35GHz Magnetron...

[Alexei.Polkhanov]

I would try actually. Cook large potato with 2.4GHz for 8 minutes and then add some crust on top with 35GHz - could be an idea for new product? Mmmmmmmmnyam ....

[tron9000]

as promised: some photo's. These are bloody heavy!

[Alex Eisenhut]

How about a microwave oven...

It wouldn't cook food in the same manner as a conventional microwave oven. The higher frequency radiation would be absorbed nearer to the surface of the food so it would be across between a grill and a microwave.

Now let's start a KS campaign for this and let's call the product " the grillwave ".

[Jay_Diddy_B]

Hi,

I have attached a datasheet for the Magnetron.

Regards,

Jay_Diddy_B

[T3sl4co1l]

Yummy, looks like solid cast Alnico magnets, I would guess!

Curious about the "heater voltage operating".  Is there that much cathode bombardment that they have to reduce heat, even if it's a mere 10W average?

Also curious if it requires hard vacuum (seems really tricky) or high pressure SF6 (??) to keep that teeny waveguide from arcing over.

Tim

[Jasterisko]

If it is for sale, I want it.
I would be not surprised if the internal vacuum is just bad, so it may need special procedure to test and eventually bake out and re-evacuate it. Reactive internal getter used by Sylvania these days is not reusable.
As far as lethal - at 10Watts average power you can only get from it a due to very small cathode, RF thermal power is far from anywhere being lethal, unless you will hold the output port right in the front of your eye, but most likely it will not do too much damage for intermittent exposure.
Driving this thing in solid state way is somewhat more complicated.

[khach]

Look for russian MI-88 magnetron [url]http://www.ostron.de/Radar-GHz/Bauelemente-bis-SHF/Magnetron-Impulsmagnetron-MI-88M.html[url]

[PA0PBZ]

Look for russian MI-88 magnetron [url]http://www.ostron.de/Radar-GHz/Bauelemente-bis-SHF/Magnetron-Impulsmagnetron-MI-88M.html[url]

That is a real angry tube:

[miguelvp]

Look for russian MI-88 magnetron [url]http://www.ostron.de/Radar-GHz/Bauelemente-bis-SHF/Magnetron-Impulsmagnetron-MI-88M.html[url]

That is a real angry tube:

Looks more like an angry waiter to me, might be because the napkin on his left arm is hot and burning his forearm. Or maybe because they only ordered a small bottle of wine so his tip won't be as much.

[BennVenn]

If it is for sale, I want it.
I would be not surprised if the internal vacuum is just bad, so it may need special procedure to test and eventually bake out and re-evacuate it. Reactive internal getter used by Sylvania these days is not reusable.
As far as lethal - at 10Watts average power you can only get from it a due to very small cathode, RF thermal power is far from anywhere being lethal, unless you will hold the output port right in the front of your eye, but most likely it will not do too much damage for intermittent exposure.
Driving this thing in solid state way is somewhat more complicated.

Many magnetrons in service at work sit on a shelf in a warehouse for many years, manufacture dates in the early 90's are common. There is usually a specific run-in routine at time of install. Some of the more exotic tubes need power applied at all times to keep an ion-pump active. Even these need a fairly lengthy alignment/run-in procedure before full power operation.

As for solid state, A magnetron of this size is well suited to a SS driver. 25us at 40kw is well within the capability of a homemade pulse transformer. Off the shelf Furuno navigation radars are available up to 60kw and are all solid state (except for the magnetron of course)

Dry air is usually good enough at these power levels even with such a small waveguide. With an average output of only 10W it would be safe enough to experiment with directional phased array. There is some pretty cool technology developed around WW2 for passive pulse steering

[tron9000]

Hi,

I have attached a datasheet for the Magnetron.

Regards,

Jay_Diddy_B

Cheers Buddy! Much Appreciated!

Yummy, looks like solid cast Alnico magnets, I would guess!

Curious about the "heater voltage operating".  Is there that much cathode bombardment that they have to reduce heat, even if it's a mere 10W average?

Also curious if it requires hard vacuum (seems really tricky) or high pressure SF6 (??) to keep that teeny waveguide from arcing over.

Tim

Yeh those magnets pulled in every loose screw I had on my bench nearly! I really am an RF noob, so I'm not familiar with vacuum in wave guides and barely remember my wave guide modules from uni! I'll have to do some reading.

