|I've had many people ask about "The Bench" over the years - and I have always shared the schematics. With the popularity of the Internet - and the exchanges of ideas, etc. - when I got flooded with requests after "The Bench" was talked about on some News groups, it seemed a good idea to update my own hand drawn schematics, and notes - and share them with everyone through a web page. This is what this page is about - as noted below - this isn't the greatest, the best - or even a necessarily good way - to accomplish the goal - but it is the way I did it - 20+ years ago... and hopefully it will provide the basis for others to build on -- and yes - improve upon. The bench - like myself - is old and tired - and past learning new tricks. So take this only as a starting point - and good luck in your project - be it a small, simple one - or one as large as The Bench. |
This project was done some 21 years ago - and - as I was in the Navy at the time - was built using what parts I had on hand - not necessarily the best or most appropriate parts - and certainly - not using modern parts available today. Having said that - there is nothing wrong with the basic design - it has worked day in and day out without a significant failure for 21 years - the only parts to have failed have been indicator light bulbs - and the 28V supply (which is a commercial supply) was damaged by lightning about 8 years ago.
|Another view showing wire rack above the bench. As many radios as we see with UsedToBe "rubber" insulation - we've just gotten used to using a lot of wire... |
The bench is divided electrically into three major subsections - The "Distribution" section - located under the bench top; the "Panel" section - which has the operator controls and circuits; and the "Heat Sink" - which is basically the backside of the bench proper. The bench is an old Navy "communicator's" desk- which is a very heavy steel framed desk - an upper shelf unit - and a recess for a typewriter. When this project was taken on - the recess for the typewriter was plated - leaving a large, open desktop - and a place for a "Tool drawer" open to the front -- as can be seen in the picture.
We'll get the "HeatSink" out of the way first since it's the simplest part of the project. The bench has a very large built-in power supply - which provides 28VDC for it's own switching and control - as well as external 28VDC devices (like Sherry's R-391 autotune) - and has a 0 - 30V 10A variable power supply accessible from the panel. The series pass elements for both supplies are mounted to the back of the bench (7 LM395 integrated transistors as pass elements, 1 LM395 as an active sink for the 0-30V 10A supply - and a 2N3771 (biased by a resistor and a Zener to 28VDC) as "28V" supply. As you can see - the sides and the back of the bench are covered with sheet metal welded to the frame members - it makes one heck of a heat sink. The "raw" voltage out of the rectifier / filters is around 35 - 36V - and drops to 34 at full load. The "worst case" power dissipation is around 490Watts. A lot of heat - but you'll seldom notice it - as the bench - all steel - weighs nearly 300lbs. - and besides - we don't run the variable supply shorted by habit.
So - on to the Distribution section. This is fairly straight forward - six duplex outlets mounted under the bench for test equipment, etc. to plug into - these outlets are not switched nor fused - All of our test equipment is fused - and several battery operated - so need power for battery charging when not in use. The 28V supply transformer is plugged into one of the outlets - and is fused. There is an additional duplex outlet - this one mounted to the top of the bench - each outlet is the output from one of the variable transformers. This duplex outlet can be seen in the picture above.
|In the shadows it's hard to see the color - but the left set is "red" while the right set is "green." |
The other two duplex outlets serve as power connectors between the "Distribution" section - and the Panel. They are NOT wired as standard AC outlets. Each duplex is color coded (one RED the other Green), and the lower outlet of each has a cutoff screw in the ground "socket" to prevent the three prong plug being accidentally inserted into it. Actually - even if a plug were inserted in the "wrong" outlet - the worst that would happen would be blowing the 5A DC in fuse in the Panel, and possibly some blown light bulbs. The reverse polarity diode would conduct on the negative cycle of the AC line power. Nothing else would be affected. AC power comes in on the "Red Lower" outlet - and goes to the rear apron terminal strip - and on from there to the two fuse holders - one for each variable transformer. The "Red Upper" plug connects the two variable transformers to the outlet on the benchtop. The "Green Lower" plug brings in the +28VDC to operated the control circuits of the panel. The "Green Upper" plug brings in "earth" ground - and the output of the 0-30V supply pass elements back to the panel.
