Author Topic: Magnetron insulators  (Read 2555 times)

0 Members and 1 Guest are viewing this topic.

Offline antennaTopic starter

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: us
Magnetron insulators
« on: December 14, 2021, 03:53:00 am »
I know this has been gone over a bunch of times but every thread warns me that it is an old thread, so starting new.

I remember hearing somewhere that its the red ceramic that is dangerous, then I heard they don't use beryllium anymore. Then, somewhere else, I read the white ceramic was dangerous too (and wikipedia shows it to be white), which confused me always hearing the red kind was bad.  The magnetron I took apart has both red and white ceramic. I would like to cut it open and see the cavities inside and how it failed so I am hoping someone can tell me which end is safest to use tools on.  Are both ends made of beryllium oxide?  Perhaps it is not worth looking inside...

Thanks!
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Magnetron insulators
« Reply #1 on: December 14, 2021, 04:35:13 am »
AFAIK red is always chrome-tinted alumina, and white can be either (alumina or beryllia).

Only beryllia is toxic, and only when fractured or ground -- it's the dust.

Find a datasheet for it?

Anyway, you're cutting into metal right?  Just don't break the ceramics, you're fine.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: antenna

Online richard.cs

  • Super Contributor
  • ***
  • Posts: 1201
  • Country: gb
  • Electronics engineer from Southampton, UK.
    • Random stuff I've built (mostly non-electronic and fairly dated).
Re: Magnetron insulators
« Reply #2 on: December 14, 2021, 09:41:42 am »
I did read somewhere that some beryllia is/was tinted red as an indicator/warning, but it was probably just on a forum somewhere. Certainly all the beryllia I have is white. In big pieces like a magnetron you can probably tell by thermal conductivity, it's a lot more conductive than alumina and should feel cool to the touch like a piece of metal.

If you have any doubts, cut the metal not the ceramic as Tim suggests.
 
The following users thanked this post: antenna

Offline antennaTopic starter

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: us
Re: Magnetron insulators
« Reply #3 on: December 14, 2021, 05:30:53 pm »
Thank you for the replies. I wonder how far into the metal the ceramics go, but regardless, I will cut into the larger cavity portion near one end and avoid the narrow sections that may contain the ceramic.  When the microwave failed, I tested the filament for continuity and for continuity between that and ground. There was still continuity between the filament connections, but also to ground. I expected that a failure would involve the filament breaking as I assumed it would be too heavy to simply stretch and short to ground, but that's the story my meter tells and the reason for my curiosity.  I will see if the thermal conductivity can provide a clue as to the materials, thanks again!

« Last Edit: December 14, 2021, 05:33:15 pm by antenna »
 

Online Gyro

  • Super Contributor
  • ***
  • Posts: 10173
  • Country: gb
Re: Magnetron insulators
« Reply #4 on: December 14, 2021, 09:01:35 pm »
Quote
I wonder how far into the metal the ceramics go, but regardless, I will cut into the larger cavity portion near one end and avoid the narrow sections that may contain the ceramic.

As a guide, here's my failed Panasonic magnetron (cracked magnet and anode structure melting)... https://www.eevblog.com/forum/chat/microwave-oven-pwm-frequency/msg3137762/#msg3137762 (See the image attachment)

First you need to pull off the sheet metal anode cap - it pulls off with pliers. Then saw through the copper pinch seal that you will find underneath. This breaks the vacuum (tiny hiss) and frees up the anode connection rod which is crimped in with it.

The easiest place to cut is between the top of the copper anode and the top steel plate - there is normally a groove. Rotate it while cutting to avoid damaging the internal structure and avoid excessive clamping pressure - the copper cylinder may look very solid but it is soft and easily distorts if squeezed too hard.

There is no Beryllium in either ceramic tube (as has been discussed at length in previous threads), but there is no need to cut or clamp them anyway and shattering ceramic won't do your eyes any good regardless!

