EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Sander on September 04, 2014, 12:12:58 pm
-
I’m looking for a part that can switch 4000 volts into a 50 ohm load within a nanosecond. The pulse duty cycle should not exceed 1 microsecond (80 amps for at maximum 1us). I was thinking of a mercury relay or a hydrogen thyratron, but the ones I have seen seem to slow or lack the specifications. Has someone used, or know of a part that is capable of this? Maybe a high power VHF/UHF transmit tube?
-
It is fast for a thyratron, they work by avalanche ionisation so I can't see how you'd get the switching time. A UHF transmit tube would be better in this respect, you should be able to slam it on fast enough :-)
You don't say what your repetition rate is but if you can pull the A-K voltage down to a hundred volts or so you're at 8 kJ anode dissipation per pulse, that doesn't sound too unreasonable. The 80 Amps is going to be interesting, I think you'll struggle to find one that specifies a peak cathode current that high *but* for such a short pulse maybe it'll be ok, meerly depleting the cloud of electrons surrounding the cathode without actually requiring all that much cathode emission. Is there 80 microcoulmbs of charge floating around the cathode? I have no idea.
No idea about the mercury relay, It seems like it should be pretty fast.
-
The repetition rate is 100Hz @ 2kV and 60Hz @ 4kV.
-
Does the trailing edge have to be that fast or just the leading edge?
The Russians make plenty of fast rise time pulse generators, I think they generally use SCR or IGBT followed by a magnetic switch to get to ns rise time and then throw it at a SOS to get it sub-ns.
PS. probably not the best time to order anything from Russia though.
-
Yeah, if the trailing edge can be slower turn off then as you suggest, a mercury wetted relay but discharging a cap into your 50 ohms. Plenty of available current and no problem snubbing the relay contacts because the current will be zero when the contact open.
-
You can also use planar triodes (including many commercial UHF parts) in transmission line structures:
http://www.slac.stanford.edu/cgi-wrap/getdoc/slac-pub-3546.pdf (http://www.slac.stanford.edu/cgi-wrap/getdoc/slac-pub-3546.pdf)
Or avalanche transistors; http://icecube.wisc.edu/~kitamura/NK/Flasher_Board/Useful/research/RSI03066.pdf (http://icecube.wisc.edu/~kitamura/NK/Flasher_Board/Useful/research/RSI03066.pdf)
-
Would a trigatron work at that low a voltage?
Eh but I don't know what the duty cycle would be...
-
Pretty sure you are going to need to use a "pulse forming network", and probably with hardline coax. Jim Williams (of Linear Technology fame) wrote some good app notes on this a few years ago.
[edit: mistakenly attributed the late Jim Williams to Analog Devices... doh.]
-
A trigatron built into a transmission line is what occurred to me as well.
-
http://www.siliconinvestigations.com/KRYT/Krytron.HTM (http://www.siliconinvestigations.com/KRYT/Krytron.HTM)
there we go. Another glass solution, but NLA...
http://www.tubecollector.org/kn22.htm (http://www.tubecollector.org/kn22.htm)
-
The repetition rate is 100Hz @ 2kV and 60Hz @ 4kV.
Sounds like HVDC transmission line stuff?
http://en.wikipedia.org/wiki/HVDC_Inter-Island (http://en.wikipedia.org/wiki/HVDC_Inter-Island)
Check out the thyristor switching hall halfway down the page. :-+
-
HVDC converters don't have to switch below a ns.
Mercury wetted relays/contactors are probably the cheapest shot at getting below a ns, it's a bit of a gamble though because you don't know the internal inductance of these things ... it's not what they are designed for. Also very high jitter and less long term reliability than fully solid state solutions. If he wants a square pulse using a piece of HV coax as a storage capacitor would indeed be the way to go.
-
That’s exactly it. I want to dump the energy stored in a piece of coax into a 50 ohm load and AC couple that pulse to a mains connector of a device under test, like the INS series from NoiseKen:
http://www.noiseken.com/uploads/photos0/131.pdf (http://www.noiseken.com/uploads/photos0/131.pdf)
They use a mercury relay, but the ones I can find are specified to only 3 amps. I could order one and hope it won’t explode…
Krytrons are probably exactly what I want, but I don’t have a DeLorean DMC-12 and I don’t know any Libyans so they might be hard to get.
The avalanche transistor solution would be excellent, easy to get cheap parts but it can’t handle a pulse width of more than 15ns. I need 50ns – 1us pulse width.
The solution with the planar triodes might work, but it will exceed the specified current by a order of magnitude. Even if I replace the final stage with a GI-39B I still exceed the specs 5 times.
-
The only other thing I can think of is an IGCT but they won't get you the PRF you want.
But you'll be able to power the commuter train for a small town.
So there's that.
What about a triggered spark gap? A DIY Kryton, get a spark gap tube, and wrap some fine wire around it and use a photoflash trigger circuit ?
