Baggage x-ray machine teardown

[BravoV]

Thanks Mike, great series and very informative as always. 

This reminds me of an interesting app note from Intersil that I read while ago here -> X-Ray Effects on Intersil FGA References

If you had a chance and not bored yet  , please do more experiments on floated silicon like buried zener voltage reference, eeprom, flash  memory to see how nasty it will affect them.

Apart from the disadvantages from x-ray exposure, I'm thinking the possibilities or advantages if this method could be used to sort of micro trimming/adjusting electronic stuff, maybe at ppm level say like an extra treatment at a well cooked buried zener voltage reference to get a really high accuracy at it's absolute value.

[Psi]

do more experiments on floated silicon like buried zener voltage reference, eeprom, flash  memory

We did some test with flash memory (CF and SD) sitting stationary in the airport xray beam for a few minutes.
The section of the flash memory that was exposed got the equivalent dose of around 8000 runs through the machine.
We couldn't fault it afterwards. The data was byte for byte accurate and we could write/read new data fine.

[mikeselectricstuff]

Part 5 :

[mikeselectricstuff]

do more experiments on floated silicon like buried zener voltage reference, eeprom, flash  memory

We did some test with flash memory (CF and SD) sitting stationary in the airport xray beam for a few minutes.
The section of the flash memory that was exposed got the equivalent dose of around 8000 runs through the machine.
We couldn't fault it afterwards. The data was byte for byte accurate and we could write/read new data fine.

I was going to do a vid testing a variety of memory devices - I got as far as exposing a couple of eproms for a minute (no effect) then my temporary keypad repair started going flaky (water based conductive ink in damp garage) & I couldn't be bothered fixing it to continue.

[HardBoot]

You need pretty high energy particles to nuke data, although x-ray does damage undeveloped film, since it is high energy light in a way.

[peter.mitchell]

Reminds me of air travel, having my rolls of film in a big ziplock bag for carry on that I had to specify not to be x-rayed

[EEVblog]

Awesome. 
Wasn't worth anything on ebay?

Dave.

[mikeselectricstuff]

Awesome. 
Wasn't worth anything on ebay?
Dave.

It had failed to sell a couple of times in the past at £100. Needs some serious transportation - we only got it into a 3.5 tonne box van by taking one of the conveyor arms off so it would get onto the tail-lift and then swing round into the van.
The seller was also struggling to sell a fully working one at £800.
I suspect the end of the Olympics might have helped flood the used x-ray. market.
I'll probably try ebaying a few oddments like the conveyor motor and detector boards.

[EEVblog]

I suspect the end of the Olympics might have helped flood the used x-ray. market.

That makes sense.
We had a big auction here in Oz in 2000 after our Olympics. Huge ticket items like hundreds of cars used to transport everyone, and even the big opening ceremony parade floats!
Not sure if there was any tech gear in all that, probably was and I missed it.

Dave.

[SeanB]

The buses from the Beijing Olympics ended up here in Durban as public transport. Saw one with the sign at the original default setting of Chinese with the english underneath saying Olympic Stadium.

[firewalker]

You can always sell the lead. ~1,5 €/kilo ?

Alexander.

[SeanB]

More money in the aluminium than the lead, Steel is almost the same price as lead. Aluminium is about 4 times the price.

Mike should send one of those to Dave as a wall decoration......

[EEVblog]

Mike should send one of those to Dave as a wall decoration......

There is something in my massively overflowing mailbag from Mike. Pretty sure I know what it is.
Wait till you see his return address... 

Dave.

[peter.mitchell]

Mike should send one of those to Dave as a wall decoration......

There is something in my massively overflowing mailbag from Mike. Pretty sure I know what it is.
Wait till you see his return address... 

Dave.

If it's legit, i might send mailbag stuff to him too!

[mikeselectricstuff]

More money in the aluminium than the lead, Steel is almost the same price as lead. Aluminium is about 4 times the price.

There's actually hardly any aluminium. pretty much all lead & steel.

[mikeselectricstuff]

Mike should send one of those to Dave as a wall decoration......

There is something in my massively overflowing mailbag from Mike. Pretty sure I know what it is.
Wait till you see his return address... 

Dave.

That was quick....

[mikeselectricstuff]

Part 6 - inside the black box

[tom66]

You say ferrite core... does this mean it's like an SMPS? Or is it simply running at mains frequency?

[SeanB]

Probably running at 15kHz or so, those look like old LOPT cores, and they work well at that frequency. Might be higher, up to 30kHz, but those multipliers use high voltage diodes and they generally are not going to be terribly fast switches, they generally have a max frequency limit of around 25kHz. The capacitors inside are likely to be hundreds of picofarads at around 20kV each, thus the big potted assembly, to keep the oil from damaging them.

there must be a driver with a push pull output in the controller driving them, likely running off a power rail of around 150VDC or so, at a current of around 5A peak.

[mikeselectricstuff]

Probably running at 15kHz or so, those look like old LOPT cores, and they work well at that frequency. Might be higher, up to 30kHz, but those multipliers use high voltage diodes and they generally are not going to be terribly fast switches, they generally have a max frequency limit of around 25kHz. The capacitors inside are likely to be hundreds of picofarads at around 20kV each, thus the big potted assembly, to keep the oil from damaging them.

I forgot to wave a scope probe near the driver when running to check the frequency.  I did however work out all the pins on the control connector, just in case..!
15K is a reasonable compromise between not being audible and high-frequency losses - loading due to stray capacitance can become a major problem in high frequency HV supplies.

there must be a driver with a push pull output in the controller driving them, likely running off a power rail of around 150VDC or so, at a current of around 5A peak.

The control box runs off 110VAC, so 150VDC is probably about right for the drive.
Details of this are at around 36:00 in part 1.

[SeanB]

150VDC is pretty common for old TV sets as line drive, What devices are they using in the driver, probably BU208/508D.

[mikeselectricstuff]

150VDC is pretty common for old TV sets as line drive, What devices are they using in the driver, probably BU208/508D.

it's not that old - probably MOSFETs.

[SeanB]

Bipolars are more common there, especially if you want to have the output devices survive flashovers that often crisp the mosfets. 50 years of designing TV line devices is still valid, and there is no real power saving of mosfet and bipolar in this application, drive power is still pretty much the same, just one has high current low voltage and the other higher voltage lower current. You can use IGBT's but they would be expensive compared to a cheap high SOA bipolar.

[G7PSK]

If that is too hard for lead it is either a lead zinc alloy or depleted uranium or alloy of.

[mikeselectricstuff]

Bipolars are more common there, especially if you want to have the output devices survive flashovers that often crisp the mosfets. 50 years of designing TV line devices is still valid, and there is no real power saving of mosfet and bipolar in this application, drive power is still pretty much the same, just one has high current low voltage and the other higher voltage lower current. You can use IGBT's but they would be expensive compared to a cheap high SOA bipolar.

Just looked and it uses two IRF630 MOSFETs.
You'd  want to be protecting things inside the oil can (as they have done), not relying on the drivers to maybe survive any nasties that might emerge.