the voltages in from those into a minicomputer, where they were punched on cards and carried to a mainframe where they were sorted for best match to get close to a 0 temperature coefficient. (So the NLS voltmeters must have had a digital output -- likely extra taps on the steppers which implemented the Kelvin-Varley divider.)
[333333333333333333333333 I designed and built test and calibration fixtures for high precision analog measurement modules so I know what a tedious PITA it is when a boring task requires constant concentration. I was very glad to see embeddable, electrically erasable flash memory to store calibration constants.
The digital output of some of my PC-interfaced DVMs is the LCD segment drive, which is easy to hash back to digits with a Select Case. Does Linux poll the printer port or can your code take full unhindered control of the I/O bit registers, as in DOS, to command external hardware when a reading exceeds a threshold? 333333333333333333333333]
These were assembled in potted circuits called "ref-amps". I once had to test several returned ones from a customer, starting at -50 (in a mixture of two silicone oils -- one to be still liquid at the -50C, and one to avoid boiling at 150C) measuring them every 10 degrees all the way up to the +150C. Yes, they were out of spec at some intermediate temperatures. :-)
[333333333333333333333333 While testing a bare IC chip and observing it under a microscope I dripped liquid nitrogen on it. The top Si3N4 passivation layer cracked like a windshield but the chip kept working.
At the high end I was checking out a tester that regulated the self-heating of a socketed power transistor by measuring Vbe. The life of the device was a few seconds at 200C, and under 15 minutes at 180C. 333333333333333333333333}
And -- someone passing by tossed a chunk of dry ice into the large beaker in which the tests were being performed while it was at 150C and still sitting on the lab hotplate. The outgassing dry ice carried silicone oil vapor over the lip of the beaker, down the side, and into the still hot coils, where the silicone oil vapor caught fire. There was a CO2 fire extinguisher ready to hand, and the fire was extinguished. The lab bench was covered with sand from the burning silicone oil vapor. :-)
{333333333333333333333333 A malicious prank like that is cause for dismissal. I don't mind harmless ones like a screen saver that occasionally flashes "REPENT, THE END IS NEAR" so it appears to be addressed to whomever is passing by at the moment. That was someone else's idea, mine was displaying Internet weather radar on an HP Infiniium scope whose probes were connected to a development board.
Dry ice makes an excellent curling stone in a long hallway because it floats on a gas layer. 333333333333333333333333}

And I think that, based on the complexity of the Strowger switches, he was mis-employed as an undertaker. :-) He really had to be a good machinist to make the prototypes, and good with relay logic to design the circuitry.
-----------------
All he lacked was an assistant named Igor.

Oh, yeah. I think I saw something like that in a very lightweight version 30 years ago. All the newest have a hinge in the middle and two extendable/removable outer sections. (I got an $89 Cosco, near identical to a $300 Little Giant 17') Oh, I did find your style after all. They're stronger than I last saw.
Muggles? (grumble,grumble)

Oh, yeah. I think I saw something like that in a very lightweight version 30 years ago. All the newest have a hinge in the middle and two extendable/removable outer sections. (I got an $89 Cosco, near identical to a $300 Little Giant 17') Oh, I did find your style after all. They're stronger than I last saw.
=================================
That's the type. Mine's older and has the wide foot on one end only, so I added another to make it more stable on dirt. Its round rungs hurt my feet unless I'm wearing thick-soled boots. A neighbor left his outdoors and the hinge latch mechanisms became hard to operate. The only easy way I've found to carry it any distance is open as a step ladder, horizontal, with a hand at the balance point on each leg.