-
They use a mercury relay, but the ones I can find are specified to only 3 amps. I could order one and hope it won’t explode…
I doubt a 1 us pulse with no need to extinguish an arc will do anything to it, I think it dumps orders of magnitudes more power into the mercury when it's breaking a contact of 3.5kV 60 Hz.
These guys (http://www.wolfautomation.com/products/804/mercury-relayscontactors-30-to-100-ampbrmdi-mercury-relays) seem to be pretty cheap BTW.
What worries me is that these things are relatively big (10 cm). The electrode seems to be a relatively big chunk of copper so inductance isn't going to be too bad ... but an impedance matched coaxial assembly it ain't.
-
Does it have to be square (or squarish) pulses? IEC 61000-4-4 uses exponential-ish pulses. I don't know offhand if there are other, more demanding, tests (MIL specs perhaps?!).
The avalanche transistor solution would be excellent, easy to get cheap parts but it can’t handle a pulse width of more than 15ns. I need 50ns – 1us pulse width.
Why can't you switch a TL with avalanche transistors?
Typically, they stay switched on for several microseconds (recombination time). The usual example (single transistor) uses a transmission line; I should think it would work with many in series too.
Tim
-
I seem to recall that firing of the explosive charges used to compress a nuclear weapon core had electrical requirements somewhat similar to this. The multiple detonators all had to be fired with sub-nanosecond timing accuracy, so it became a matter of generating a very fast high voltage pulse, and coupling that to all the precisely measured lengths of coax leading to the individual detonators.
Of course in that case there was no pulse repetition requirement.
I think somewhere I had saved an article on the technology used, but can't find it now.
-
Krytron (http://en.wikipedia.org/wiki/Krytron), previously discussed. Dismissed due to following disadvantages: delicate, inefficient, export controls, angry Libyans showing up, having to collect pinball parts for exchange. :)
-
The Russians make plenty of fast rise time pulse generators, I think they generally use SCR or IGBT followed by a magnetic switch to get to ns rise time and then throw it at a SOS to get it sub-ns.
I was looking for complete HV pulse generator for a while now. Best I was able to find are broken and require hard to find replacement parts.
If you know any supplier that sells them please share!!
Here is the link to a document describing "modernized" design of a sharp HV pulse generator.
URL: http://www.lle.rochester.edu/media/publications/lle_review/documents/v133/133_07_Solid.pdf (http://www.lle.rochester.edu/media/publications/lle_review/documents/v133/133_07_Solid.pdf).
I am trying to build down-scaled version of it right now for time of flight laser 3D scanner.
No tubes or Krytrons - I bet you can see those only in spy movies and in North Korean radars these days ;D Pulser stalk by itself is pretty old idea but with good MOSFETs and right cores you can get very good performance.
Here are parameters from that doc.
-
I like the design. The linear layout reminds me of a distributed transmission line amplifier while connecting the primaries in parallel and the secondaries in series reminds of how an Ft doubler works.
Transformer coupling is not suitable for long duration pulses.
-
Maybe this application note is interesting :-+
-
If you know any supplier that sells them please share!!
http://www.moose-hill.com/technolo.htm#Technology%20Transfer (http://www.moose-hill.com/technolo.htm#Technology%20Transfer) (might be a zombie website)
http://www.megaimpulse.com/ (http://www.megaimpulse.com/)
http://eng.iep.uran.ru/naudep/imp/razr/razr_2.html (http://eng.iep.uran.ru/naudep/imp/razr/razr_2.html) (academic, but they seem to have sold some of their devices to other academics)
There's also multiple companies making Pockel Cell drivers which are HV and sub ns :
http://www.fidtechnology.com/apps/fds/fds2-1M101B201.html (http://www.fidtechnology.com/apps/fds/fds2-1M101B201.html)
http://www.kentech.co.uk/index.html?/&2 (http://www.kentech.co.uk/index.html?/&2)
Here is the link to a document describing "modernized" design of a sharp HV pulse generator.
Hard to get sub-ns with transformers ...
It wouldn't let me download the PDF BTW. Had to to view it in google and move it to google drive to download it. Maybe only for US'ians?
Why would you need a HV pulse for a ToF laser scanner?
-
Maybe this application note is interesting :-+
This application note was one of the first things I considered mentioning but avalanche pulse generation has severe limits on pulse duration.
-
I used design based on Avalanche Transistors similar to one described in app note that eurofox linked. It is similar to what Jim Williams described in one of his app notes - it also uses regular germanium transistors (like 2N2369) in avalanche mode. Zetex claims that their transistors are designed specifically for avalanche mode, so you don't have to pre-select ones that work from large set and plus they can work reliably in avalanche mode for a long time. I have read somewhere that when almost everyone stopped making germanium transistors they found a factory somewhere in Russia that still makes them and bought it. Now they charge something like $12 / transistor.
Many laser scanners work in phase shift mode - laser diode sends continuous wave signal and distance is measured by phase shift from reflected one. TOF is on other end uses short pulse and its accuracy really depends on how sharp is the leading edge. One way of doing it is to send sharp and short pulse (~200V) to laser diode. It should be short enough not to kill the diode. Another way is to get laser working in continuous mode and then make pulse by turning the Pokels Cell on/off - in that case you need HV pulse. These are not all design alternatives of couse.
