The AM Forum

I am in the think-tank mode for modifying my remote antenna tuning system so I don't have to run out to the tower every time I change bands.  I found a couple of identical, big-arse dual wafer ceramic rotary switches, each about 5" in diameter with 10 positions, indexed with spring loaded cams.  I would need to gang the two together and figure out some way to rotate the whole thing remotely. I think a reversible DC motor would be too difficult to configure so that it would automatically stop precisely every time at the proper position for each tap position on the switch.

I have seen mechanisms that rotated with a latching cam that looks kind of like a rotary saw blade.  A solenoid with a hook on the armature engages a tooth on the cam, and when activated, pulls the cam down one notch, then releases it. I would like to find something  like that ready-made or at least modifiable to work in my application.  Otherwise it will involve homebrewing from scratch, carving the parts out of hunks of solid steel, not exactly an easy task.

Anyone have anything lying around in the junk that might be made to work, or know of something easily available that would fit the bill?

Ledex has traditionally been the primary supplier of these solenoids. 
http://www.ledex.com/rotary-solenoid/rotary-solenoid-comparison.html?gclid=CIaI-4r30aACFdk55Qodh0UC1Q
I'm sure I have a couple in the storage unit, but I have some doubts about them rotating a 5" diameter switch, since they are attached to a ~2" multi-pole gang of switches.  Do you have a required torque spec on the 5" switch?

73,
Mike  KC4TOS

Hi Mike,

That is a wonderful website you have created, pertaining to the B-17 and B-29 aircraft, and early Army Air force avionics equipment. Very nicely written, too.

http://aafradio.org

Anyone who is interested in aircraft and airborne electronics equipment of the Second World War period is encouraged to take a look.

I especially enjoyed the information and interior photos relating to the Enola Gay; this is, of course, the aircraft that dropped the atomic bomb on Hiroshima.

I did not know the Enola Gay had been moved from the Smithsonian to the museum you had referenced in Dallas. I may visiting Dallas later this summer; if so, I intend to pay her a visit.

Thank you for making this information available, Mike.

73,

Bruce

My pleasure, Bruce. 

Regrettably, it's at Dulles, VA (IAD), not Dallas, TX.  It's about 18 miles west of the downtown National Air & Space Museum, and is known as the Udvar-Hazy Center - http://www.nasm.si.edu/UdvarHazy/

clothes and dish washing machines of the 80'sand 90's have solenoids and rotary switch timers and other parts that may be useful to your application....

a power screwdriver, with a worm gear driving a standard gear, plenty of torque, reversible, and with a little ingeniuety you could put an indicator or stop on it.

Probably what you want is something like the TV antenna rotators with the repeater indicating switch position in the shack?  they arent' THAT big and have plenty of torque.  You could easily calibrate the dial face to show the switch positions. 

Just thinking out loud.

Thanks for the clarification, Mike. How I confused Dulles, VA and Dallas, TX is beyond me; I guess I could use age as an excuse.

The next time I am in that part of VA, I plan on scheduling a visit to the Udvar-Hazy Center. I would really enjoy seeing the Enola Gay, realizing of course, that the likelihood of getting inside the airplane is probably close to nonexistent.

Keep up the fine work of restoring the Enola Gay's avionics and interior. As an example, that kind of work makes the restoration of a '67 Corvette big-block to factory original/as-built condition seem like a veritable walk in the park. I'm fairly sure there are very little authentic OEM or reproduction parts available for B-29 restorations nowadays.

73,

Bruce

You would not believe how many signatures it takes for me to get inside to install something I've restored.  Once the "artifact" is turned over to the collections staff by the restoration staff, it becomes a big deal.  Just part of the bureaucracy...  That's one reason I took so many photos of the interior and put them on the website.  On the other hand, the practicality of having thousands of grubby little hands rubbing the gubbins in the interior of a priceless aircraft should be obvious.  There are a few museums who have tours through one or two of their airframes and the yearly loss to theft and vandalism is just something they tolerate.  I wish there was some way to limit access to just hams or some other criteria, but the political consequences of such a proposed action is fairly predictable.