If it is for sale, I want it.
I would be not surprised if the internal vacuum is just bad, so it may need special procedure to test and eventually bake out and re-evacuate it. Reactive internal getter used by Sylvania these days is not reusable.
As far as lethal - at 10Watts average power you can only get from it a due to very small cathode, RF thermal power is far from anywhere being lethal, unless you will hold the output port right in the front of your eye, but most likely it will not do too much damage for intermittent exposure.
Driving this thing in solid state way is somewhat more complicated.

I'm not sure about selling it at the moment - it was given to me in the premise that it was to be "thrown out" i.e. it should be on its way to landfill! I'd have to do some asking about the in's and out's if I was to sell it: I don't want to loose my job if my boss saw it up for sale on ebay for example!

If it is for sale, I want it.
I would be not surprised if the internal vacuum is just bad, so it may need special procedure to test and eventually bake out and re-evacuate it. Reactive internal getter used by Sylvania these days is not reusable.
As far as lethal - at 10Watts average power you can only get from it a due to very small cathode, RF thermal power is far from anywhere being lethal, unless you will hold the output port right in the front of your eye, but most likely it will not do too much damage for intermittent exposure.
Driving this thing in solid state way is somewhat more complicated.

Many magnetrons in service at work sit on a shelf in a warehouse for many years, manufacture dates in the early 90's are common. There is usually a specific run-in routine at time of install. Some of the more exotic tubes need power applied at all times to keep an ion-pump active. Even these need a fairly lengthy alignment/run-in procedure before full power operation.

As for solid state, A magnetron of this size is well suited to a SS driver. 25us at 40kw is well within the capability of a homemade pulse transformer. Off the shelf Furuno navigation radars are available up to 60kw and are all solid state (except for the magnetron of course)

Dry air is usually good enough at these power levels even with such a small waveguide. With an average output of only 10W it would be safe enough to experiment with directional phased array. There is some pretty cool technology developed around WW2 for passive pulse steering

some useful information there, thanks - but like I said will have to do some revision, get some other projects out of the way, and I'm right in the middle of moving house at the moment, so unfortunately there may not be an update for a while

but many thanks to all those who offered suggestions and information thus far, its been a real help!

[TerraHertz]

Here's a nice collection of antique magnetrons: http://www.tubecollection.de/ura/magnetrons.htm

And here's one I have. A Philips JAN CV3602
Brand new, never been used, still has the waveguide end cap. Which screws on and is too tight on the threads to unscrew by hand. Hence no pics down the coaxial waveguide in this set.

Some info from http://www.radiomuseum.org/tubes/tube_cv3602.html
Tube type:     Pulse magnetron
Identical to    CV3602 = 5J26
Filament    Vf 23.5 Volts / If 2.2 Ampere / Half indirect /
 Tunable magnetron, 1220 to 1350MHz, 600KW peak pulse. Coaxial out, external magnet.
British spec for WE 5J26 magnetron.

I can't find data on the JAN CV3602, but there's a lot around for the 5J26.
Too bad I have no magnet for it. 

[tron9000]

that looks an interesting piece of kit!

Wonder if they would like the spare one for their collection? Or maybe cos I'm learning, should I have a spare one.

So it summarise from the spec sheets typical rating for my Magnetron, a list of things to achieve:

• locate the anode/cathode and heater element wires
• determine if the waveguide cavity needs a vacuum and a suitable end cap
• develop a circuit that will heat the heater with 6V at 2A for a while or until it has reached optimum temperature and then ramp down to 1.6V & maintain a specified heater temperature
• develop a circuit that can deliver a pulse to the anode/cathode wires up to 12kV in a 0.25microsec pulse.
• develop a method of measuring output power safely
• construct a test-rig with safety systems and interlocks - have to bear in mind that i'll be doing this in a garage probably, plus a curious 3 year old wandering about - I intend to be conducting these experiments after bedtime!
• start experimenting with ramping the output power up to its average rated power and then peak power - peak power I highly doubt I'll reach, but will be fun to try. I understand that a little power in RF terms can go a long way if directed properly,

My goal I suppose is it see if I can get it up and running. Then think about an application. But before all that, I need to clear some projects out the way first. As I said; it'll be a while before this one starts, but at least I have a plan!