I think it would be cool to build a 3D scanner that can scan entire mountain. Artillery range meters can easily handle distances like 80km, so it should be possible to make 3D scanner that capable.
-
Marx generator? http://en.wikipedia.org/wiki/Marx_generator (http://en.wikipedia.org/wiki/Marx_generator)
-
I think it would be cool to build a 3D scanner that can scan entire mountain. Artillery range meters can easily handle distances like 80km, so it should be possible to make 3D scanner that capable.
A diode ain't going to do that ... you're going to need mega or gigawatts of peak optical power.
-
I’m looking for a part that can switch 4000 volts into a 50 ohm load within a nanosecond. The pulse duty cycle should not exceed 1 microsecond (80 amps for at maximum 1us). I was thinking of a mercury relay or a hydrogen thyratron, but the ones I have seen seem to slow or lack the specifications. Has someone used, or know of a part that is capable of this? Maybe a high power VHF/UHF transmit tube?
If I am reading your requirements correctly, there is no solution for your <1ns response time. It is possible to make such a pulse but the time from stimulus to pulse is normally 2-4ns, twice your requirement.
An old rule of thumb is under 2ns requires a pre-trigger.
Turning on 4kv with ~200ps rise time is definitely possible but I believe you are talking about clamping a non characterized transient.
I believe you will need to use a suppression method to extend the response time. Hopefully your application can compensate for such colouration.
Sent from my iPad using Tapatalk
-
A diode ain't going to do that ... you're going to need mega or gigawatts of peak optical power.
Range finders on laser diodes exist that can measure distances up to 10km. YAG lasers used for range meters that capable of 20km. I am talking about off the shelf models. Here is one example: http://www.laseroptronix.se/dispu/dm10000e.html (http://www.laseroptronix.se/dispu/dm10000e.html). Longer distance is possible but it requires favourable weather conditions to use of course. It is not all about power. Since diode laser used to pump the crystal inside YAG laser it is technically a hybrid.
-
Okay, I guess low single digit MW is enough on a clear day ... it's not all about power, but the peak power of laser diodes is enough orders of magnitude too low for power to be the limiting factor. You need a rod or fiber laser (preferably near 1.5u, much smaller wavelength and it reaches the retina, much longer and it dumps it all on the outer surface of the cornea). Pump sources can be lasers diodes, but they don't need to be driven with ns pulses.
-
If I am reading your requirements correctly, there is no solution for your <1ns response time. It is possible to make such a pulse but the time from stimulus to pulse is normally 2-4ns, twice your requirement.
An old rule of thumb is under 2ns requires a pre-trigger.
Turning on 4kv with ~200ps rise time is definitely possible but I believe you are talking about clamping a non characterized transient.
I believe you will need to use a suppression method to extend the response time. Hopefully your application can compensate for such colouration.
Sent from my iPad using Tapatalk
The switching element needs to go from non-conducting to conducting within a nanosecond, the switch delay may as well be in the millisecond range. I will probably use a mercury wetted reed switch, embedded in a coil. So the search continues for a mercury wetted read switch rated 4kV, 80 amps for a microsecond and preferably 50 ohm impedance…
Today I tried the read switch setup with a normal, non mercury wetted, reed switch. Just at 40 volts and with a 22nF cap instead of 200 meter coax. Measurements with a 500 MHz scoop got me under 2.5ns, getting to the end of the scoops bandwidth. The first bounce was about 1 us from first contact. Next test will be 100 meter coax and some test gear with an higher bandwidth (spectrum/network analyzer and some matlab/excel reverse fft?).
-
If I am reading your requirements correctly, there is no solution for your <1ns response time. It is possible to make such a pulse but the time from stimulus to pulse is normally 2-4ns, twice your requirement.
An old rule of thumb is under 2ns requires a pre-trigger.
Turning on 4kv with ~200ps rise time is definitely possible but I believe you are talking about clamping a non characterized transient.
I believe you will need to use a suppression method to extend the response time. Hopefully your application can compensate for such colouration.
Sent from my iPad using Tapatalk
The switching element needs to go from non-conducting to conducting within a nanosecond, the switch delay may as well be in the millisecond range. I will probably use a mercury wetted reed switch, embedded in a coil. So the search continues for a mercury wetted read switch rated 4kV, 80 amps for a microsecond and preferably 50 ohm impedance…
Today I tried the read switch setup with a normal, non mercury wetted, reed switch. Just at 40 volts and with a 22nF cap instead of 200 meter coax. Measurements with a 500 MHz scoop got me under 2.5ns, getting to the end of the scoops bandwidth. The first bounce was about 1 us from first contact. Next test will be 100 meter coax and some test gear with an higher bandwidth (spectrum/network analyzer and some matlab/excel reverse fft?).
Oh that allows many more possibilities.