Moving to the Panel.
The left "section" contains a small 3A variable transformer - which is suitable for controlling the temperature of an Iron - or other light duty loads. It's control switch, fuse and volt meter are all in a handy cluster. To it's right - is a larger 8A variable transformer - that serves as the main "under test" power source. More about it in detail further on.
In the "center" are the controls, power jacks (banana jacks) and metering for the 0-30V 10A DC supply.
The right hand "third" of the panel has controls, regulators, and gauges for 0-125 Pounds air pressure, 0-20In. of vacuum, and a vacuum cleaner- the last item proved to be a bad idea - the stiff hose attached to a fitting at the panel proved to be too difficult to "maneuver" to be practical - and in fact - if I were doing this project again - I would leave all of the pneumatics out- compressed air is not that useful- it blows dust everywhere - the vacuum source had a lot of promise - trouble is - I haven't had a continuous solder sucker work for more than a short time before clogging and needing to be cleaned - and I've tried "heads" from Ungar to Pace... and while Pace's are certainly a huge leap better than the others - plain old solder wick is (for me) consistently faster and easier to use. I had assumed that when recapping an entire chassis that a vacuum powered solder sucker would be just great - reality proved to be something less. "The Bench" gets used nearly every day of the world - the pneumatic side hasn't been "powered" in a year or more. So much for the best laid plans!!!! Anyway - back to the subject at hand...
In between the two variable transformers is the Main Switch- which is an automotive type key switch. The Off position is just that - Power Off. The On position is the normal "running" position - but coming from "Off" - only supplies lamp power. This gives you time to glance over what will power up when power is actually supplied to the bench. Turning the Key on to the "Start" position - and a relay (K1) latches in - providing power to the control relays and other switches throughout the system. Totally unneeded - Totally neat. After all - that's part of "doing it yourself" - putting your signature on it. Once K1 latches - it will stay that way until the Key switch is turned off. The "Status" readout is provided by some very special switches.
|LEFT PICTURE: With the main switch turned from Off to On - the switches indicate what the various items "will" do... the #2 Variable Transformer is shown to be "set to" off - and is in fact "off". |
CENTER PICTURE: Pressing the switch to set to on - the left side of the switch shows that the switch is now "set on" - while the right hand side shows the unit is still "OFF"
RIGHT PICTURE: Turning the Switch on over to the "Start" position - the control voltages are enabled - and the bench comes to life - and the switch now indicates that the #2 Variable Transformer is ON.
|"THE" switches. The two switches that control the variable transformers are a bit extreme. Practically speaking - most people are not going to use switches like these. I got these surplus for 20 cents a pound years ago at the Guam Naval Station / Defense Supply Agency Surplus "outlet". Admittedly - at nearly $100 bucks - if I had to buy them now - I'd certainly use something else - but hey- like I noted earlier - I used what I had in my junk box. So what's so special about "these" push button switches? Not much - just four pole double throw Push - Push latching mechanism - four independently controlled lamps illuminating the changeable legend(s) in the push-button - in other words - NASA spec'd (and priced) switches. You could use four "boots" over the lamps - each a different color - and change the switches color; or use "position" of the lights to indicate status, etc. - or a combination of above. This is the option I took. Since you can't tell if these switches are on or off just by looking at them (look at them in the overall picture of the panel where there not lit) - it makes sense to indicate "off and on" with the lamps. The upper two lights - indicate "on" and have red boots. The lower two lamps indicate "off" and have yellow boots. But without power - you can't tell - and once powered - too late? What if we had a "two step" power on sequence - that lit the bulbs first - so you could turn whatever on / off as desired - then actually power-up the bench. That way when "lamp" power is first applied - the left side of the switches will indicate what position the switches are in - while the right side always indicates what the "real" power status is.|
|The Circuits - Schematics - and all that jazz... |
We've already covered the "Distribution" section- it's straight forward - nothing fancy -
AC comes in the plug - commons through the half-dozen duplex outlets - sources the 28VDC power supply, and passes power on to the "panel" through the lower "red" connector that is half of the left "connection" duplex. Note that obviously - these outlets have had the "commoning" strap removed - so that the upper and lower outlets are totally isolated from each other (and again - the "ground" lugs aren't used in these four outlets).