Transitron made 'scopes? It must have been after I left them (in the early 1960s). At that time -- the only thing they *sold* were solid state devices -- including the Ref-Amps described later in the article, and something developed by total accident. A four-terminal device called a "Binistor". It was a Mesa silicon transistor in which the base bond was accidentally offset so it was partially on the mesa ring, and partly on the adjacent silicon. They went ahead and bonded another base where it belonged, and tested it -- and discovered that it had two stable states, pulse on one to turn it on, pulse on the other to turn it off. IIRC, it was labeled the "3N35".
Not as rainy where you were, then. Lots of rain in the DC area during the summer -- and lots of snow during the winter, neither of which is good for machine tools stored outdoors under *no* tarp, ripped or no.
O.K. A difference in location -- this was between Texas and Northern VA via commercial aircraft -- with the (rather small) box in my lap, not checked. No stops where any available safes were likely. Drove from the airport directly to work (Luckily it was daytime), and also the classification was not that high. It still gave me a shock to see the devices at a hamfest a few years later.
It was a prototype device to go on top of a helmet to generate audio tones indicating which direction you were being targeted with a laser from. :-)
The "Russian Woodpecker" -- which made a mess of ham bands, among others.
:-)
The test points were small enough to be easy to pull back off if they welded.
I used that approach when debugging a board for the Altair 680b. I got one of the first of them in the area (680b used the Motorola MC-6800 CPU, while the 8800 used the Intel 8080 CPU). I wound up doing repair work for the local vendor. They had one board which was bought as a kit, and when assembled, would not work. Low voltage with enough current so I could trace voltage drop with a 0.001 V resolution DMM. It turns out that the problem was a short between two traces -- and they wanted to call it the kit builder's fault, until I pointed out that it was under the overcoat on the PC board.
:-)
The opposite was a Mitsubishi alternator in a later Mazda GLC. It was one of the early ones with built-in regulators, and the regulator failed in a "charge the short circuit" mode -- Burned out headlamps, boiled the electrolyte in the battery. It was under warranty, at least. :-)
I forgot to mention that the diodes or transistors were connected in series, with a constant current through them, and steppers taking the + and - probes of the DVM up the chain -- and feeding it to the computer along with the DVM's measurement output.
I can imagine. That was a bit early for that -- about 1961 IIRC, and also no microprocessors, which would have been great for the task otherwise. (Actually -- before ICs too. :-) Thus the minicomptuer controlling the run.
Also -- some other in-house test equipment used OP-AMPs -- back when the only ones available in modular form were the Philbrick ones with two tubes plugged into the top.
The series connections, with all the junctions coming out to a multi-pin connector were put into silicone oil at the desired temperature (controlled temperature). After running the test at that temperature, you (well someone -- not me) you unplugged it and carried it to the next temperature station -- and I presume waited for N minutes to let the temperature stabilize. :-)
Easy enough with C -- does/did the Microsoft BASIC have something similar?
I have no idea. I haven't used the parallel printer port for I/O like that. I would have been more likely to wire-wrap up a custom MC-6800 or MC-6809 board with enough PIA chips to do the readings, and transfer them via serial port.
Or before the PCs, the SWTP 6800 and SWTP 6809 computers had parallel I/O boards -- which could be used for talking to parallel printers, or any other multi-pin parallel connection needed. The PIA has two 8-bit parallel ports with handshaking pins, and if wired right, can have 16 bits occupy two consecutive addresses (and the control and I/O registers on an adjacent two bytes). Transfer it to an index register on the MC-6800, or to the two accumulators (which could be treated as a single 16-bit register) on teh MC-6809.
OS of choice for me was the SSB DOS-68 for the MC-6800, or OS-9 for the MC-6809.
Ah -- but for how long? Easy for contaminants do do it in once the silicon is exposed.
Constant current to the base? And maybe variable C-E voltage to get to the higher temperatures.
It should be, indeed. I was a summer employee when it happened, and the fellow who did it was a full-timer. At least, he was the one who climbed on a stool with the extinguisher and put it out.
Sounds like fun.
I should mention one which was brought back from a user's group meeting for the CDC 6600. The 6600 console had two *big* green round CRTs. Typically, one displays the status of the system, and the other can be used to examine or modify one particular job at a time.
Anyway -- he brought back a tape, loaded into the system, and set it up for execution late in the evening shift -- after the operator was likely to be a bit drowsy.
Suddenly, the two monitors blanked, then a pair of eyes slowly rose from the bottom of the console onto the screens. The eyes looked at the operator, down at the keyboard, up and towards where the clock likely was, back at the operator, and sunk back off the screen. After that, normal operation resumed, and everybody left in the center (including the one who brought the tape back) came in laughing after the operator screamed. :-)
Hmmm ... Someone who used to work for Ampex, maintaining the cake carrier/washing machine disk drives. In general they were quite reliable, except for at one site.
As the jobs roll around, he winds up working at that site as an operator. First time on night shift, after doing the system backups, and carrying the disk packs back to the safe, an old-timer introduced him to the sport of curling -- using the disk packs. :-)
Enjoy, DoN.

