Hilarious design job - anyone want it...!

[botcrusher]

To add to the list of their recent f*k ups, add the Canadian Government's new payroll system...

[PartialDischarge]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!

Some switches from Behlke comes close to this.

How do you even measure that frequency/voltage??? al probes have a huge derating with frequency

[nctnico]

I regularly get random emails with requests to design stuff - this one is particularly hilarious...

Subject:  High Voltage switch enquiry
Hi,
We are based in Malvern UK and are looking for an engineer to design and build a high voltage high frequency switch.
This project has a commercial application and we have a budget of £2000.
Ideally the general parts for the project should be available off-the-shelf (RS, FEC, Rapid).
General Spec:
Our aim is to provide two polarised plates (in air) with a DC (8KVDC) pulsed frequency (variable by tuning) between 40-60Mhz.
The plates will not be shorted and no arc is to be produced, but in the probable event of an arc the circuit needs to gracefully close down and restart.
There will not be any load across the plates, these are just used for polarization purposes.
The frequency of between 40 and 60 Mhz needs to be controlled by tuning is some way, this could be digital or analogue or a combination.
Our ultimate goal is to productionise this module, so any design considerations will be biased towards this, keep it simple.
If you can help or you know of someone that can please will you let me know.

Sounds like a great project to me IF they can add a zero to the budget (at the left side of the dot ofcourse).

[HighVoltage]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!

Some switches from Behlke comes close to this.

How do you even measure that frequency/voltage??? al probes have a huge derating with frequency

You probably have to build your high voltage probe, to get this high frequency measured stable.
Not impossible.

[daqq]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!

Some switches from Behlke comes close to this.

How do you even measure that frequency/voltage??? al probes have a huge derating with frequency

Very, very carefully High voltage high frequency probes do exist and are quite expensive. Or you can probably guess it from the effects of it on the aura field of the quantum chackra?

[tggzzz]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!

Some switches from Behlke comes close to this.

How do you even measure that frequency/voltage??? al probes have a huge derating with frequency

You probably have to build your high voltage probe, to get this high frequency measured stable.
Not impossible.

... but surprisingly difficult in practice.

The Tek 6015 (a bargain at £1600) is 20kV/75MHz, but - as is usual with probes - not at the same time. See figure 1.4 in http://www.tek.com/high-voltage-probe-manual/p6015a

You could have 8kV 6MHz, or 60MHz 1.5kV-ish; I doubt their customers would notice any difference

[setq]

I'd take the £2k for this job but only 100% up front.

I'd send them a box with some potted sand in it and a mains lead attached.

If it didn't work I'd tell them they didn't believe in it so it didn't work.

[John Coloccia]

Or just repackage one of these:

http://www.homedepot.com/p/Bite-Shield-Electronic-Racket-Zappers-with-Protective-Grids-RZ02C/100616483

Every now and then they'll hear a "zap". You can tell them that it's one of those arcs they warned you about, but your design is able to handle it.

[setq]

Genius!

[PartialDischarge]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!

Some switches from Behlke comes close to this.

How do you even measure that frequency/voltage??? al probes have a huge derating with frequency

You probably have to build your high voltage probe, to get this high frequency measured stable.
Not impossible.

Ok, how. What is the technique. I'm guessing it must be active.

As mentioned the tek hv probe gets derated at 1.5kV for 60MHz

[HighVoltage]

Ok, how. What is the technique. I'm guessing it must be active.

As mentioned the tek hv probe gets derated at 1.5kV for 60MHz

I would build a low capacity 1000:1 capacitive divider and probe it with a good high speed probe.
The capacitance and resistance of the probe has to be included in to the system calculation.
If you use a 200 MHZ differential probe, you can measure the voltages perfectly.
This way you will get very good and repeatable results, since it is not a universal probe but specifically made for this application.
I have done this for up to 60 kV or so.

The Tek probe is great for slow signals but it loads the DUT a lot at very high voltages.
Plus there are different P6015 probes around with different capacities.

[Marco]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!   

Won't necessarily induce a ton of power.

There are probably medical devices operating in this range, they use all kinds of weird shit in dermatology.

[bitshape]

I bet they have lot's of snake oil for sale. 

[T3sl4co1l]

Well, meh, not entirely sure I'd no-quote it -- but, I'd want to be damn sure of the application and what kind of medium is between those plates (you know, for electrical purposes -- I don't care if they're cooking a chicken with it, or gluing plywood, or gluing ply-woo-woo or what), and what shielding they have around it (in particular, a full coverage Faraday shield to keep EMI inside and fingers outside; potentially limiting their intentions, but hey, there are exposure limits, on the lawbooks no less!), and I would quote no less than* five times their "expected" budget, with a lengthy product development and law and patent research phase just in case.

*Meaning I'd start with at least double the figure, and negotiate from there.

In any case, I'm guessing the outcome is the same... but Mike's approach to responding to them is definitely more efficient

Tim

[HwAoRrDk]

Been to their sister site? http://computerclear.com/

"But help is at hand in the quest for EMF Protection! The Computer Clear software for Microsoft Windows uses homeopathic techniques to strengthen your body's natural resistance to the EMF radiation that comes off your computer. It is no exaggeration on our part to say that Computer Clear has transformed people's experience with their computer."

Holy-moly, that's hilarious!

I like how they also try and link EMF radiation with a legitimate medical complaint, Computer Vision Syndrome (CVS). CVS is nothing to do with with EMF, it's just caused by the eye's focusing muscles spending too long stuck in one position (i.e. staring at a screen) and excessive drying of the eyes due to lower blinking frequency when staring at a screen. The solution? Stop staring at a screen for so long! Get up and do something else!

I'd love to get a hold of a copy of that software just to try and examine it's inner workings and see if it does anything at all, or whether it's just a bunch of placebo buttons and knobs.

