Retrotechtacular: DC To DC Conversion By Vibrator

Electricity comes in two basic forms: Alternating Current (AC) and Direct Current (DC). DC is handy to use and is easy to analyze. However, AC has some useful properties too. In particular, AC current can operate a transformer which can step it up or down easily. Power is conserved, of course (well, actually, you get less power because of losses in the transformer).

You can’t do that trick with pure DC. You can reduce a voltage, although that typically wastes power in heat (for example, a voltage divider or linear regulator). You can’t readily increase a DC voltage unless you convert it into some sort of AC first.

This was a particularly bad problem in the era of tubes–especially tubes in car radios. The car’s voltage was probably 12V but the tube’s plates might take hundreds of volts. What do you do? Some old car radios used what is called a dynamotor. This is just a motor and a generator in one box. You could spin the motor with 12V and have the generator produce a different voltage (even a DC voltage).

The Electric Dynamotor

If you think about it, a transformer is really just a generator (the secondary) with the shaft replaced by a moving magnetic field (the primary). However, with a dynamotor, you can use a DC motor to spin the input. With a transformer, you must have AC input because that’s what moves the magnetic field. Note you don’t have to have a sine wave, necessarily, just current that switches polarity fast enough to drive the primary.

So how do you get the DC in a car radio into AC? There’s a clue in another part of the car: the blinkers used to flash the lights. Today, you’d probably do the blinking with some electronics, but old cars didn’t have a lot of electronics. What they did have was a special kind of device. When you apply current to the device, it would heat up a bimetallic strip. The strip would slowly move and break the current flow. This would cool the strip down which would resume its original position and start the current flowing again.

If you wanted things to go faster, you could do the same trick with a relay. Have the relay coil take current through a normally closed contact. When you energize the coil, the contact opens and breaks the current flow. The contacts then close and the process repeats.

Testing, 1, 2, 3

RCA-tube-tester-at-Oklahoma-History-CenterOld tube testers in the drug store (see right) often had a test for vibrators. In those days, that didn’t have the meaning you’d expect today. In fact, a vibrator was a relay wired to interrupt DC current into AC. You might expect that these failed pretty often and a trip to the drug store would let you test vibrators and tubes, buy a replacement, and repair your own devices.

The name vibrator came from the characteristic buzzing noise made by these devices. Once the DC was broken up by the vibrator, a transformer could step the voltage up (or down, but it was almost always up). This produces a higher AC voltage that the circuit would then rectify and filter to get the desired higher voltage.

Heathkit_VibratorThere was actually a lot more to making a good vibrator power supply than you might think. The mechanical parts moved constantly. Also, the contacts would spark and that would eat at the contacts. It also created a lot of interference or hash. A good design would suppress sparks and hash. The resulting rectified current needed a good bit of filtering, too. The photo to the left shows a pair of Heathkit vibrators, including the internal structure that was normally hidden.

Although the vibrator seems inefficient, it beat the dynamotor. It was quieter, had fewer moving parts, was smaller, and cheaper. At the time, it seemed like progress.

Today

Vibrators are long gone except in vintage gear. Transistor power inverters became practical and edged them out. Today, you would be more likely to use a switching mode power supply to get the same effect. The principles aren’t that different, but the conversion of DC to AC is electronic and the control is more precise.

If you want to see more about vibrators–including schematics–check out the video below.

Vibrator photo By WikimediaFoundationVEVO (CC BY-SA 4.0)

42 thoughts on “Retrotechtacular: DC To DC Conversion By Vibrator

  1. Of course, at one point there were tubes that could run off low voltage directly, but those came about the same time as transistors.

    Another scheme was transistorized replacements for vibrators. Not much more than a couple of power transistors, Popular Electronics had a cover article about building one about 1962. There was a wave of dc-dc converters that used the same basic idea, see the ARRL Handbook circa 1971.