I am skeptical about a reed being fast enough. I suspect the physical restraints will cause a Ton well over 8ns and a much longer Toff. Unless there has been great strides in reed design to overcome the inertia of the contractor to move through the self conducting range. To move the contractor beyond the breaking point would be much longer. So long that you will need to calculate contact life biased on arc over and in motion transitions. One thing that may help is there has been some new gasses developed in the last while that I don't have the details on. So please keep testing my assumption may be out of date now. I am just skeptical.
Sent from my iPad using Tapatalk
-
The first fast reference level pulse generators used mercury wetted reed relays to good effect but I do not see how a reed relay would work at high voltages.
http://www.amplifier.cd/Test_Equipment/Tektronix/Tektronix_other/109.html (http://www.amplifier.cd/Test_Equipment/Tektronix/Tektronix_other/109.html)
http://www.linear.com/docs/29168 (http://www.linear.com/docs/29168)
-
I think even if it was possible to get fast and clean rising edge out of mercury switch falling edge will be nowhere near desired result. If I understand right that is important requirement for most if not all pulse generators. Sander mentioned that he wants switching rate in KHzs in original question as well.
-
I think even if it was possible to get fast and clean rising edge out of mercury switch falling edge will be nowhere near desired result.
That's not really a problem with a transmission line pulser and a matched load.
-
A number of years ago I built acircuit for switching a Pockels cell. I don't recall the exact specs but it was a few kV switched in a few pS. this was before electronic record keeping so I don't have access to the schematic now that I am retired. I think by Googling "Pockels Cell Driver Circuits" you can come up with something close to what you want.
-
A friend of mine who used to work at United Electronics in the 50's figured this one out for us. The 4PR60 tube was made for this sort of duty. I would use 3 in parallel.
-
A number of years ago I built acircuit for switching a Pockels cell. I don't recall the exact specs but it was a few kV switched in a few pS. this was before electronic record keeping so I don't have access to the schematic now that I am retired. I think by Googling "Pockels Cell Driver Circuits" you can come up with something close to what you want.
I'm pretty sure I'm not going to find anything in the single ps range :) That's on the level of the fastest electronic pulse generators period.
Can you remember if you used a stack of avalanche transistor? That is mostly what google turns up for the really fast Pockel Cell drivers.
-
I Google is right 4PR60 is a "TERODE" tube, designed for RF not for fast pulse switching. It could be that this tube was used for pulser due to it accessibility?
In description of modern solid state designs I often see Thyratron mentioned as the default core of tube-based designs. 4PR60 is not a Thyratron .
-
I Google is right 4PR60 is a "TERODE" tube, designed for RF not for fast pulse switching. It could be that this tube was used for pulser due to it accessibility?
In description of modern solid state designs I often see Thyratron mentioned as the default core of tube-based designs. 4PR60 is not a Thyratron .
I don't want to be a jerk, but..... My recently departed friend Howard Brady designed that tube in the 50's at United electronics. It was used as a pulse modulator in radar systems and is related to the 5D21. the name is 4PR60
4 electrode, "PR" pulse requirement 60 watt anode dissipation.
typically they are limited to 18 amps pulse. but if you have a 1 microsecond pulse and a low enough duty cycle that the anode dissipation is not an issue, they can provide considerably more anode current that rated.
Here is a later data sheet from eimac.
http://frank.pocnet.net/sheets/140/8/8252W.pdf (http://frank.pocnet.net/sheets/140/8/8252W.pdf)
Not trying to be rude, but I just happen to have a little, personal,experience lets say with that design.
-
Oh yeah.. Thyratrons.... They have a practical limit of very slow speeds due to the ionization characteristics of the gas used. Typically hydrogen. In practice the limit is 1khz and in heavy pulse it can take well over a ms to turn on all the way. For precision switching you need a HEAVY DUTY pulse forming network because once they ionize they will conduct until there is no longer any real plate current because the anode cathode voltage will only be a few tens of volts to sustain ionization. Ignitrons are even slower, but when they conduct, BANG!
-
Oh yeah.. Thyratrons.... They have a practical limit of very slow speeds due to the ionization characteristics of the gas used. Typically hydrogen. In practice the limit is 1khz and in heavy pulse it can take well over a ms to turn on all the way. For precision switching you need a HEAVY DUTY pulse forming network because once they ionize they will conduct until there is no longer any real plate current because the anode cathode voltage will only be a few tens of volts to sustain ionization. Ignitrons are even slower, but when they conduct, BANG!
No, those are the common (for power applications) xenon and mercury thyratrons -- typical turn-on time is comparable to a fluorescent light (the gas part, not the filament part) or neon light, or... thyratron, and is in the 10us range. Compare to semiconductor thyristors, which deliver a cascade of charge carriers within a few microseconds, latching on (similar physics, really). The deionization time is the hard part, taking upwards of milliseconds (I guess one way to think of it, the ionization potential of atomic mercury vapor or xenon gas is low, so there is less incentive to suck electrons back up).