That's actually how I got started in restoration work.  I used to restore Corvettes!  My last one was a ground-up '67 big block.  I was afraid to drive it when it was finished.  Also, the thought of getting under a car in my later years made me shift to old radios, which you could actually use without fear of someone scratching it or rear ending you at a stoplight... ;D  Amazingly enough, I still pick up an occasional part for WWII aircraft avionics on ebay.  Recall that there were hundreds of millions of parts produced during the war, and there are still old warehouses here and there that have undiscovered corners.  I found an original aircraft ash tray in an old "empty" Glenn Martin plant in Baltimore the same way - http://aafradio.org/flightdeck/Peripherals-misc.html

Don, what do you want to switch?  you have, I think it was, two 10 pos. dual wafer switches?  do they have stops?  does the contact make with each tap one at a time or connect the current one plus all previous as it goes around?  Are you switching coil taps, feedlines, .padder caps, all of the above ...the reason for all the questions is it might be that rotary switches are not the best way to go.  Depending on what you want to do, you might have it easier and more reliable to have a switch network of relays.

Rob

That would be my preference as well, Rob.  Back in the dark ages of my youth, if you wanted to switch something remotely, it was the only reasonable way to go.  A rotary switch is inherently a manually rotated device, though I can understand Don's reluctance to abandon such beautiful switches.

I got over to Dulles the first year they had it open - WOW! I got to see the Enola Gay which is suspended overhead but you get to be 3 feet from the nose and just above the canopy so you can see right in close up.

They had lots of crazy stuff like the last SR-71 - the one that set the record on it's flight east on its wheels.

And they had the Enterprise OV-71 the first orbiter on its wheels.

And Air France donated an awesome SST which is somehow hanging in mid air.

This place is BIG.

And they had a lot of classic foes nose to nose like Zero vs P40 and MIG 15 vs Sabre and so on.

Mike WU2D

I have 5 different tuners that have to be switched, both input and output, some of them feeding open wire tuned feeders at a high voltage point.

Right now I do it with one manual rotary switch to select the input and the output is selected using a series of  DPDT knife switches thrown in a sequence to  form  kind of a logic circuit.  I  would need several high voltage ceramic insulated relays with wide spacing between the contacts, all with the same coil voltage.  Don't have anything like that in my junkbox, and a complete set would probably cost a fortune and take for ever to round up.

The wafer switches have spring loaded stops, so conceivably a shaft coupling with a certain amount of flexibility would get them close enough that the indexing stops would seat in proper position.  Coupling the shafts together would allow me to remove the stops from one of the  switches if it took too much torque to operate both.

I assumed that from your first post.  Removing the stops from one switch is probably essential.  The Ledex style solenoids have quite a bit of torque, but you would have to try one of them to see whether it was enough.  The indexing is also something that will need to be looked at - they have a whole series of rotor angles that the solenoids are designed to turn.

It's mounted on a pair of $100,000 scissors jacks to get it up in the air. ;D The catwalk was a deliberate attempt to get you as close as possible, along with the extensive fiber optic lighting to allow you to see the inside lit up.  It's a shame the crazies forced the museum to mount the plexiglass wall between the public and the plane.  The unloaded weight of the aircraft makes for some interesting effects - the balance is normally nose heavy, but in its current condition, I've felt the nose wheel rise when I've been in the radar compartment.  (I've gone on a diet since then, so no nasty comments, please...) ::)



The landing gear on the Concorde really jacks it up in the air an amazing distance, but it's not actually hung on cables.  The arched design of Udvar-Hazy was designed to hang a lot of tonnage, but it wouldn't support a Concorde...

Mike,
That wasn't the last SR71. That was the last one that made the news.
Yeasr later I heard one flying around Palmdale and a guy I work with saw it.
Bad ass sounding motors.

Back in the '70's my brother and I went to the local SAC base.  Early morning, nothing to do.... Got to watch a '71 do touch-and-goes....