    Michael

    1. Thanks for reminding me of this, Michael. Since Hackaday pointed out some time ago that older Popular Science is available online ( http://hackaday.com/2014/12/03/popular-electronics-magazine-archive-online/ ) it’s a snap to find those articles. One front page article is of a build that is refined to the point of fitting in a vibrator socket. It is shown in the December 1963 edition (pp 41-45) authored by R.L. Winklepleck of Terre Haute, IN. An earlier build with larger form factor is written up in October 1960 (pp 64-64) by Patrick A. Gainer, who worked at the Langley Research Center at the time. (PS revisited this subject every few years). The online archive is easier to use than the incomplete set I have stored in boxes.

        1. Glad to be of service! Certain new-old-stock vibrators can be had, but they only had a certain life cycle. It was apparently common and reasonable to have the car radio pulled out of the dash and have the 0z4 (usually) rectifier and vibrator replaced as PM on an annual basis. One ‘soapbox’ message on the subject of dynamotors: If you are restoring a military receiver or commercial/military transmitter that uses a dynamotor, seek out the exact maintenance instructions about lubrication and cleaning, checking the brushes for sticking, hardness, etc., –and then consider making a sturdy, ‘fire and forget’ solid state replacement instead. The reason is, original DM-* dynamotors for the WW2 shortwave receivers now routinely sell for as much as the partner radio does without one. Early amateur radio operators were aware of the maintenance cycle of the old Signal Corps sets on the surplus market and removed and destroyed the majority of these dynos–hence, a maintained, functioning unit is a rare treat to curate on an infrequent basis, to people who never got to experience things like an old Army Air Force BC-348 with a DM-28 singing away in its aluminum case, a ground unit BC-312 with a DM-21, or a fully racked SCR-274 set hosting a bank of DM-32’s. Seventy plus years down the road, this once ubiquitous ex-military scrap is rare in unmolested form, getting more attention for its historic importance and gaining some collectior value.

          1. Packrat, I agree! Your comments on old WW2 radios brought back fond memories of my early ham gear. I had a BC-348 and SCR-522, modified for the ham bands.

            Old vibrators are not likely to work without some repairs. The picture doesn’t show the foam rubber sound dampener that was often inside the can. Over the years, these have usually disintegrated and contaminated the contacts. I’ve had to open the case and clean this mess out to get it to work again.

            The old metal-cased power germanium transistors and their heatsinks.were too big to fit in place of a vibrator. But today we have tiny MOSFETs that can easily do the job. I wonder if you could use the vibrator’s contacts to switch the transistors? Then the contacts would last “forever”, and the transistors would do the high-current switching. You’d have the same “look and feel”, but without the reliability headaches.

            Dynamotors are also an interesting technology. The ones made for the military are solidly built, and likely to still work today (with a little cleaning and relubrication).

          2. The picture above DOES show the foam rubber sound insulation, if it’s that yellow stuff.

            Yeah you could use the vibrator to drive a MOSFET but there’d still be arcing as it made and broke the circuit for it’s own coil, that’s a fair amount of current in itself, and coils give high-voltage back EMF so you’ll still get sparking, at something intentionally buzzing away at 60Hz or so it’s not gonna last that much longer, I shouldn’t expect, just with a modification.

            I’d just replace the whole thing solid-state. But if you really want authenticity, then a vibrator is just a relay. There’s reallyno difference in principle. Maybe they tune the weight and springiness of the moving contact for a particular frequency, or maybe they don’t. But it’s still just an ordinary relay cutting it’s own power off. So you’d need a DPST or DPDT relay and wire it up so. One set of contacts powers it’s own coil, the other set goes to the transformer. Then replace that with a new relay whenever you need to. You could even hack up a little adaptor. It’d take up more space than the original thing, which looks optimised for space. Which is reasonable enough. So a relay with similar ratings wouldn’t fit into the housing of an old vibrator, but they why chop up an old one anyway? If you must do, then have the adaptor run off to a relay socket. Otherwise from scratch. 1x relay, 1x bit of wire. Can’t go far wrong!