Any device containing Beryllium will have a prominent warning label.
« Last Edit: December 14, 2021, 09:04:57 pm by Gyro »
Best Regards, Chris
 
The following users thanked this post: antenna

Offline TerraHertz

  • Super Contributor
  • ***
  • Posts: 3958
  • Country: au
  • Why shouldn't we question everything?
    • It's not really a Blog
Re: Magnetron insulators
« Reply #5 on: December 17, 2021, 12:57:07 pm »
Every single time oven magnetrons are discussed, there's always someone warning about potential BeO and the dangers.
Another recent eg: [HP-Agilent-Keysight-equipment] OT: microwave oven blows fuse

I flatly refuse to believe BeO would ever be found in consumer microwave magnetrons. Even made in a China.
Reasons:
1. There's no need. BeO is used for its high thermal conductivity plus electrical insulation. Both the output stub and filament base insulators in magnetrons don't need high thermal conductivity. Alumina will do fine.
2. Never once saw any BeO warning markings in microwave ovens. Any manufacturer that included BeO without warning markings (including the 'do not dispose of' one!) would get legally slaughtered.
3. BeO manufacturing processes tend to kill workers unless extreme dust containment measures are taken. Too much trouble for ultra-cheap consumer items.

They just use alumina. Apparently of random colors.

Anyway, boring. I'm sure I'm going to read "Be careful! Warning! BeO is deadly" in every single magnetron thread forever.
Meanwhile, I have a salvaged high power X-ray tube from an XRF machine, that has a pure Beryllium metal output window. You should see the warnings related to that. Fundamentally "DO NOT EVER DO ANYTHING" to it. No touch, no cleaning, no 'disposal.'

Back to magnetrons. Here's a pic of one I opened by cutting away the whole output stub side, using a lathe for neatness.
This magnetron has the common 'electron beam welding' effect. Some asymmetry (perhaps of the magnets) results in uneven heating of the copper vanes. Once their shape changes even a little the magnetron will just keep melting itself until it totally stops working.

Btw, if opening a magnetron using a lathe, be careful. The cylindrical body is a thin cylinder of well-heat-annealed copper. Very soft. These pics were after successfully cutting away one end plate. Then I also tried to cut off the opposite end plate. No pics of that. Because it went horribly wrong. As soon as the cutting bit began to separate the end plate, it turned out that was all that was supporting the copper cylinder in the lathe jaws. Without support it instantly deformed, dug in the cutting bit, and the remains of the magnetron ejected from the lathe at speed.  I don't seem to have pics of the result.

I do have a grinder attachment for the lathe. Cutting off the magnetron end plates with slow chuck RPM and high speed abrasive cutting disk would work. And I have another dead magnetron atm. But grinding on the lathe is a pain - grit is bad for it. Needs a lot of dust covers and cleanup after.

Collecting old scopes, logic analyzers, and unfinished projects. http://everist.org
 

Online richard.cs

  • Super Contributor
  • ***
  • Posts: 1201
  • Country: gb
  • Electronics engineer from Southampton, UK.
    • Random stuff I've built (mostly non-electronic and fairly dated).
Re: Magnetron insulators
« Reply #6 on: December 17, 2021, 01:28:20 pm »
I flatly refuse to believe BeO would ever be found in consumer microwave magnetrons.
I am inclined to agree, for the reasons you state, but:
  • Some oddball 1970s magnetron with BeO may exist. Not necessarily from a consumer microwave oven.
  • It does no harm for beginners to become aware that BeO ceramics exist, even if they are (probably) not in magnetrons.
  • Alumina dust isn't entirely harmless either, generally discouraging people from grinding ceramics without care is good.

It is pretty hard to prove a negative, and I am unwilling to definitively state that no magnetron that an amateur might encounter has ever been manufactured with BeO. But it is probably true unless they're getting them out of military radars or something.
 

