833a Amp

[Richard]

....and wait till you get that 150,000V HT on then we'll gaze towards the heavens and say softy to the grandkids, look, see that supernova, that was uncle Philip that was...

[Richard]

-- and they'll whisper back, yes Grandad... but it looks more like a TV stand...

[pre65]

I remember a couple of years ago when I did an 845 power stage that used the same 950V HT, several forum friends warned me of the dangers and begged me to be careful.

Those warnings have been with me since then.

[pre65]

The plan is to use the 300b output transformers (my "small" 5K Hammonds) to drive the 833a grid using the 16 ohm tap.

The OPT in the picture below are my "big" 5K Hammonds and are rated at 200ma, and they are bloody heavy.

The amp below is my 300b SET driven by ECL82 with cathode follower (CCS loaded) and my dual mono PSU.

This, all combined with the two tv trolleys, is "the beast".

Originally Posted by Richard

....and wait till you get that 150,000V HT on then we'll gaze towards the heavens and say softy to the grandkids, look, see that supernova, that was uncle Philip that was...

that comment cracked me up

[pre65]

Originally Posted by Dave Cawley

As admin on this site I must point out this project has the capacity to instantly kill and also to burn a house down, and easily both.

The proposed very long wires could easily set up a feedback oscillator with might be a big problem. The use of 1,500 volts at half an amp would instantly kill you.

But perhaps the most worrying is the 1,500 volt return through a SMPS which will simply not handle it and is likely to instantly catch on fire and put a destructive voltage through the mains affecting your own and adjacent properties.

Please feel free to talk about this on this forum, but it must be noted that the forum owners cannot approve or endorse any such home build. Further more this forum cannot be held liable for loss of life or property by ignoring this notice.

As Dave has invited free comment about these alleged dangers I will attempt to put the matter right.

The cathode on this amp is zero bias, so the only voltage on the cathode is the filament voltage of 10V.Dave seems to suggest that HT will be on the cathode (and hence the SMPS), but this is not so. It can't happen.


Also the HT is 950V and not 1500v. My main PSU will not give more than 950V in cap input even though I would have liked a bit more HT.

If this breadboarded version works as intended I may have to get two bigger transformers for version two.

[NickG]

Not trying to fuel the discussion but:

The cathode on this amp is zero bias, so the only voltage on the cathode is the filament voltage of 10V.Dave seems to suggest that HT will be on the cathode (and hence the SMPS), but this is not so. It can't happen.

"Can't happen" should never be considered in enginearing. What you mean is should not happen, which is entirly different. It may take more than one fault condition to happen at the same time, but its still entirly possible.

The obvious one would be if the 0v link on the grid was disconnected (cap falls off) and then the triode becomes a diode, passes enough current to disconnect the link between the finnament and earth. Once that happens the SMPS is connected via a forward bias diode to your B+ supply.

[pre65]

OK Nick, I accept what you say.

The grid is not cap coupled though. I suppose the transformer that couples the driver to 833a grid could fail.

How fail safe can any valve amp be ?

[NickG]

Well, it can be made as safe as you want, just the normal cost/benefit trade off.

For example you could avoid this problem with a 1:1 isolation TX in front of the SMPS, you are not stopping it from failing, but you are preventing the failure from escalating.

I think the comparitive impedances of your B+ supply and the supply into the house would mean any effects of connecting the two together would remain local to your mains supply. it would certainly cause any transformers you had on the mains to saturate though :-)

[pre65]

Would adding a grid leak resistor as well safeguard against coupling transformer failure ?

In fact, my first sketch for this amp had a grid leak resistor, but I took it out for the subsequent drawings.