[T3sl4co1l]

Anyone know if magnetrons are linear power?  I know you can throttle and tune an average one by varying the current (apparently, the V-I curve is not power-law as you'd expect for being a vacuum diode, but more like an avalanche diode -- the threshold voltage corresponding ultimately to the spiral electron path and resonant frequency), but I don't know how far that holds true (we're talking orders of magnitude here!).

If it's the case, then I would suggest, heat it with a 6.3VAC transformer (doesn't have to be anything special at all), and apply some high voltage (maybe with an NST, a stack of diodes, and some high voltage capacitor, using the typical circuit a microwave oven uses).

You'll want to buy/construct a waveguide to horn transition.  Construction will be fairly straightforward if you have a machine shop, just follow the dimensions of the flange and cavity.  Bolted together aluminum should be fine.  I don't think the horn dimensions are very critical; it should be easy to find some standard shapes (angles and lengths; even if you find a drawing for one at a different frequency, just scale everything linearly, no worries).

If that's a bit beyond your capability, you're probably better off buying.  I wouldn't suggest attempting it with, say, cardboard and tin foil, not for something that small and precise.

You won't want to run it very hot for very long of course.  As a vacuum tube, it should be perfectly happy running 10W continuous (assuming it even oscillates at that level), or 100W at some duty cycle, or 1kW at even less, etc.  Unless there are specific physics reasons I don't know about (and as much physics as I know, I don't know the subtleties of magnetrons), that should be just fine.  You're dealing with thermal time constants, so you should be able to cook it for a surprisingly long time before it's a problem.

As for a detector, with a low average power rating, just seeing if it heats something up isn't going to be a good test.  But something like a neon tube with formed lead lengths should light up quite visibly.  Let's see.  You want the leads 1/2 wavelength from end to end.  At 4.3mm, the electrodes inside the bulb alone are probably enough, external leads cut flush to the bulb.  Or 12.9mm (~1/2"), enough for some thru-hole lead length on a salvaged bulb perhaps.

Tim

[BennVenn]

I think there would be a threshold where the electrons would have enough energy to overcome the influence of the magnetic field. Insufficient voltage would give you current flow but purely as a diode as the plate would be some unintended part of the body, not the resonant cavities they were intended for.

As for the horn, the dimensions are more or less critical. Its purpose is impedance matching to the air. Its been a while since our RADAR theory courses but I believe speed of RF in a waveguide is 0.9 that of light and it has a lot to do with it.

If you can find a TR cell, you can use that to detect RF power, they are usually on the edge of ionisation either from a radioactive gas within, or a voltage applied across an arc gap. Any RF coming through them will cause a breakdown and can be detected. These are used to prevent RF from the magnetron entering the sensitive receiver at TX time.

[T3sl4co1l]

After some reading (quite a bit more than skimming the pages of one of the MIT Rad Lab books -- man, they're dense -- and *useful*!), I suspect the V-I characteristic is roughly exponential, but the toe-in is probably diode-like.  There's also an oscillation threshold.  And mode hopping (depending on how good the design is), which changes with respect to V, I, B and time -- seeming similar to the misbehavior of laser diodes and other coherent oscillator components (which produce little to no output, or incoherent noise or mode hopping, outside of their ideal range).

V and B are roughly proportional, so that you could get a much lower oscillation threshold and output power level (even going CW) if you had a variable electromagnet, and simply turned it way down.  But, er... so, yeah.  Pulsed it is. 

As for the horn, the dimensions are more or less critical. Its purpose is impedance matching to the air. Its been a while since our RADAR theory courses but I believe speed of RF in a waveguide is 0.9 that of light and it has a lot to do with it.

Yeah, you want to avoid too much SWR.  I shouldn't think the dimensions are terribly critical (i.e., accuracy better than 1/10th wavelength -- that's "marking with chalk" to a machine shop) in order to keep it within, say, 1.5:1.  More than that and I'd be worried.  Don't know what SWR is actually tolerable or desirable (for SWR > 1, the waveguide length will matter, as it will act, at least in part, as a resonator, loading the cavity with some phase shift that will pull the resonant frequency somewhat, or make it hop modes or something), but most anything can handle that, so I can't imagine it'll be *too* bad.

If you can find a TR cell, you can use that to detect RF power, they are usually on the edge of ionisation either from a radioactive gas within, or a voltage applied across an arc gap. Any RF coming through them will cause a breakdown and can be detected. These are used to prevent RF from the magnetron entering the sensitive receiver at TX time.