Hydrogen thyratrons are extremely fast, in the 10s of ns range I think (turning on), presumably due to the much smaller size of the ion (about two orders of magnitude), and the higher ionization potential (so the voltage drop will be higher). Deionization time I think is microseconds?
As for the 4PR60 and related devices, they were specifically made to handle extraordinary currents and voltages, but are otherwise absolutely plain thermionic vacuum devices. Notice the heater power (26V x 2.1A = 55W) is almost greater than the total plate dissipation alone -- these are massively disproportionate tubes compared to their continuous-wave brethren! They are made for extreme peak current (space charge, backed up by a massive filament) and high voltage (unusually large plate-grid spacing, though not at too much expense to operating characteristics, a nice achievement).
In peak power versus weight, you probably can't match these devices with semiconductors -- nah, that's not fair, if you count die weight versus whole envelope. But if you look at active cathode area and compare that, you'll probably come out ahead. 300kW is nothing to sneeze at from something three inches across! (Thinking about this further, a 4" puck SCR might have a 3" diameter die inside, which can switch 40kA peak at 4kV = 160MVA. The die itself might be 20 mil or so, just guessing; maybe it's not so easy to compete, despite the higher voltage. The speed-power tradeoff might still win, though you'd then have to compare against IGBTs or MOSFETs to be fair. Dunno.)
Tim
-
In my experience thyratrons are just not that fast. they begin conducting, then go into a limbo state for a bit, then by about .1 ms the arc begins, and that has a "Bloom" as the old timers put it. What i work with mostly dates to the late 70's at most recent, but the the rule of thumb I was given was over 1khz, look for something else. Also as they get hotter you get to a point particularly with mercury ones that they loose grid sensitivity and will just arc when sufficient anode voltage is present varying by temperature. We have had this happen actually in some of our welders.
-
Thanks for the datasheet calexanian. I would love to play with one of those and I can see that they are still available at affordable prices - "New Old Stock". I have a concern however, datasheet that you linked noted that these are pumped out to hard vacuum. After so many years lying on shelves somewhere could it be that some air can get in? One I found on Ebay also look kinda rusty :-\ I have purchased old photomultiplier tube in the past just to find that it has some air that got into it somehow.
-
Well, only way to know for sure is to test it.
There's usually a getter flash (a thin film of barium or other reactive metal) deposited on the glass, that makes it apparent what condition it's in. Though, high power transmitter tubes were often made of adsorbent materials (graphite, tantalum, etc.) which don't indicate gas presence -- and also must be operated at elevated temperatures (the plate usually glows red hot) to reach useful gettering activity.
It doesn't look like there's a flash on that one, so you probably need an electrical test to tell.
Bring it up slowly, and check things as you go. Start by applying filament power, run it at somewhat reduced heat and no electrode voltage for some time, then bring it up to full filament voltage, and check that the current draw is within spec. If it's too low, it may be worn (filament evaporation, excess filament voltage in a previous lifetime..?!); if too high, it may be running cold due to convection cooling (i.e., lots of gas).
Then, test the electrodes at low power to see if the characteristics match up with the graphs, then operate it at rated power (60W at DC should be easy to achieve on that sucker!) for a while. You probably don't want to go straight to full power, because some gas will have diffused in over the years (and out from the materials), which could cause internal arcing and damage. Allowing everything to reach normal operating temperatures gives time for the getter to do its job and actually improve the vacuum.
As for the vacuum itself, glass is an incredibly good container, and the glass-metal seals are also top quality. The most likely invisible failure is an extremely small leak at a poorly made glass-metal seal -- but one so slow that it wasn't caught during manufacturing burn-in test. So, these are very rare indeed. Otherwise, if it doesn't have obvious visible damage (smashed envelope, cracking around seals), it's probably fine.
Tarnish on outside electrodes is normal from old age, not really a good sign, but nothing fatal; it might be worthwhile brightening the higher amperage connections, just to keep contact resistance low. But don't be hasty; if you use chemicals, you may cause more corrosion later on, or if you use abrasives, you might cut right through a low resistance or [somewhat-]corrosion-resistant layer.
Do make sure to operate it with the specified hardware -- cooling ventilation as needed, heat sinking plate connector, etc. Overheated seals are a prime cause of early (operational) failure!
Tim
-
Thanks for the datasheet calexanian. I would love to play with one of those and I can see that they are still available at affordable prices - "New Old Stock". I have a concern however, datasheet that you linked noted that these are pumped out to hard vacuum. After so many years lying on shelves somewhere could it be that some air can get in? One I found on Ebay also look kinda rusty :-\ I have purchased old photomultiplier tube in the past just to find that it has some air that got into it somehow.