[SL4P]

I wonder what they've budgeted for EMC testing? Could be a nice little never-ending consultancy job right there.

Half of their dev budget... GBP 1000

[joeqsmith]

8kV @ 60MHZ !!  Might as well just go stick your head in a microwave!!

Some switches from Behlke comes close to this.

How do you even measure that frequency/voltage??? al probes have a huge derating with frequency

Very, very carefully High voltage high frequency probes do exist and are quite expensive. Or you can probably guess it from the effects of it on the aura field of the quantum chackra?

Mr Pickles!!  LOL!!   

That would be a fun probe to try and build as a home project, not! 

[SL4P]

Have the Roohparvar's moved to the UK?
Or - perhaps if Bob is following - he should apply for this opportunity.
I'm sure their experience and qualifications are eminently suite to this con.

[JimRemington]

There will not be any load across the plates, these are just used for polarization purposes.

The strange thing about this specification is that they have a high frequency and high voltage in mind, but do not understand that "polarization" is a load.

[Brumby]

There will not be any load across the plates, these are just used for polarization purposes.

The strange thing about this specification is that they have a high frequency and high voltage in mind, but do not understand that "polarization" is a load.

The strange thing about this specification is that they have a high frequency and high voltage in mind, but do not understand <insert phrase here>.

You can start with these:
* the modial interaction of magneto reluctance and capacitive duractance
* inverse reactive current as used in unilateral phase detractors
* metapolar refractive pilfrometer
* the difficulty of nubbing together a regurgitive perlwell and a supernative weinal sprocket

[Cyberdragon]

Voltamort would be happy to consider your offer for 2000 pounds...

Time to get some large scale glasswork equipment...because you know where this is going...a giant ass water-cooled transmitter tube (valve). Just make sure I sign off all liability...and warn of x-rays (I'm sure they like X-rays)

Then we wait...

"Experiment Turns Deadly
Several people were found dead yesterday at a new age healing facility. The police were called after a woman heard screaming coming from the building. Apon arrival, several members of the society were already dead. There bodies were found electrocuted and covered in RF burns between two large metal plates in the center of the room. These plates were connected to a large glass device covered in wires and pipes marked "Voltamort Labs 8K60MC" which was still very hot and glowing. A quote from one off the officers on the scene was "The air was so full of electricity that I could feel my hair standing up". The machine was promptly switched of to prevent any further harm. Investigators later discovered that the facility signed a realese of liability with manufacurer of the device, Voltamort Labs. A survivor of the accident claims they were trying to "

But seriously, what is the current potential of this voltage they are using? Because in all likelyhood these idiots could actually kill themselves.

[T3sl4co1l]

But seriously, what is the current potential of this voltage they are using? Because in all likelyhood these idiots could actually kill themselves.

Well.............

If we suppose wave-of-the-hands plate sizes of 1m x 1m and 1m spacing, that's 8.84pF.

Except it's more because of fringing, and it's more again because of self-capacitance (basically, field that's fringing out the back side, over to ground, and back up again to the other side).  Probably more like 20 or 30pF?

At 60MHz, 30pF is 88 ohms reactance.  8kV into 88 ohms is 90A and 723kVAR.

At a respectable Q of 500 (you'll need a pretty healthy spool of litz cable* to achieve this), you're still dissipating a toasty 1.45kW, which will need quite a bit of electrical power (which is actually available, from a UK wall outlet), and will quickly warm up the office (which is a nice feature, this time of the year).

The coil will need to be tunable by tap setting, from 0.23 to 0.53uH.  To handle 80A at up to 60MHz:

*Regular litz cable craps out circa 10MHz, due to capacitance between strands.  To keep going at higher frequencies, you need to increase the separation between strands.  Likely, you'd have to build your own super-cable, using a cage structure.  It might look something like this:
https://upload.wikimedia.org/wikipedia/commons/6/61/Solec_Kujawski_longwave_antenna_feeder.jpg
You might use many-pointed-star shaped insulating plates to support the strands in a cylindrical shell arrangement.

This structure is then twisted into a modest helix, having the effect of circulating the strands between the inside and outside of the coil.

The cable, thus described, gets wound onto a cylindrical form, to make a solenoid coil.  Not very many turns would be necessary; maybe a single loop would be sufficient.

Otherwise, solid tubing should be used, which will need to be quite large in diameter to achieve such a high Q.  It will still get hot.

Needless to say, the RF electric field around anything nearby will also be quite intense, leading to partial discharge, dielectric heating and corona discharge (even if not resulting in complete hot-arc type breakdown).

Of course, any heating and average-power-consumption effects will be greatly diminished by duty cycle, if this is a pulsed application.

As for the "pulsed DC" part of it, who knows.  Do they mean to say they don't want pulsed wavelets of RF, but very short, unipolar pulses?  Should they be alternating or consistent polarity (thus giving an average -- DC -- bias)?  How much undershoot and rebound is tolerable, if any?

Such signals can still be generated, but it requires very fast switches, such as a hydrogen thyratron, or a pressurized gas or liquid filled spark gap.  The repeat rate certainly will not be up to 60MHz.

Tim

[EEVblog]

The strange thing about this specification is that they have a high frequency and high voltage in mind, but do not understand <insert phrase here>.
You can start with these:
* the modial interaction of magneto reluctance and capacitive duractance
* inverse reactive current as used in unilateral phase detractors
* metapolar refractive pilfrometer
* the difficulty of nubbing together a regurgitive perlwell and a supernative weinal sprocket

 

[MK14]

This is my proposal for them. It would take about 6 weeks to complete, but would go over budget by about 50..60%. See attachment(s) for details.

[Cyberdragon]

It's not that hard to make that kind of power, all they need is a klystron tube.