So--- it's on the main schematic -
Variable Transformer #1
Following along on the schematic - note that power comes in through the 3A fuse. S2 controls K2 - and the variable transformer (T2) feeds back through the "connection" duplexes - and on to the "left" outlet on the bench top. The meters (also surplus - and very weird) are 2.75ma. full scale with an internal resistance of 12Ω. So the 47K provides proper scaling - and 10K variable provide proper calibration "tweaking". The diode bridge (in my case 4 1N914s) rectify the AC for the meter. Since variable transformers put out a "mostly" sinusoidal waveform - the meter can be set to read a good approximation of RMS (though it's really reacting to average).
Variable Transformer #2
The right hand variable transformer is the part that will interest the most people. Well at least the RMS current sensitive "circuit breaker" that's in it's control circuit. But let's take it step at a time. The control (except K3's ground) is the same as above - as is the voltmeter.
|The Panel swung out. It's on casters - so it's easy to get in and out.|
|Close up of the circuit board. The voltmeter circuits are on the far right - the AC ammeter and Electronic circuit breaker next over. The components in the upper left are meter and other circuits to the 0-30V power supply. |
Since this is the "unit under test" power source - we're interested in how much current is being drawn - so an ammeter is included in this circuit. We run the AC through a 15A bridge - whose "output" has a .1Ω resistor across it. Why not just run the current through a resistor then use small rectifiers like the Voltmeter above? Has to do with the voltage drop across the diodes. Small signal rectifiers have about a half volt (.55V actually) across them at normal forward current levels. That means to indicate any current - our shunt would have to deliver at least 1.2 volts. Let's say that is what it delivers at 1Amp (i.e.1 ohm shunt) Ok - then at 10Amps - were dropping 10V - way to much - and 1) the shunt is hot (10 watts) and 2) we're not delivering full voltage to the load... Using high power diodes (or a bridge) each element drops roughly 1 volt - so a bridge drops 2V - but that's regardless of current. And if our shunt only drops another say .3V at 10A - That (.3V) is the only variation in output voltage that is current (ammeter) caused. If we take our voltmeter off behind the ammeter - then it indicates what the load is seeing. By placing the shunt across the output of the bridge - all of the AC current flows through the shunt - (now pulsing DC) and the diode drops aren't between the shunt and the meter. The meter is 2.75ma. full scale - 12Ω - so calculating: .00275*12=.033V. The shunt we are using is a piece of a 10Amp shunt out of an old Simpson 260 multimeter. The part we're using is about .1ohm - and we have the "tap" at about .03 ohms or so. So .03Ω * 10A = .3V. The meter needs .033V - so a little calculation .3V-.033V = .267V .267V/.00275A= approx. 97Ω. Since it's very hard to make "accurate" adjustments with the "tap" we just get it close (on the high side) and use the variable (VR3) to "tweak" it in.
OK Here's whatcha came for. Measuring average AC current is easy - RMS is a whole "nother" problem. Sine wave is bad enough - but inductive spikes and capacitive phase shifts can wreck havoc with conventional circuits. Getting a circuit that is forgiving of spikes - yet kicks out at the desired trip point is not easy. Fast reacting modern components (like LED opto-isolators) will react to the slightest provocation - such as any capacitive, inductive, or whatever disturbance -- and be more of a nuisance than help. That's why for this circuit we're interested in something that will act - and react like to RMS - not some artifact. In the early days of measuring AC - people didn't rely on opamps, charge pumps, and other techie solutions - yet they were able to measure RMS with great accuracy often far exceeding what we do with modern instruments. The true definition of RMS is that (voltage, current, power) which does the same work as the given amount of Direct Current. In other words - 110VAC RMS does the same amount of heating as 110VDC -- regardless of waveform, frequency, etc. Using that principle - the ancients made accurate AC meters by passing the AC through a given load - and seeing how much heat it gave off - that heat being translatable into work. Such a device is called a thermocouple meter. In the "usual" thermocouple meter - two dissimilar metals are combined (usually bismuth and antimony) are heated by a resistance wire through which passes the current to be measured. The most important point about this is it's immune to waveshape and frequency / harmonics. It delivers the same "reading" from DC through RF. On the bad side - it's sluggish - and very nonlinear - so that the meter scale must be made accordingly. For our purposes - we need these "good" points - and since were only interested in two specific "trip points" we don't need to worry about being nonlinear - nor "sluggishness" - in fact we take advantage of that last "benefit" so that our circuit will ignore very short spikes and other transients. Since we are only looking for "trip" points - and not looking for a "quantitative" measurement - we can use a "thermal converter" that is a lot more common - and very easy to work with. Radio Shack (and others) have a little light bulb that runs on 1.5V @ 25ma. It's filament is small - just about the right "thermal mass" to provide the response we would like.
