An interesting audio tube possibility

[KM1H]

I never heard of a 6JD5 but it sure looks promising as a driver or low power modulator tube. Especially as its cheaper than a 2A3

http://www.radiomuseum.org/forum/6jd5_forward_grid_bias.html

I'll let the deep thinkers here hash it out

Carl

[Opcom]

It's a very rugged tube IIRC from the color TV repair days. It seldom needed replacement in TV sets, unlike the much-abused horizontal output tube. I'm not an expert on this tube but I have toyed with it back in the day, as used TV tubes were essentially free and 6L6's were not. The following is just an opinion. I have not seen one of those tubes in over 20 years (except maybe in the truckload I got a couple weeks ago had a lot of TV tubes in it..)

I've never tried this tube for audio because of some serious issues discussed below, but I want to disagree on principle with the website's suggestion of tying the beam forming electrodes to the plate. They should be grounded so that the beam is formed and strikes the correct area of the plate designed for the purpose. If that is not done, the available plate dissipation will be reduced, but no one has ever said by how much. It is not clear in the author's article what might be expected, but then the datasheets are not exactly forthcoming in that respect either.

My concern with higher voltages is the ability of the socket to withstand them. These were always in high quality sockets.

The grid dissipation rating is not stated but the grid is pretty decent sized. Possibly zero bias at 400V? but if I were to try the tube for low power I would be more likely to use 600V and a little bias, because it would be very easy to drive in AB1. Would a "D" cell work for bias at 600V?

What worries me about using it with the beam forming electrodes tied to ground is that it takes a lot of plate voltage to get the plate current up so that the 35 watt dissipation can be taken advantage of.  

If the contacts for pins 5,6,8,9 were removed, the socket could take it. The tube has internal connections, so the question is, what are they connected to (supports/ elements?), and would arcing outside the tube from one pin to another be a problem?

2A3 at zero bias: 170V/240mA

6JD5 at zero bias: 170V/about 13mA (hard to tell so closely, and we know that individual devices may not all act the same in such areas of a published curve)

Assuming the 2A3 curves extend a bit, at zero bias it would take close to 180V for 250mA.
For the 6JD6, about 1480V would be needed to hit 250mA.

The reason I point these out is that the two are much too different to swap. But I think you didn't mean to swap, but to design-in. To get real power worthy of its plate dissipation will be a problem due to the high voltages involved -if the beam plates are grounded.

To run the 2100V necessary to make 250mA with zero grid voltage at the crest of the audio wave (stay in AB1), the idling current could only be 14mA for a 30W dissipation (per tube) and the bias would be about -6V. The plate to plate load impedance would be very high. 20K?

OK so those are the reasons it is not very attractive when used as intended. I expect you already looked at these facts in the curves.

The author is running +6V peak on the control grid (about +3.8V bias) and I think he says it looks like a 200 Ohm load. That is probably safe for the grid. He also says that a higher plate voltage might be used, but has not written about trying that. If the plate voltage is raised quite a bit more to 500-600V, maybe less positive control grid voltage would be needed. The question also comes in as to how much the beam forming electrodes not act as a second plate. They are good sized but not of the same metal and who knows how much heat they can deal with. Might it be reasonable to put a resistor between the beam plates and the anode to limit the current to the beam plates? Would it also be a good idea to drive the grids with MOSFETs?

If you try the author's method and also use higher voltage, please keep us informed because here we have another tube most of the audio guys probably do not know how to use.

One application the tube might be more easily and fully used with beam forming electrodes connected as designed would be a class C stage at 1000-1500V and then modulated, but its unusual capacitances might make it a challenge at high frequencies and the plate dissipation would have to be watched carefully due to the duty cycle being more than 2.5%, so perhaps 2-3 in parallel sharing the current would be interesting.

Have you contacted the author and asked if he has any further comments since it was written? I bet he has made some improvements or found some limits. The article shows he is very ingenious.

[KM1H]

Very good and interesting comments Pat.

I only glanced at the curves as Ive no particular interest in exotic or expensive driver tubes. Im more than satisfied with 6L6's and 807's and any percieved benefit from 2A3's is lost on me anyway....too many years with headphones chasing DX and contesting, dragstrip and stock car racing, and shooting guns

No, I didnt mean for the 6JD5 to be a swap, just a new design replacement at the same power level as the 2A3 family for those who must use triodes. Id say the designer included pins 5,6,8,9 as HV isolation and there are no tube pins looking at a Fleaybay photo and Id have to think that the socket didnt include contacts either. The 6HZ5 and 6JH5 seem to be similar and interchangable.

It should be an interesting project for someone with spare time.

Carl

[W2PFY]

i have been playing around with these tubes for some time. I purchased about 3 dozen of the above and variants. So far without fail all the tubes I have received are NOS and ALL of them are very gassy. I have been able to De-gas the tubes by running them at the rated filament voltage for about 200 hours each. You can see it will take a long time to get the gas out of about 1/2 dozen. Because of their gain, I thought they would make an excellent HV mobile  modulator or maybe even a GG stage for something like a Flex 1500.

Those were my thoughts and now there on hold till next winter. There are some hi end experimentation going on with the tubes but they failed to take popular demand for the tube.   

[KB3DKS]

Another one in that family, same pinout and spec, is the 6HV5A.
I have been eyeing those up for awhile. Nice big plates. Looks like a triode sweep tube.
If you ever run out of 5u4s there is the 9pin compactron 5BC3A. Looks a little better than the 5U4 but close.
Bill

[kg8lb]

Isn't a 6AV5 , triode connected  similar to the 2A3/6B4 performance ?  Often found for only a couple of bucks each . I am using a pair in a HB speech amp.

[k4kyv]

The 6JD5, according to the link, is a high mu triode, μ=300.  Nothing like a 2A3.  It probably has a very high plate resistance, just like a beam tetrode. Might be good for a low power class B modulator stage, but not a good driver for a class-B or AB2 stage.

The reason the 2A3 - 6B4 series, as well as the 845 make good class-B driver tubes is their low internal resistance.  This requires a low-μ or else a lot of negative feedback. Also good for driving a speaker, since a speaker, like modulator grids, may present a widely varying load to the previous stage.  This is probably the reason why audiophools have scarfed them all up and you don't see them at hamfests any more.

[KM1H]

Here is some info on the 6HV5

http://www.radiomuseum.org/forum/6hv5a_forward_grid_bias.html

[kg8lb]

And the 6AV5:


  http://www.tubelab.com/6AV5.htm