The zirconium coating on the anode will begin sorbtioning any evolved gas at just about cherry temperature. I can tell you this though. that tube was in the 1X10-7 scale and screaming hot when it was tipped off. The chances of a good tube having gone gassy, but not just outright bad is highly unlikely. In the exhaust process by about 200C the gas film on the inner wall of the glass has been driven off and by 300C where tubes are typically baked out at any gas occluded or diffused into the glass is mostly gone. At that point glass derived glass is pretty much just from decomposition of the glass itself at that temp. The metal parts will have already been degased by RF bombardment up to orange heat. The zirconium coating on the anode will give off most of its gas at that heat, and will actually re adsorb most gases, excluding noble gases, from cherry to red heat. This is a convenient means of refreshing transmitting tubes to a certain degree. Also the tungsten filament will combine with any remaining oxygen or nitrogen at red heat range, and give everything back of that was not converted to a compound on its surface at incandescence. The seals will be of tungsten/nonex glass at the bottom and the top will either be of tungsten nonex as well, or a kovar to intermediate glass, to nonex graded seal and thats going to be pretty good. I am reasonably sure if you pick one up NOS its still going to be pretty good. Please bear in mind those cathode current numbers are in pulse condition. Trying to draw amps from it DC will just strip it, or cause ionic bombardment and kill it that way. Bonded thoria (Distinctly different from thoriated tungsten) really is the preferred type of filament for this sort of affair.
-
I see one circuit on line for the 4PR60 which hits the grid with a 1 kV pulse from a variant avalanche transistor Marx ... is that the kinda circuit you would need to actually get it to turn on in the ns range?
http://www3.telus.net/schmaus2/elect/ftron.html (http://www3.telus.net/schmaus2/elect/ftron.html)
-
Check out the operating point, drive specs, curves... you'll need a hearty bit of drive, yes. A cathode follower from a smaller transmitter tube might do, or if you can get enough from something like that, sure.
Uh, worth noting by the way... this particular tube boats 50pF max Cin (Cg1k + Cg1g2), and looks to have about 2" leads into the electrodes, so easily 40nH per lead (not quite half for the cathode since it has two pins, but both grid and screen need consideration here!). Which is a cutoff around 112MHz. You'll be lucky to get nanosecond edges off this particular one.
More specifically, for pulses, you'll have a hard time terminating that stub properly. A planar triode is better for this, since you can just lay it in a stripline and call it a day. And even then, you'll probably only get so many dB per stage, so you might expect to use one 100W (average) device to drive two in parallel (for the final), and so on back to your signal generator.
Wonder if a shock line might be better. See what #61 or higher frequency ferrites do? Advantage is, it can be switched with an IGBT (or stack), or something like that.
Tim
-
that may have too much loss for switching that fast now that i think about it.
-
Now I remember where I saw that FBI rubbish about triggered spark gaps and terrorists with nukes. This is pretty funny:
http://www.fbi.gov/news/stories/2007/december/counterproliferation_121307 (http://www.fbi.gov/news/stories/2007/december/counterproliferation_121307)
http://www.rell.com/products/High-Energy-Transfer-Products.html (http://www.rell.com/products/High-Energy-Transfer-Products.html)
http://www.excelitas.com/downloads/DTS_Triggered_Spark_Gap.pdf (http://www.excelitas.com/downloads/DTS_Triggered_Spark_Gap.pdf)
Considering that those things are way slower than OP needs.
-
Slower but able to supply far more energy ... that said, I doubt Iran would have much trouble making a triggered sparkgap or Marx.
-
I’m still working on this project, so far I’ve selected the Russian GI-39b radar triode for the power stage (16A@20kV in pulsed operation, 1.2GHz BW). To get down to the specified 80A@4kV (50 ohms output) I was thinking of a 1:5 transformer on the anode of the tube.
Big question is, how do you select the appropriate core and how do you calculate this kind of transformers?
http://www.electronics-tutorials.com/basics/wide-band-rf-transformers.htm (http://www.electronics-tutorials.com/basics/wide-band-rf-transformers.htm)
This seems to give calculations once you have selected the appropriate core, but what kind of core is suitable? Ferrite? Air core? Something else?
Primary:
20kV peak
16A peak
1250 ohm input impedance
Secondary
4kV peak (40Vrms)
80A peak (800mArms)
50 ohm output impedance
Max dutycycle 1us
Max repetition rate 100Hz
Max rise time 1ns -> calculates to 340MHz bandwidth, let’s say 500MHz BW to be save.
-
The calculations shown are very basic calculations. The actual design calculations are extremely long, and rarely done by hand now. Last time I did the calculations by hand it took almost two weeks. Even then I had them verified by a transformer designer.
To get away from the extensive calculations and having custom core manufactured, a bit of trial and error is used.
Find out what core materials are available to you that meet you BW, Al and Mu requirements. Then find the core type of that material that meets your power, size and shielding requirements.
After these steps you should be down to only a few options.
You can attempt to have your design calculated by a transformer manufacturer. If you do you will receive a very long report detailing construction specs. Then follow their instructions exactly.
If you can't, guess. If you are too far off you will likely sacrifice efficiency. If efficiency is too low you risk core overheating. If this happens use a new core and try again.
You will need to do allot more research on transformer design. To determine the best wind pattern and buildup for you requirements and selected core. More knowledge will reduce your trial and error iterations.