|The lamp has the red and white wires sticking up. The Phototransistor is just below the heat-shrink tubing. Probably - would be better to use an "endfire" phototransistor - and use the heat shrink to lock them together. The Nichrome shunts (R5 & R6) can be seen in the barrier strip - Relay K4 is to the left. |
So If we place yet another "shunt" in the current path - say 1.2 Ohms - if we pass 1A through that shunt we'll get 1.2V across it - and if our lamp is also across it - the lamp (I20) will light - A photo transistor (Q1) placed close and in-line with the bulb - will start conducting - a transistor (Q2) it's hooked to will also conduct - pulling in a relay (K4) - who's contact's just happen to be 1) set to latch the relay "on" - and 2) the NC contacts are in the ground path of K3 - So here's the deal: A load starts pulling current - and it increases through 1 Amp. As the current passes 1A - the lamp (I20) glows - turning on the phototransistor (Q1) - biasing the relay transistor (Q2) "on" -- K4 pulls in and self latches - and at the same time opens the ground to the coil of K3 - which opens - and interrupts the AC flow from T3. Another set of contacts provide power to light the "tripped" LED indicator. Also note that since K3 has dropped out, one set of it's contacts has removed power from the "On" bulb and turned on the "OFF" bulb in S3. Depressing S3 removes power from both K3 and K4 - K4 drops out (the "tripped" LED goes out). Depressing S3 once more causes K3 to pull in - again supplying power to T3. Notice the switch S4. It is a double pole - double-throw - CENTER OFF switch. In the Down position - it completely shorts out the shunt. No voltage across the shunt - no breaker action. Center position (OFF) works as described above. Up position shorts out only a portion of the shunt - leaving .325Ω - so the circuit now trips at 4 amps. Notice that S4 should be rated 10A. The shunt - R5 and R6 is actually a piece of nichrome wire - that is mounted in a terminal strip - with an appropriate "tap" under one of the screws. By adjusting the sizes of R5 and R6 - any trip points you like can be set (and you can add more if you like).
OK that's all fine and dandy - so what does the little bridge above the shunt do? - and is that really a dead short between the DC+ and DC- terminals? Yes -- that is the correct circuit. Remember - the bulb is designed for 1.5 volts. A couple or more volts and ZAPPP!! so... that is a 6amp bridge across the shunt - and the bulb. 6A bridges are dirt cheap - and like all good bridges - they have a (usual) voltage drop of 1V per element - the bridge will start conducting at 2V - which should save our bacon - er - bulb! It's just insurance for our bulb.
I hear that comment / question - "If we didn't want to use an ohm before - how come it's ok here?" Because the extra voltage drop introduced in this additional circuit will never be more than 1.2V - any higher and the "circuit breaker" kicks in. So regardless of 1A or 4A - the greatest additional drop will never be more than 1.2V - a long ways from 4V (1ohm at 4A) or 10V (1ohm at 10A). And while 1.2 volts isn't great - it's tolerable.
OK the last "item" of the bench. The 0-30V 10A DC power supply is the Lab supply circuit published by National Semiconductor - and can be found in their document LB-28 - which you can view / down load from their site. The URL is
|Returnto the Lobby|
|Copyright © 1996, 1999 Randy Guttery|