Transitron made 'scopes? It must have been after I left them (in the early 1960s).
4444444444444444 It was a Tek scope with a Transitron property tag. George Roach found it for me. 4444444444444444
Easy enough with C -- does/did the Microsoft BASIC have something similar?
4444444444444444 The final version of QBasic includes the structures and syntax of Pascal such as SELECT CASE and DO loops, subs and functions and one level of pointers, without the academic fussiness. I preferred it to C because I could understand and modify what I had written 3 months previously, as could the project engineer. It still retains GOTO for hasty exits from unrecoverable errors, otherwise I coded in proper nested loops. Library Includes and variable definitions at the start are optional if you prefer them. I never had a problem with the dual definition of "=" retained from earlier Basic. 4444444444444444
I have no idea. I haven't used the parallel printer port for I/O like that. I would have been more likely to wire-wrap up a custom MC-6800 or MC-6809 board with enough PIA chips to do the readings, and transfer them via serial port.
4444444444444444 The idea was to connect applications boards for new ICs to the customer's lab computer via the parallel port, which was still common in 2000, especially on older hand-me-down lab machines. Windows messes with the port and Visual Basic lacks the instructions to read and write it, so we fell back to DOS and QBasic which do allow full access to the hardware registers in BIOS space, interrupted only by the brief clock update every 55mS. I coded an I2C serial interface using the printer control and status bits.
QB has limitations like no USB or mouse and poor integration with Windows, though the mouse can be added by calling a pointer to an interrupt routine stored as a string.
The graphics capabilities were sufficient to display a representation of the development hardware on-screen showing the voltages measured with that DAC A/D converter, numerically and by color. I made the address bits for a programmable memory cell array walk around the border of the array as they were written in, so it looked like an animated video game. Later, not on company time, I figured out how to create moving graphic images called Sprites and play tunes.
The overall effect was to turn the computer into an enormously fast and powerful Arduino with I/O bits, mass storage, a development system, keyboard and screen. I configured a boot floppy to install and run DOS on the host's hard drive. QB's integrated development system is particularly nice, allowing interpreted or compiled execution. The compiler is quite efficient, an integer FOR loop executes in 2 CPU cycles, 2nS on a 1GHz CPU. SpeedStep ruined that timing capability on later machines. 4444444444444444

... I have no idea. I haven't used the parallel printer port for I/O like that. I would have been more likely to wire-wrap up a custom MC-6800 or MC-6809 board with enough PIA chips to do the readings, and transfer them via serial port.
===============
Here's how to control the printer port in C:
A useful first exercise is to write a program that displays a graphic of the port and the state of each bit.

I think a lot of the maintainability with C is down to style and organisation. My background was in mechanical engineering but got into computing during my degree and went on to get a job at a software house and trained in C where the company's in house style made the code easier to follow IMO than K&R style and my engineering background helped with organisation. I provided a university technician with the source code for a port of his PCB program to Windows in C and he found it easy to follow whereas he often found students C code difficult because of how they were taught to write C.

.................... I think a lot of the maintainability with C is down to style and organisation. My background was in mechanical engineering but got into computing during my degree and went on to get a job at a software house and trained in C where the company's in house style made the code easier to follow IMO than K&R style and my engineering background helped with organisation. I provided a university technician with the source code for a port of his PCB program to Windows in C and he found it easy to follow whereas he often found students C code difficult because of how they were taught to write C. =============================
Yes, I've seen some that looked like the programmer's main goal was job security. Perhaps engineers are more relaxed when programming because they have other valuable skills.
I began with an LSI-11 flavor of Pascal customized for hardware control, an uneasy fit due to Pascal's strict typing versus variables that were bit-mapped binary register contents, but most of my early programming experience was hand-written assembly code for the 8080 computer I scratch-built. It forced me to be careful to comment everything well enough that I could understand and improve it later. Eventually I wrote a text editor that could assemble new or disassemble older code line by line, but by then I had coded perhaps 10,000 lines by hand and needed to expand the 16K memory.
As soon as I had acquired enough RAM I dropped the habit of writing tricky, tightly condensed code. The company made production memory wafer testers for National Semi et al and the engineers received pre-production samples of new devices like the 6116, which were typically too slow (500nS) to use for much beyond evaluation, so after they were done they gave them to me.
It can use the same socket as a 2716 UVPROM. There's also a 2816 EEPROM variant. I backed them with NiCads so I wouldn't have to boot from the Teletype each time.