            If you knocked up a fully MOSFET-based alternative from scratch, I wonder what difference operating frequency would make? You could go fully-modern, but I think the PSU is part of the radio too so would be good to keep it as close to original as possible, I think replacing an old relay with a new relay isn’t too much of a crime, especially if it keeps the radio singing.

            I wouldn’t go with a hybrid relay-driving-MOSFET as you suggest, worst of both worlds there for no advantage, and doesn’t make the relay live much longer.

  2. You can still find this method of DC/DC conversion in use today. The joke “shocker” pens that give your chosen victim a jolt of high voltage often have a small solenoid embedded.

    From what I remember they don’t do much in the way of regulating the output, it’s more taking advantage of inductive kick (current through an inductor can’t instantaneously change, so voltage tries to go to infinity as the field collapses). But, it’s a simple example of this tech that you can still find today.

    1. Otherwise known as back-EMF… yes I’ve felt it, by being a little slow connecting the spade lugs from a 400A relay to the terminals of a 7Ah SLA battery. The coil on the thing was huge and the repeated disconnections as I tried to put the spade lug on gave me a big zap!

      I reached for the circuit breaker after that.

      1. When I was really young I tried connecting a 9 volt battery to almost everything with a cord attached.
        A large fan motor zapped the urge to try that again right out of me!

    2. The joke pen I’ve seen uses the mechanism out of an electric-sparked cigarette lighter. A spring is compressed then released, driving a hammer to hit a little rod of piezo crystal. Wires carry the voltage to the spark point, in a lighter, or to the electrodes in a pen.

      I found one once in a shop with my mate, tried it, “ARGH!”. My mate, having seen this, asks “what does it do?”. “Oh, it plays music”, so the gullible sod tries it…. “ARGH!”

      1. Talking of piezos… there’s an unusual, but very compact, method of high voltage generation, the piezo transformer. One piece of piezo material is run at some KHz by a circuit at low voltage, and it’s mechanically coupled to another (generally two piezo elements on the same piece of metal), which outputs high voltage. Smaller and doesn’t generate any magnetism.

          1. “LEDs and solar cells”.
            A dis-integrated opto-isolator!

            I have charged some solar lights by leaving the bathroom incandescents on for a night. And I felt bad doing it, lol.

  3. Nice article.Have a couple of those lying around, and i actually still find them as useful as SMPS for some uses…
    I might be wrong, but shouldn’t the title read “DC to AC…”?

  4. This vibrator and even a DC brushed motor being used as a generator is still following the same rule — there is AC in some form happening. Edison, in spite of all his wisdom, seemingly refused to acknowledge the realities of alternating current existing in the windings of his DC generators. It was simply being rectified by the commutator and brushes.

  5. “In fact, a vibrator was a relay wired to interrupt DC current into AC.” No, it was designed to create pulsed DC. (Been around long enough to have had a lot of experience fixing those old, tube type car radios.)

    1. Designers were pretty clever, and managed to make vibrators more reliable than you might think. It wasn’t unusual for one to last the life of the car back then (perhaps 1000 hours of radio usage). Imagine how many times those contacts had to switch in that length of time!

      Synchronous vibrators had a second pair of contacts on the secondary side to do the rectification. They converted DC to DC directly (no diodes needed).

      The output voltage of a vibrator power supply was regulated by the turns ratio of the transformer. The vibrator produces square wave DC, and the transformer turns ratio stepped this up into a fairly predictable higher square wave voltage. Just like a conventional 120vac to 12vdc transformer-rectifier supply, the no-load to full-load voltage might vary +/-20%, which was good enough for a radio.

  6. it wasn’t just car radios, a lot of home radios had vibrators.
    In this day and age it seems inconceivable that people in rural areas didn’t have reticulated water or grid connected power.

    Here in Oz I’ve seen quite a few shelf and cabinet radios that were battery powered and used vibrators.

  7. These vibrators were the key component to American Flyer’s “air-chime whistle,” a toy-train accessory from the 1950s. The catalog made a big deal about how faithfully this device recreated a diesel horn, but it really just made a buzzing noise that was picked up by a speaker in the locomotive.