Sent from my iPad using Tapatalk
-
Sub ns rise time with something like a spiral wound air core HV transformer is right out.
You'd need some type of transmission line transformer.
-
Maybe this ;D
http://www.ebay.co.uk/itm/Kentech-Instruments-Special-Avalanche-High-Voltage-Step-Generator-100pS-2kV-80kW-/271601660104?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&hash=item3f3cb870c8 (http://www.ebay.co.uk/itm/Kentech-Instruments-Special-Avalanche-High-Voltage-Step-Generator-100pS-2kV-80kW-/271601660104?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&hash=item3f3cb870c8)
http://www.ebay.co.uk/itm/Kentech-Instruments-HMP1-Avalanche-High-Voltage-Step-Generator-100pS-2kV-80kW-/271601660382?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&hash=item3f3cb871de (http://www.ebay.co.uk/itm/Kentech-Instruments-HMP1-Avalanche-High-Voltage-Step-Generator-100pS-2kV-80kW-/271601660382?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&hash=item3f3cb871de)
-
This seems to give calculations once you have selected the appropriate core, but what kind of core is suitable? Ferrite? Air core? Something else?
Try Mu metal, Supermalloy and Amorphous cores such as ones made of METGLAS. These are made of thin strips of amorphous metal placed inside a plastic or sometimes aluminium. Here is good example: http://www.ebay.com/itm/Metglas-FineMet-FT-3-KM-K2214B-Single-Core-Lot-of-3-/161161180107?pt=LH_DefaultDomain_0&hash=item2585f473cb (http://www.ebay.com/itm/Metglas-FineMet-FT-3-KM-K2214B-Single-Core-Lot-of-3-/161161180107?pt=LH_DefaultDomain_0&hash=item2585f473cb). These materials are best suited for pulse applications.
To avoid all sorts of problems with unknown cores you will need to build a rig or purchase tool that will allow you to get Magnetic Hysteresis curve for your core on your oscilloscope or computer. Also please pay attention to details such as if core material was annealed or not. Because core is harder to find and change I would start with obtaining suitable core(s) and then calculating number of turns after obtaining its exact properties experimentally.
-
The calculations shown are very basic calculations. The actual design calculations are extremely long, and rarely done by hand now. Last time I did the calculations by hand it took almost two weeks. Even then I had them verified by a transformer designer.
I'm curious, what kind of calculations would take two weeks to perform? Surely not E&M solutions, because that would take either months of pure grinding, or months to years (if possible at all) doing analysis.
Anyway, you're looking for a transmission line transformer. The bandwidth is approximately the electrical length equivalent of the transmission line(s) used, so you need to use less than about 2ns worth of line. Which isn't much.
With most of the energy in the 1-300MHz range (centered around 5 or 10MHz maybe?), you'll need a high frequency ferrite, such as #43. You want a part which has enough I.D. or winding area to hold the transmission line, with an impedance curve well enough above your spec. You also need enough core area (Ae) to withstand the pulse, which puts this beyond the scope of mere small-signal analysis.
NiZn ferrites saturate around 0.3T, so let's say you run up to 0.2T peak. A single turn needs to handle 4kV for 1us, or 4000uWb. Or proportionally less for more than one turn. So, N*Ae > 0.02 m^2.
And, you need a perimeter (winding length per turn) under the limit of 2ns at say 0.67c = 0.4m line length, divided by the number of turns (5*N for this case).
Suppose the core cross section is circular (the best case, since perimeter can't be any smaller for the enclosed area), and that the transmission line thickness is zero (so we don't have to add it to the perimeter -- a lie, of course).
Solve for N (the secondary turns), given that the ratio is 5:1 (Np = 5*Ns, Ns = N).
Let: total transmission line length = L, core area = Ae, r = radius, p = perimeter.
L = 5 * N * p
Ae = N * pi * r^2
p = 2*pi*r
so p = 2*pi*sqrt(Ae/(pi*N)) = 2*sqrt(pi*Ae/N)
so L = 5 * N * (2*sqrt(pi*Ae/N)) = 10 * sqrt(pi*N*Ae)
For L < 0.4 m,
N*Ae < 509.3 x 10^-6 m^2
We already know,
N*Ae > 0.02 m^2
which is 39 times too big. Which, I think, means your amplitude*bandwidth factor is a little optimistic for this material -- you'll saturate it.
Put another way, the largest enclosed area of 0.4m perimeter is a single turn circular loop, with area 0.012 m^2. Applying 4000uWb of flux to that loop (this doesn't depend on having a core) creates an average B field of 0.33T already, which no high-frequency ferrite can do.
Other ways to put that:
The radius of said loop is 63mm. Assuming mu = 100 (a modest figure for something like #43 at frequencies in the 10s-100s MHz range), the velocity factor is maybe 0.03, so the radius corresponds to a roll-off frequency of ~140MHz (give or take what wave pattern is rolling off), not counting skin effect due to lossiness. The effect will be, the ferrite closest to the transmission line will first saturate, then more and more, as the saturation follows a wave towards the center, until finally the whole thing is saturated, but by then your pulse is almost over. The bulk effect is, a change in the apparent permeability (and lossiness) of the core; it has much lower dynamic inductance than you would've expected from small-signal estimates alone.