The last time around the pilot musta known a bunch of JNs was watching. He lit the afterburners, and she went straight up and up and went away......

klc

Don: I have been using large 24V solenoid type RF relays from Russia. They are available from a variety of the vendors on E-pay, QRO parts, etc. They work great for switching antennas and RF tank components. Picture attached
Since you already have rotary switches they are not all that hard to utilize. I assume the ones you have are similar to the T 368 bandswitch and have the really stiff roller follower that rides on an alloy or nylon camwheel. To drive this with a dc motor you must gang an additional wafer with the same indexing to run indicator lights. Use a DC motor like a headlight door (best) or window motor coupled with a LOOSE chain drive. When you do this the motor rotates the switch until the camfollower snaps to the next detent and the chain goes slack, allowing you time to get off the motor switch. This assures that the switch will stay solidly in the detent and contact will be made properly. Drake did this on the remote antenna switch accessory for the "C" line. I have one of the remote boxes if you want to borrow it to check out the operation. If you do not do this trying to time the "coast" on a DC drive will drive you nuts.
Skip

Don... Car door window motors....car antenna up and down motors....sunroof actuators..windshield wiper motors (that can move a rotary actuator 1 click with each rotation.......etc   12 volt and available at junk yard prices or free in some cases...lots of possibilities in the older auto world of parts...  good luck   Steve

Don....I've not tried this but the windshield wiper motor with the associated controls that allow the 1 wipe cycle might work very well....1 wipe cycle for each click of fhe rotary switch...a simple peg and cam system ..and windshield wiper motors are high torque.......just a thought

How about old phone-type stepping switches? They have a lot of torque and many small steps. I have some that have contacts about half way around, and I don't know about how easy the shafts are to couple to, but they are strongg enough to turn other switches. I could take pics if you are interested in thinking about that possibility. otherwise the only thing I could suggest would be gearmotors and use reed relays to set the 'detents". There is an article in QST April, page 44 about this right now.

Those look like they would be exactly what I need, except my switches are indexed at 36° per step and the closest thing they list is 35° per stroke. They mention something about building a special unit to order with custom stroke angle, but I suspect that would be way out of my price range. The 35° ones are close, but 1° error per step would accumulate too much error over 5 steps for the switch contacts to line up properly.

The round part of the ceramic wafers are actually 4" in diameter.  The 5" figure includes "ears" or protrusions on each side of the wafer, making it more like an oval shape, to accommodate the mounting hardware. With the spring removed from the detent cam, the switches rotate with very little torque. With the spring engaged, it takes considerable torque, but I haven't attempted to measure it.

I have several 12 vdc reversible gear motors that turn at about 1 rpm.  I used one of these, geared down further with a worm drive, to rotate the tuning cap on my 160m dipole tuner (before I  ran over the control cable with a lawnmower).  The torque is great enough that I can't stop it from turning by hand.  The problem is figuring out how to make it stop exactly for each switch position. On the 160m tuner, I used my el-cheapo SWR meter and and let it run till I hit a null on reflected power. But on this thing I want to be able to just hit the button and have it stop precisely in position each time.

I looked over the QST article, but I doubt the reed switches and magnet idea would be precise enough to make it always engage the detent cam precisely on.

Would not programmable stepper motors be the way to go for such a project?

Hey Skip nice switching arrangement. What is the value of those RF chokes strapped to ground?

Agree.  There would need to be a careful balance of detent spring pressure and solenoid power to work properly, as the last 1° should easily be taken up by the detent mechanism.  The solenoid will then simply reset for the next step.  I didn't mention it before, but you also have to have a switch section in the mix that will be the inverse of the selector switch in your shack so that when the switch gets around to the correct position the solenoid stops energizing.


A cheap fishing scale on a lever arm should allow you to measure the torque with several different spring pressures to see whether it's within the published capabilities of the solenoid.


Exactly!  A servo system can be designed to do it very precisely (there is a lot of CNC machining equipment out in the marketplace that do this) but the feedback loop with a position sensor is an added complication for something that should be fairly simple.  If you have access to a machine shop, I think it would be fairly easy to cut out an escapement wheel with ten sawtooth positions and then use a linear solenoid with whatever stroke is needed for the radius of the wheel.  Larger radius = less force required, but longer stroke, of course.  I think that's how I'd approach it, but I have a pretty complete machine shop in the outbuilding.  Anyway, it's something to think about....