  8. Probably my first real hack was running a 6 volt car radio on a Lionel train transformer. They free and still used common tubes instead of the latest (deckchairs on the Titanic) low voltage tubes. I was having trouble with the vibrator and messing with the contact. I jammed it closed and surprise the radio came to life. It worked better with the AC and not both the vibrator and AC.
    The weirdest hack was a cover article on one of those G rated men’s zines.
    Shave in your car!
    Run your then common 110 volt AC/DC motor shaver on the B+ of your car radio! A double pole switch and a 110 chassis outlet in your metal dashboard and an afternoon and you’ll shine. You can’t listen at the same time of course, besides those were RFI bombs on radio & TV reception.

    1. Vibrators were generally used for low power applications that needed less than 100 watts (like receivers). Dynamotors were used when more power was needed (transmitters). One that drew 60-100 amps (360-600 watts at 6v) would be quite large; more than a normal car or truck generator could handle. Aircraft used large dynamotors like this, but they were 24vdc systems.

  9. This article makes it sound like the vibrator was used to chop DC into (square-ish) AC which was then fed into a transformer.

    The way I remember it – there was no following transformer. The much higher voltage back EMF from the coil was directly rectified into the higher voltage DC for valve (tube) anode.

    1. The vibrator power supplies for car radio equipment did indeed chop the incoming DC into AC, run it through a transformer to step it up, and then rectify it for use. The circuit was a classic 2-switch bridge with a center-tapped transformer, exactly like you find in modern switching power supplies but with switches instead of transistors.

      The rectifier was often an 0Z4, an interesting tube that had a “zero” volt filament (i.e. no filament).

      The circuit was trickier than it looked. There was a resonating capacitor, so it operated in resonant mode so the voltage across the switch just happened to be zero when it switched. This greatly extended the life of the contacts.

  10. Interesting article. I came across these beasts twice: My dads old portable photo flash had a briefcase size box containing a lead/acid battery (Liquid acid in a portable device!), the vibrator and a coke can sized cap to store the high voltage for the xenon tube in the flashgun. The other was a portable scientific chart recorder that contained both valves as the pre-amps and germanium transistors to drive the pen. The vibrator was used to create the HV for the valves.

  11. Sandwiched in between the 120 VAC kit for auto alternators and Solid state 12 VDC to 120VAC inverters where 12 VDC to 120VAC 60Hz dynomotors. Around here practically every every oilfield electrical contractors service truck had one mounted on it. To tun power tools while constructing secondary distribution lines and installing motor control boxes. Many of the electricians said they worked great when camping. Didn’t take too long before the dynomotors where superseded by inverters.
    Almost every time I read an article on the vibrator radio power supplies I read about how often they needed to be replaced. Thing is I don’t recall the radio in the family cars being nonoperational at all. Tube radios until our family got a used car that had a transistorized radio in in the late ’60s.

    1. I worked as a TV/radio repairman in the 1960’s and 70’s. Like Doug, I don’t recall having to replaced vibrators very often. The tubes in car radios didn’t fail that often either, especially once they switched from generic tubes intended for home radios to the more ruggedized versions.

      Tubes and vibrators lasted “long enough”. Keep in mind that the life expectancy of a car (in hours of operation) isn’t that long! 100k miles at an average of 50mph is only 2000 hours.

      Dynamotors and motor-generator sets are still used because they can be extremely rugged. Most cheap electronic inverters produce poor waveforms, and can’t handle “tough” loads like motors. The output of a good motor-generator is a clean sinewave, and it can handle huge peak loads to start motors etc. They are also bidirectional; you can feed power in either side, and take it out the other.

      PS: A “dynamotor” is a single motor with two sets of windings. If it’s DC-DC, it has separate commutators at each end. If it’s AC-DC, it has a commutator at one end, and slip rings on the other. A “moor-generator” is just as the name implies; two separate machines connected by a shaft. They are bigger, heavier, and less efficient; but easier to build.