Online Gyro

  • Super Contributor
  • ***
  • Posts: 10173
  • Country: gb
Re: Magnetron insulators
« Reply #7 on: December 17, 2021, 01:31:46 pm »
Back to magnetrons. Here's a pic of one I opened by cutting away the whole output stub side, using a lathe for neatness.
This magnetron has the common 'electron beam welding' effect. Some asymmetry (perhaps of the magnets) results in uneven heating of the copper vanes. Once their shape changes even a little the magnetron will just keep melting itself until it totally stops working.

Your failure mode is the same as my Panasonic one (melted down anode segments) but yours is much worse, and very asymmetrical. Hard to believe the anode went out of shape with all those cooling fins around it. Total magnet failure on one side?

This is one occasion where I definitely prefer the hacksaw method to the lathe. Much thinner blade and a lot less swarf.
« Last Edit: December 17, 2021, 01:43:35 pm by Gyro »
Best Regards, Chris
 

Offline antennaTopic starter

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: us
Re: Magnetron insulators
« Reply #8 on: December 31, 2021, 03:39:23 am »
off topic, but I thought magnetrons were supposed to be relatively pure in frequency, I guess not. Just for kicks I turned on the spectrum analyzer in the bedroom and went to the kitchen and turned on the microwave.  I never could have imagined them being so broadbanded. I don't understand how my wifi router can even function when that thing is on.
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Magnetron insulators
« Reply #9 on: December 31, 2021, 03:53:17 am »
They're FM to applied current or something like that, so they're spread spectrum at line frequency.  Delightful, eh? :P

IIRC, phase noise at DC is alright, like most any LC oscillator you'd expect; drifty, not like you'd get from a crystal controlled synth of course, so, not really suitable given typical channel size/BW up there, but just doable.  And for WBFM voice or something like that, yeah, works out.  Oh, uh I suppose you'd normally have a PLL servoing DC, then modulate on top of that to get reasonable performance.  But the PLL only works so fast, it can clean up close-in phase noise but misses farther-out noise, or added modulation obviously.  And drift can be helped by keeping temp stable, or regulating the magnets with an external servo coil or something.  Since those have a... negative tempco usually, isn't it?

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline antennaTopic starter

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: us
Re: Magnetron insulators
« Reply #10 on: December 31, 2021, 05:43:36 am »
They're FM to applied current or something like that, so they're spread spectrum at line frequency.  Delightful, eh? :P

IIRC, phase noise at DC is alright, like most any LC oscillator you'd expect; drifty, not like you'd get from a crystal controlled synth of course, so, not really suitable given typical channel size/BW up there, but just doable.  And for WBFM voice or something like that, yeah, works out.  Oh, uh I suppose you'd normally have a PLL servoing DC, then modulate on top of that to get reasonable performance.  But the PLL only works so fast, it can clean up close-in phase noise but misses farther-out noise, or added modulation obviously.  And drift can be helped by keeping temp stable, or regulating the magnets with an external servo coil or something.  Since those have a... negative tempco usually, isn't it?

Tim
I'm lucky to make an audio amplifier work decently so I'll pass on making an FM transmitter out of my pizza reheater, I was just shocked at the level of power getting outside the oven and the bandwidth it covered.  What you say sounds logical though, but honestly, it's way above me.  I don't think I'll be trying to stabilize my microwave oven, its bandwidth is probably a benefit when it comes to cooking.  But with that first sentence you wrote, am I correct to assume that if I remove the 60Hz ripple on the HV side with better filtering, it wouldn't do that?  I thought the frequency was controlled by the geometry of those little cavities the electrons zip past? I guess I need to hit up wikipedia and get a crash course on magnetrons, klystrons etc.
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Magnetron insulators
« Reply #11 on: December 31, 2021, 07:43:40 am »
Heheh, yeah.  Certainly wouldn't recommend you do so; but, have seen an amateur build such a transmitter before -- more just to say it's possible, and how well (or badly) it works.