What about setting the switch so the solenoid is 'spot on' at slot #3.  Then you'd only be out a degree / maybe two at each extreme....

Just a thought...  Might make the 35 degree actuators usable in the 36 degree switches.


SR71s still fly out of Edwards rarely.  Usually at night, they have a telltale VERY distinct sound.

Then they fly around in circles waiting to be fueled.  What's REALLY cool is when you hear a plane, and a half minute later, you look up and see 'donuts on a rope' going across the sky.

--Shane

Just another thought, Don...how far is it from the shack to the tower?

Something like http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&item=260574075127 might be another alternative if you're running your cables in plastic conduit.

Plenty of Ledex stuff on there as well.  Found this for $6, not sure if it's 33 degrees or not:

http://www.tradebuyerseller.com/electric-solenoid-/-ledex-/-new-/-nice-/-p.n:-7810/290376359342/12576/

Don: the relays are about $20 each including shipping from either the guy in the Ukraine or UR4LL in St. Petersburg. They will usually negotiate on a quantity deal. I think I have bought 20 or so over the past several years. Relays are easy to deal with and these even have aux contact to run an indicator light.
I can e-mail a link if you want to investigate further. I also have a multi wafer dc motor driven switch assembly that I purchased from a surplus place on a Korean trip if you want to try that. It was actually built for that purpose. I suspect it came out of some military equipment.

Dave: I think the chokes (static drains) are a 1 Mh. They look suspiciously like Valiant plate chokes !
The picture shows "version 1" and I have since rearranged the tuners. I don't have the room for decent antennas so I tune what I have at the feedpoint.
Skip

What kind of contact spacing and voltage rating are these?  I can't really tell much about them from the photos.  I assume from the picture they are single pole, so two relays would be needed for each tuner output to line.  Are they single throw or double throw?  I need some with at least 1/2" spacing between open contacts, since some will be switching the high voltage point of open wire tuned feeders, the kind of thing for which you would normally use a big-assed knife switch (as I am presently using).  They would also need to handle several amperes of rf, since others would be switching the line at a high current point, as well as the input line from the shack at up to 3 or 4 amps or more, depending on tuner settings. It appears I am talking about a total of at least 20 relays (10 each if they are available in double pole).

 



I might be interested.  How many poles and how many positions, and what kind of contact spacing?  What kind of insulation?

I may just stay with manual switching to change bands, and gang all the variable tuning caps together with the reversible motor so that I can at least make a  major QSY within a band without a trip to the  tower.

Another problem with the switches I have is that they are not constructed with the shaft extending rearward so you can gang them end to end.  I had planned to couple the front shafts together and turn them with a National 90° right angle drive, similar to the tuning capacitor configuration using the worm drive in the HRO.  That way, I can gang one two-wafer switch that selects the tuner to the line from the shack (now coax but maybe later open wire line), to the other to select which tuner feeds the open wire line that goes up the tower, but I still need another switch to select either the open wire line that goes up the tower to the horizontal dipole, or the single wire that attaches to the base of the tower to form the vertical.  I have a couple of additional identical wafer assemblies, but mechanically coupling one of those to the assembly just described would be a Rube Goldberg mechanical nightmare that would probably be flaky even if I initially got it to work.

Don, I have a bigassed TMC remote tuna that has a motor driven rotary switch.  Here's how they do the indexing:

There is a disc on the same shaft as the rotary switch with a cam profile.  The dips in the cam correspond to the switch stop positions.  A microswitch senses when the shaft is at the dip, and shuts the motor off.  To start the motor, there is a pushbutton back at the control box in parallel with that microswitch - once you get the motor started, the cam profile is such that the motor then keeps moving until it hits the next dip.  So you just momentarily hit the button to advance the switch one position, or hold it down to keep it moving.

One section of the rotary switch has a resistors, wired so they are progressively in series.  A regulated DC voltage is fed thru them, to a milliammeter on the control box so you can see what position the switch is in.

Of course you'll have to remove the switch detents.  The only real tedious work required is to machine up the cam disk and get the microswitch aligned correctly.