      1. Thanks for the great clarification I plan on making a Dynamotor.

        Heads-up to anyone who wants to make a low wattage one. Some cheap rotary shavers actually have a ~90 volt dc motor inside. With no isolation!
        The warning to use a GFCI is there for a really good reason. D:

        1. To be fair, bathroom sockets are usually isolated. There’s the special shaver symbol that means it meets some standards. One of which is a fairly low-powered isolation transformer. No GFCI for in in homes I’ve been in, but that might well be in the distribution box mounted on the wall in a cupboard or under the stairs somewhere.

          In the UK there are no sockets at all allowed in bathrooms except for the one isolated shaver socket. Which has a different pinout, with 2 round pins, so it’s incompatible with British mains devices and so is the shaver.

          We also have separate mains runs for lights and sockets, earthing has been compulsory since the dawn of time, our plugs are all fused (except the illegal Chinese ones that put insulating sleeving on the Earth pin for fuck’s sake, who even teaches them electricity!?!?). Our plugs are so sturdily built they function as improvised caltrops. Whether you like it or not. You don’t forget a good foot-gouging from a couple of inches of solid brass!

          Actually with shavers using different mains plugs from other appliances, I wonder if they even make GFCIs with those pins? I mean the plug-through GFCI type you use with lawnmowers etc. I don’t know where the shaver socket connects to the distribution box and if it’s protected. It’s common now to have GFCIs in the distribution box to protect all the sockets, so maybe it’s covered that way. The isolation will prevent a lot of problems though.

          If your house hasn’t got an isolated shaver socket you should probably fit one.

      2. So what do they do? Have a set of AC stator windings, a set for DC, and an unconnected spinning permanent magnet on the shaft?

        The problem with bidirectional operation would be the same one transformers have. That is, due to inefficiences, a 240V:12V transformer doesn’t have 20:1 turns. It’s more like 18:1, to compensate for inefficiencies. Running one in reverse means you’re working against the inefficiencies and the measures put in place to compensate for them!

        Still back in the day they built stuff tough, so I suppose you could just rely on slack in the system, built with wide enough tolerances.

  12. One of my projects was to convert a tube car radio to use AC power by replacing the vibrator and transformer with a transformer from other tube equipment from the junk box. Worked. (Didn’t call them “hacks” back then.)

    1. That’s cool. :)
      I wish I was born back in the ‘tube days’! I probably would have gone absolutely crazy with them.

      This article makes me want to make a radio from ‘scratch’. As in no semiconductors and no PCBs.
      High voltage, dead-bugged tube sockets, and a motorized DC-DC convertor. :D

      I’ve been through a junkyard hundreds of times and found a radio vibrator only once. I wish that my curiosity hadn’t destroyed it but a hammer-and-screwdriver was the only way to figure that one out. :(

      +1 to all of the older hackers here!

      1. If you don’t mind putting up with AM, tube radios still work! Ideally just the strong AM signals too so you can just build a simple TRF receiver and not bother with heterodyning and positive feedback and stuff.

        There was a book I recall at school as a kid. The book is from the early 1960s but I’m not. Budgets, eh? Plenty of money for shitty overpriced computers, if you know how to speak their language, educationalists are a rich market of suckers. Anyway it’s a Ladybird book on how to make your own transistor radio. I read it wondering WTF the apparently common “red spot” transistors were. And it’s made on bread board. Not breadboard. Bread board! For cutting actual bread! That’s make of wood! The board not the bread of course. Bread board and countersunk brass screws!

        https://archive.org/details/MakingATransistorRadio-LadybirdBook/mode/2up

        It gets as far as an audio amp and a radio RF amp, starting off from just a crystal set with a germanium diode. So get on your best tank top, get your mum to cut your hair, and call up the Universal Engineer Spirit (who is the father of all of us not lucky enough to have engineer fathers around), and sally forth! Katie COBOL!

        No it’s not valves / tubes, but I bet it’s easier to make and adjust. Still valves seem straightforward in principle, never used them. Interestingly for a book from 1961 or whenever, most of the components still look the same. The obvious exception would be the tubular resistors rather than the bobbly ones now.

