Yes,
Fuses from some manufacturers are extremely inconsistent when purchased at different times. At first I thought I would need to test under 100 samples to get useful data, it now looks like sample size may rise to 1000+! So it will be a lot longer to complete the research. Also a bit busy at the moment.
The current conclusion is to only buy established name brand fuses.
My first experience with fuse failure came at a baseball park. Some of the low power gear would blow fuses after operating for months. Had to install higher current fuses to keep the system running! Of course had to measure current draw first. Not easy as one would assume!
Anyway, corona can exist undetecable by eye, but that is of minor importance.
Important is to keep the average fieldstrength safely below 4kV/cm to prevent high power discharge channels.
In my 40+ years expirience with 6-12kV transmitting tubes (industrial rf-heaters) i found a distance of 1cm/kV
to be plenty enough to prevent serious discharges.
(some occasional flashmarks still possible, but nothing serious enough to damage components or interrupt production)
Important is to keep the average fieldstrength safely below 4kV/cm to prevent high power discharge channels.
In my 40+ years expirience with 6-12kV transmitting tubes (industrial rf-heaters) i found a distance of 1cm/kV
to be plenty enough to prevent serious discharges.
(some occasional flashmarks still possible, but nothing serious enough to damage components or interrupt production)
Discharge distance is changed by a humid day! I have seen a demonstration from a fellow with a screwdriver learning what was safe to do in the winter was not come summer! Fortunately low enough current that he survived the lesson.
1 KV per CM seems to me to be reasonably cautious!
So far have only lost one friend from carelessness. Perhaps why I am a bit more cautious than some folks here.
1 KV per CM seems to me to be reasonably cautious!
So far have only lost one friend from carelessness. Perhaps why I am a bit more cautious than some folks here.
Purchased a bunch of these. Was a good tip. New they are $ 250 each.
Two in series with proper balancing resistors should be good for 4kV.
And I blew a heater supply, after I heard a discharge.
No idea what happened, maybe I should get proper HV cabling ...
Jan
Attachments
The fact that they must be able to deliver amperes of Ia (albeit short pulse) is the reason for the relatively high heater power.
Those high anode currents need a lot of electrons in a short time. But the average power dissipation is low.
Jan
Adrian Immler from Austria was so kind to put an LTspice model together for the GI-17.
It looks like a reasonable match to my measurements.
I have modified my current amplifier model based on the 6HS5 and next will see if I can port the GI-17 model into it.
The original design used the 6HS5 in grounded grid, driven with a cathode current.
With the GI-17 that is complex as the heater supply is connected at one side to the cathode so will swing the parasitic capacitances, which causes losses.
I now want to drive the GI-17 at the grid, but with feedback from a small sampling resistor in the cathode lead to remain in current drive mode.
This way the triode will look like an (ideal) pentode.
Exciting!
Jan
It looks like a reasonable match to my measurements.
I have modified my current amplifier model based on the 6HS5 and next will see if I can port the GI-17 model into it.
The original design used the 6HS5 in grounded grid, driven with a cathode current.
With the GI-17 that is complex as the heater supply is connected at one side to the cathode so will swing the parasitic capacitances, which causes losses.
I now want to drive the GI-17 at the grid, but with feedback from a small sampling resistor in the cathode lead to remain in current drive mode.
This way the triode will look like an (ideal) pentode.
Exciting!
Jan
Jan
IMHO you can check in yours local automotive shops for those 7mm in diameter spark plug silicone cables with insulated cooper stranded wire in the center ,those HV cables can be good choice for your amps , think that those cables are good for at least 20KV if not way more .
Attachments
This guy has all I need: https://www.benl.ebay.be/str/highvoltageshopc0m
And he is in my back yard ;-)
Jan
And he is in my back yard ;-)
Jan
This is my current amplifier:
https://audioxpress.com/article/you-can-diy-edd-a-direct-drive-amplifier-for-electrostatic-speakers
Jan
https://audioxpress.com/article/you-can-diy-edd-a-direct-drive-amplifier-for-electrostatic-speakers
Jan
Thank you for that write up! And for putting your life on the line handling potentially lethal voltage and current combinations. Somewhere I remember it's not the voltage that kills, it's the current, and the most lethal for the human heart are between 100 and 200 mA. I'm not sure I could trust myself to work single handedly, I'd slip up at some point or another. A friend of mine used to work with high voltage transformers in the nuclear plant business and made a mistake at one point. He was literally thrown backwards a few meters. Great to see the comparative distortion measurements, too. I can't remember that I've seen a comparison like that before. You mention very high SPLs, could you measure those as well?
Jan,
check out this pdf (vol 5, Glasoe and Lebacqz), <www.introni.it >pdf>05-Pulse Generators
Starting from page 106 it is about vacuum tube properties for use in high voltage pulse generators.
At the time, there where no special "impulse" type tubes yet, so they investigated
"radio" tubes that could handle, a.o.t., the high anode voltages needed for pulse radar
check out this pdf (vol 5, Glasoe and Lebacqz), <www.introni.it >pdf>05-Pulse Generators
Starting from page 106 it is about vacuum tube properties for use in high voltage pulse generators.
At the time, there where no special "impulse" type tubes yet, so they investigated
"radio" tubes that could handle, a.o.t., the high anode voltages needed for pulse radar
Last edited:
@gorgon53, I don't see any PDF on that page, or in the left side menu.
Nothing I see is about pulse generators.
Is it possible to download the PDF and send to me or send a link when it is open?
I'm definitely interested.
Edit: Google found it, it is a book by Glasoe and Lebacqz.
Definitely interesting. Thanks for the tip,
Jan
Nothing I see is about pulse generators.
Is it possible to download the PDF and send to me or send a link when it is open?
I'm definitely interested.
Edit: Google found it, it is a book by Glasoe and Lebacqz.
Definitely interesting. Thanks for the tip,
Jan
@Tubelab_com: George, reading thsi again, see the attached from the Beveridge direct drive amp. Vg2k is 36V and apparently that is enough for this application.
Maybe I should check that out.
Jan
Attachments
The beam regulator triodes like the 6HS5, 6HV5 and others are essentially a sweep tube with the screen grid left out. In my experiments using them on 650 volts a positive grid voltage was needed to get reasonable plate current. The high Mu (300) and Gm (65000) of these tubes makes them prone to spurious oscillation with my clip lead construction techniques. With 3500 volts on the plate (max spec is 5500) you will be in negative grid region and within the range of an opamp chip. Curves are furnished in the 6HV5 data sheet. The other variants 6HS5, 6JD5 and 6JK5 are similar, but often cheaper. These come in two plate sizes, 30 watt and 35 watt. I have seen both size plates in tubes with the same type number on them.