Beware, the purpose of a TR cell is, of course, to reflect power.  So if it's sealed right to the magnetron, you're shorting the output.  And there'd be nowhere for the 40kW pulse to go if that's the only load.  Which, I think, would result in excessive cathode bombardment.

The whole reason such a small device can handle so much current in the first place -- and also why the heater power is reduced in use -- is because a substantial quantity of electrons are accelerated by the field and thrown back at the cathode, whacking out tons more electrons, greatly enhancing its current density without causing destruction.  I would fear that, if the RF were fully contained, it would tend to spark and destroy the cathode surface, resulting in quickly degraded performance and early failure.

Tim

[BennVenn]

Beware, the purpose of a TR cell is, of course, to reflect power.  So if it's sealed right to the magnetron, you're shorting the output.  And there'd be nowhere for the 40kW pulse to go if that's the only load.  Which, I think, would result in excessive cathode bombardment.

True, but to detect if the magnetron is CW capable at 10W it should be OK. A neon should also work ok but would probably require the horn being fitted or not much power would leave the waveguide flange.

I've wondered the consequence of high SWR in a magnetron. The output is coupled from only one cavity so I would assume the resonance would either be destroyed or amplified. Amplification would lead to arcing between the entry points of the cavities? I suppose one way to find out is to run a junk microwave cooker without a load till failure then do an autopsy...

I read somewhere that the original magnetron design was built using the tooling for an old pistol revolver drum?

[tron9000]

I read somewhere that the original magnetron design was built using the tooling for an old pistol revolver drum?

All I was told is it was first commercially made in Sheffield, just down the road from me, so probably was made from old weaponry tooling, since it was around the time of the 2 wars and there was probably plenty of spare tooling floating around. Seems fitting judging from the amount of metal that's in them that it was made in the city of steel!

There's lot of reading to do here, even though both BennVenn and T3sl4co1l, you are both stating horn dimensions are less critical, I'd still like to find out exactly how to design it. I do have access to a metalwork shop, so I could CAD something simple up.

I just found this brief basics guide - http://www.radartutorial.eu/08.transmitters/Magnetron.en.html#abs9 - I think I have a better understanding now of how it works as a diode: essentially your firing electrons off the cathode and stopping them from reaching the anode with the magnetic field, since mine is a permanent magnet, there much be a certain V&I to get the sweet spot (red arrow in figure 5).

but I'm pretty sure its aperture coupled (fig 15), so definitely needs re-vacuuming!

[PA0PBZ]

but I'm pretty sure its aperture coupled (fig 15), so definitely needs re-vacuuming!

Read your own link: Aperture, or slot, coupling is illustrated in Figure 14. Energy is coupled directly to a waveguide through an iris (made from either glass or ceramic).

[tron9000]

but I'm pretty sure its aperture coupled (fig 15), so definitely needs re-vacuuming!

Read your own link: Aperture, or slot, coupling is illustrated in Figure 14. Energy is coupled directly to a waveguide through an iris (made from either glass or ceramic).

Ahh I read it wrong I thought it was on about the other end, it's the end going into the body it means. In which case then it should be ok, providing it hasn't been damaged during storage.

[T3sl4co1l]

FYI, at rated voltage, the heater current will be above spec, if the tube's at atmospheric pressure.

Horns: http://www.microwaves101.com/downloads/Slayton/William%20Slayton%20Electronics%201955.pdf

Tim

[tron9000]

cheers Tim much appreciated!

[BennVenn]

If I get a chance, I'll go for a walk over to one of the ships and see if I can get some quick dimensions of the feed horns. Most of the gear running at that frequency is contained within pressurised domes but lower frequency stuff is accessible. There might be some correlation between waveguide dimensions and horn dimensions. It's likely all formula derived but given the speed of light is constant, you never know!

[tron9000]

If I get a chance, I'll go for a walk over to one of the ships and see if I can get some quick dimensions of the feed horns. Most of the gear running at that frequency is contained within pressurised domes but lower frequency stuff is accessible. There might be some correlation between waveguide dimensions and horn dimensions. It's likely all formula derived but given the speed of light is constant, you never know!

don't go too far out of your way, I mean if you have to climb up something then its really no bother I can work it out!