  13. Ran into one of these a few years back when my dad bought a 57 chevy to relive his younger years. The radio didn’t work, so I offered to fix it. It was easy to remove – thankfully – and presumably designed that way because it required maintenance. Had to buy a new OZ4 rectifier, adjust and clean that vibrator/interrupter, replace two tubes, and replace a couple of caps that dried up. It worked remarkably well. The amusing thing in this day of “instant on” was waiting for it to “warm up” before you heard anything out of it, and also amazing that the RF bands it can receive are not obsolete. Only drew about 5A when running. I wonder if I tried to boot up my current smartphone (which performs the functions of a “radio” – in that I can listen to music that is streamed digitally over RF) in 59 years if the hardware would at least boot up – forget about the fact that the cell providers would have obsoleted LTE long before 59 years had gone by.

    1. Here in 2020, 2G is hanging on by the skin of it’s teeth. Or put another way, it’s flourishing! Many people just don’t want a smart phone, so they stick with feature phones. Which I wouldn’t be surprised to find out all run on the same Mediatek SoC. I’ve had wristwatches run on it and seen identical software on other cheap phones. It’s a shame nobody seems to have produced an SDK for it, would be ideal being a wristwatch, and so cheap ($7, seven dollars!).

      The chip must cost virtually nothing, and the networks in the UK at least weren’t decisive enough to just shut 2G down with a snap, give ’em 6 months notice or the like. So as prices fell through the floor more and more 2G handsets appeared instead of disappearing.

      2G is something like the IBM PC of obsolete technology that won’t die! Although I suppose 2G did a good job in it’s day so has earned a respectful goodnight. Bloody China though aren’t going to let that happen!

      Apart from that, I think they were making the first steps to obsoleting 4G as soon as they were putting the towers up! Eventually something much more cognitive and negotiate-y will appear and that’ll be the end of Generations altogether. Phones will just negotiate space from each other, all radios will. And it’ll all appear like noise, optimally compressed, so I hope the aliens caught sight of us while we still looked like we were transmitting anything.

  14. I still have two of the 12 volt DC to 120 volt AC Powercon units from Cornell-Dubilier Electric Corp. One is the Sportsman which is 60 watts continuous and 80 watts maximum. I don’t have the original paperwork so I don’t know how long it can be used at 80 watts. Long in the distant past, I used one like it to power an AC-only CB radio in a car (that dates me, doesn’t it?).

    The other unit is a Sportsman Senior rated at 140 watts, Model 12SS14. It’s built into a metal case with a hinged lid and a lift-up handle on each side. There is space in the unit for a 12 volt lead-acid battery (I think it’s a group 27 – the unit is from 1957 and that battery size was popular then). Also included in the case is a charger, a DC voltmeter and a +20, -5 amp off-center-zero ammeter so you can monitor the battery voltage and charge/discharge current. The control panel includes a 2 prong AC outlet (female) and a 2 prong AC input connector (male). There’s a three position switch that selects taps on the 120 volt side of the transformer for
    low charge, high output voltage
    medium, charge, medium output voltage
    high charge, low output voltage.

    With a light load, you could extend the battery’s working life (maybe 3 hours) by starting in the “low output voltage” position and going up as the battery discharges. In one advertised application (small portable TV), that would have been watching for the screen size to start shrinking (indicating low AC voltage) and then turning the switch up.

    The highest charge position has an in-the-lid caution that it’s only for a few hours and then must be turned to medium which can be left charging. The low charge is a listed as a storage setting.

    What good is an inverter that only delivers 140 watts (in the days of multi-KW pure sine wave inverters less than half its size)?

    When we were moving into a house that didn’t yet have AC power, it let me run a corded drill to mount a window air conditioner so we’d be cool when the AC was turned on ;-) With today’s almost universal availability of lithium-powered portable tools, that’s not an application that most people would think about. In the 60’s, it was a big deal to have a drill that didn’t need commercial power.

Leave a Reply to notarealemailCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.