I like that!  Don could use one of the existing detent cams on either of the two switch assemblies.  Just mount a microswitch with a roller on it where the spring assisted detent follower currently resides.  You will still need a small added wafer to tell you which position the switches are in using an array of small lights or LEDs in your control panel.  Great idea, John!

Bird made some motor driven rotary coax switches using the cam and microswitch method.

Ive also seen stepper driven big Variacs.

With access to a machine shop anything is possible.

Carl
KM1H

It doesn't need to be a cam with a detent it coulld also be a disc with say allen cap screw heads to catch the microswitch. Just allow enough slop in the system so the switch lands in its own detent when the moror stops

Thanks for all the ideas so far.  I have some n.o.s. microswitches somewhere in the junkbox, complete with the little roller cam assembly.  I could remove and modify one of the detent cams to make it run the microswitch.  Not sure if I would need to remove the other detent cam.  It might help make sure the switch was properly indexed and seated each step, since the only satisfactory 90° angle drive I have has a small amount of slop due to backlash.

Instead of the added complication of a feedback circuit to indicate position, I might just use the  receiver to indicate if an antenna is connected, and count steps to  determine which tuner is selected, since I will be using only 3 of the 10 positions on the dipole, and if another wafer section is Rube Goldberg'ed into the assembly for the vertical, I'll use only 2 of 10 positions of that one.

First, I'm going to have to completely rebuild the dog-house.  The existing one is falling apart, with the original plywood almost completely de-laminated, plus wood rot and termite damage. I don't think it will make it through another season.  Believe it or not, it was made from an old broadcast radio floor console cabinet dating back to the late 30's or early 40's.  It had a 78 rpm turntable plus an AM/FM  radio that covered the old 40 mHz FM band.  I covered the top with flashing and the sides with asphalt shingles to protect it from the rain, and nailed on a sheet of plywood for the back "wall" and made wooden doors for the front.  This was supposed to be a temporary JSment, to last a couple of years until I could build a good one.  It has lasted about 28 years instead.  Besides falling apart, the existing one is too small to accommodate an additional tuner I plan to add. After replacing the shelter I can then think about revising the switching set-up. I'll build the new dog-house with pressure treated and/or synthetic materials.

Go high tech, just a disc with a hole every 35* and add a few LED's and a photo transistor outputted to the control circuit. Use a motor geared down to some comfortable speed.

I just gave myself a bright idea for switching input networks when using oddball 36* index Centralab JV series switches. ::)

Carl

What exactly are those?  36° is what my switches are indexed, not 35°

Sorry, my goof, I was referring to the big Radio Switch Corp Model 88 switch. Now Multi Tech Industries. I used to sell loads of those and Model 85's to serious QRO builders ;)

But here is the Centralab info anyway.
The old JV 9900 series Centralab ceramic power switches as used in many, many commercial amps from Alpha, Dentron, Ameritron and almost all others in the 1500W+ category. It uses an oddball 20* index which makes it hard to mate a series phenolic switch to control a relay switched input board.
Currently the Electroswitch E-4 series since 1992

http://www.electro-nc.com/rotaryus/e4.pdf

Carl
KM1H

Hey, I neglected to notice that Don's switches already had cams  ::)  That's an even better idea, using the existing cam!

Don, ya only need one extra wire (and another switch bank, of course) to run the meter position indicator.  You could also just bust the stops off a pot and drive the pot with the same shaft.

Don,

One way to go is the Zenith TV "clicker tuner" method... basically all you need to do is to remove the detent mechanism (or use a much much lighter spring, or modify the depth of the detent gear...) and add a wafer that has the same degrees of rotation per contact as does the main switch. Then all you need is a simple "up down" motor (DC). The added wafer does the sensing. When you hit the up button the motor turns on, and keeps going until it hits the next contact, then it stops, or if you hold the button, it keeps going until it hits the last one which is wired so that it can then only go down...etc... clever system.

You can rig an indicator with two wires, just add one more wafer and that can be a voltage divider or half of a bridge (same idea) that can read out on a meter or whatever you rig up...

              _-_-bear