If you can get by with pulse widths no more than 25ns, I think you'll have an easier time. On the upside, you can use quite modest lengths of transmission line for the PFN.
Tim
-
The calculations shown are very basic calculations. The actual design calculations are extremely long, and rarely done by hand now. Last time I did the calculations by hand it took almost two weeks. Even then I had them verified by a transformer designer.
I'm curious, what kind of calculations would take two weeks to perform? Surely not E&M solutions, because that would take either months of pure grinding, or months to years (if possible at all) doing analysis.
Yes it was. The electrical calculations are ok but when you add in the magnetic domain it got painful. It was manual as in pen and paper, and it was not fun at all. We didn't keep going through the cycle endlessly. We just repeated until the errors where within specs. Then sent in the buildup to a contact to put it through the computer to make sure we didn't miss something. That was defiantly a special case. It worked in the end but wouldn't like to do it again.
Sent from my iPad using Tapatalk
-
As far as I know only extremely exotic materials will add much at 500MHz, except for loss that is. Air is the only real option I see.
How about a Stacked Line Transformer? Wind 25 pieces of (micro-)coax on a cylinder, on one end connect them all in series, on the other 5 parallel 5 series.
-
I feel little bit lost here. We are talking about fast (~1 nanosecond) HV pulses. Saturable core is what going to allow us to make fast switching of 100 Amp. So if we don't have core what are we going to saturate? Also if in our calculations we are trying to avoid core saturation then how can it be "saturable core"?
Here are few extra papers with good overview on subject that I found in my collection. I think it explains the concept pretty well. If anyone here have links to better sources, papers, webpages - please post!
Most of designs use METGLAS, it is accessible, but not as easily available as other amorphous materials. Disadvantage of materials with low saturable induction is that cores have to be big.
This for example: http://www.aliexpress.com/item/Nanocrystalline-core/1940784676.html (http://www.aliexpress.com/item/Nanocrystalline-core/1940784676.html) shows saturable induction of 1.2T. Metglass is in comparison 1.57T
-
I feel little bit lost here. We are talking about fast (~1 nanosecond) HV pulses.
Well OP "just" wanted to step down a pulse from his radar 20kV triode, which was already ~1 ns rise time to begin with ... so a saturable reactor pulse sharpener wouldn't help him much. Also I doubt you could get below a ns with a saturable reactor (generally they are used to ram a silicon opening switch for that). No magnetic core will do much at 500 MHz, so if he wants a normal transformer the only core option is air ... luckily since it only needs 50 Ohm output impedance he doesn't need a whole lot of inductance, a Stacked Line Transformer might work.
Though working at 20 kV to get a 4 kV pulse is a questionable design decision IMO.
-
Though working at 20 kV to get a 4 kV pulse is a questionable design decision IMO.
I want to watch the smoke test . . . on video or from a safe distance.
-
Triode won't give you 1ns, maybe 500 ns or much worth. I think you need saturable core to make it sharp. I think the paper that I attached explains it pretty well.
-
It's 1.2 GHz capable, so sure it can ... just throw a mercury wetted relay on the anode.
-
No magnetic core will do much at 500 MHz, so if he wants a normal transformer the only core option is air ... luckily since it only needs 50 Ohm output impedance he doesn't need a whole lot of inductance, a Stacked Line Transformer might work.
Sufficiently small (or layered) ferrite cores work just fine at 500MHz (lossy, but some impedance is better than almost none), the problem is, nothing will give anywhere near the flux capacity or average inductance to handle a huge 1us long pulse. Please read my analysis -- a loop of maximal area of maximum perimeter (limited by transmission line length and risetime), with or without a core, must contain a flux density over 0.4 tesla -- not only is that beyond the range of any ferrite nearly fast enough for the job, one can only imagine how much current draw must be required to achieve that much flux density in free air! It will short itself out in a matter of nanoseconds.
Tim
-
Triode won't give you 1ns, maybe 500 ns or much worth. I think you need saturable core to make it sharp. I think the paper that I attached explains it pretty well.
What? What the heck triodes are you even talking about, have you even used one before? I can't even think of a device so slow; a 12AX7 can be used for pulses far faster than that. Perhaps you mean "semiconductor triodes" -- archaic terminology, but technically apt?
Planar triodes have been used in nanosecond pulsers many times. Cite: Stanford papers on pulse generation; there are many to pick from, all interesting reads. Methods range from magnetic compression (usually with IGBTs pulsed beyond datasheet specs, working into nanocrystalline material or square ferrite) to shock lines (ferrite / ceramic dielectric loaded transmission lines) to planar triodes, avalanche, step recovery and so on.
Tim
-
a huge 1us long pulse.
Oops, forgot about that.