As an oscillator, it depends on cavity dimensions yeah, but also whatever's coupled to the load (so, what's going on in the chamber), and various other conditions.  It has some natural bandwidth, of which it mostly oscillates in the middle, but can be perturbed to one side or the other -- this is what gives the tuning range.  It's only a 100MHz or so, so the fractional bandwidth is pretty low -- as you'd expect -- but compared to the band it operates in, still a pretty wide range.

Just like, you can pull a crystal, but only by a few kHz; you just can't couple to it strongly enough to push it around further.  And indeed, with good reason, else it wouldn't be a good frequency reference, hehe.

I forget why exactly current causes FM.  It causes AM too of course; the V(I) curve is not like a tube diode, but a zener diode -- the magnetic field causes a threshold effect, below which the electrons spiral uselessly back to the cathode.  So of course, up at that threshold voltage, more current means more electrons blowing the cavities like a whistle, more power output; but it turns out it also modulates frequency.  (FM receivers are designed to reject AM, so the AM doesn't matter in practice.)

It would be a bit of doing to put a filter on there, but that is indeed the correct reading, at least as far as I know.  You'd need a few more capacitors (at least one more, but probably of a much larger value in order to get useful filtering effect), and another diode, and maybe an active circuit (to regulate the current, saving on filter size).  Not recommended, of course; the voltages are nontrivial. :)

Tim
« Last Edit: December 31, 2021, 07:53:32 am by T3sl4co1l »
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: antenna

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16385
  • Country: za
Re: Magnetron insulators
« Reply #12 on: December 31, 2021, 10:07:00 am »
Yes, allowable leakage for a microwave oven is pretty high, much higher than for intentional radiators like WiFI AP's, and the associated units. 0.5W is the limit in most cases for the microwave, and more for the larger commercial units, while your AP can only dream of ever getting to 0.01W right at the antenna port.

Guess which one people are concerned about.......
 

Online tom66

  • Super Contributor
  • ***
  • Posts: 7334
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Magnetron insulators
« Reply #13 on: December 31, 2021, 05:15:28 pm »
Most Wi-Fi routers and user devices implement "microwave oven avoidance".  Either they use the region code or the RF measurements to determine when the microwave is not actively transmitting and transmit in that period.  Most microwave ovens use simple capacitive voltage doublers which leads to a quiet period at 50 or 60Hz during which time transmission can be scheduled.

I do wonder how this applies to ovens that use inverters, as they use a power-factor correction circuit that should ensure current through the magnetron is a sine wave at 100Hz, which should mean overall utilisation of the 2.4GHz band is closer to something like 70-80% rather than the 40-50% of a traditional microwave.
 
The following users thanked this post: antenna

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8526
Re: Magnetron insulators
« Reply #14 on: January 01, 2022, 12:01:54 am »
I have troubleshooted and worked around problems with dropped connections while the microwave oven is on by changing the WiFi channel to one or the other end of the range.
 

Offline BrokenYugo

  • Super Contributor
  • ***
  • Posts: 1214
  • Country: us
Re: Magnetron insulators
« Reply #15 on: January 04, 2022, 07:00:53 pm »
Most Wi-Fi routers and user devices implement "microwave oven avoidance".  Either they use the region code or the RF measurements to determine when the microwave is not actively transmitting and transmit in that period.  Most microwave ovens use simple capacitive voltage doublers which leads to a quiet period at 50 or 60Hz during which time transmission can be scheduled.

I do wonder how this applies to ovens that use inverters, as they use a power-factor correction circuit that should ensure current through the magnetron is a sine wave at 100Hz, which should mean overall utilisation of the 2.4GHz band is closer to something like 70-80% rather than the 40-50% of a traditional microwave.

The only inverter microwave I'm familiar with is an older Panasonic that kills all wifi communication in the room. The home theater receiver has a wireless stereo speaker unit that uses B/G wifi channels for the link, you can hear the magnetron cathode heating up before the audio cuts out entirely and it's about 4 meters from the microwave. It has always done that so I assume it is safe/normal operation.
« Last Edit: January 04, 2022, 08:17:26 pm by BrokenYugo »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf