4-65a Interesting Eimac DH Tetrode

[Paul Barker]

I have had to do some digging aound looking for "affordable" Tetodes or Pentodes with screen capability of 600v. The search led me to the 4-65a.

My quest is for a valve I can triode connect at 600v.

I realise the 813 will cope with 1kv or more in triode connection, but it is filament hungry at 50 watts for the job in hand. Though there may be a job for it to do in a future power supply project. For this project I want shunt valves that are more than up to the task in espect to reliability but not excessively so in espect of space and weight constraints which are alredy a problem in such a project. Otherwise I would be using my TY4-400's like the one seen on my recent 833 output stage at Owston.

IT is true that the KT88 proper allows 600v g2, so yes this is pefect for my task. BUT I am suspicious that the 6550 has a very much limited g2 capability. WHy this matters to me is because I want reliability and I don't trust the modern production facilities to have made two different internals at all. I cannot be sure there is any diffeence internally between a kt88 and a 6550. Therefore I don't feel that where dependability and ruggedness matter a great deal that I can trust the production methods, materials and quality assurance of present day valves of this type.

As the original KT88 costs a small fortune it is not an option.

There is always the TT21 and TT22 but on an ebay search I quite surprised that although still cheap for what they are, eventually the market has copped on to them. SO even though I have a few of these which I could quite easily use for the job, I kept looking.

For affordability if we can limit our g2 requirements to 500v I consider best option is the RS components EL34 which I believe is a present day bargain. But I am not recommending this valve on an experience basis. I have never owned one. It is purely on a reading menus from right to left (cost) basis. But it is likely that this valve would do the job, because a G2 rating is for when it is fixed against a moving anode. When it is tied to the anode it can be used beyond the rating. However we are left here in the dark largely by the manufacturers as they don't often give is figures. In a power supply shunt valve application the last thing we want is an unespected valve demise. So I would rather stick to well tried and tested ratings.

Anyway I managed to win one of the 4-65a's on ebay for a reasonable price delivered to uk £18 although to be fair they normaly go for more.

But there are a few features which make them interesting. The heater is fairly economical 6.3v 3.5 amp. 65 watts dissipation with 10 watts g2 dissipation.

They were expecting it to be used in AB1 and 2 audio amplifiers aswell as transmitters. As audio output they could be an extremely efficient means of obtaining power output.

We don't have triode curves but with a g2 capability of 600v it is a lakely candidate for SET connection. And it has to be considered as a fun project for using the 10 watts screen grid as the anode and you then have another DHT to try possibly capably of 2 to 3 watts. The absence of beam plates / supressor grid would be an advantage when considering the valve as a triode. We have 3 options,

1/ground the anode and use g2 as anode
2/ connect g1 and g2 together in triode mode
3/ traditional triode mode

Then we have two oddball connections the well known Inverted Operation, but there is also a connection the details of which I forget now, but it goes something like screening the cathode from g2 with g1, using g2 as control, whence g1 becomes a "vertual cathode". The efficacy of this connection I have long since forgotten. But it could be experiemnetd with. There is a very significant reason for the "vertual cathode" which niow escapes my grey matter but I believe it would be worth a try.

Note, this is not "super triode connection" in which g2 is modulated and g1 is grounded. It is "vertual cathode" operation in which g2 is modulated and g1 is held slightly positive to the cathode so that it becomes a vertual cathode. I just can't remember enough detail about it, but there was a very good reason for the type of connection. It must have been something I read in Terman years ago.

My personal application is simply a shunt valve triode connected. But this valve looks worthy of experimentation.

[Paul Barker]

It turns out that there is prior art on this valve for audio:

"I am using the 4-65A at present and have traced the triode curves on the tester up to 400V . 845 curves are close enough , Ra is around the 1.6k mark , mu of 5 . Currently running mine with James 6123HS at 530V/50mA , really these need to be run at around 650V / 55mA to get some decent power . Here's a pic of the Mk1 breadboard with T20 driver , it's totally different now and has LCL DC filament supplies , I would recommend implementing DC fils with all the big Eimacs with spiral filaments such as 100TH , VT127 and 4-65A . Also don't forget the heat dissipating caps !!!!!! "

taken from.

[nic]

It turns out that there is prior art on this valve for audio:

"I am using the 4-65A at present and have traced the triode curves on the tester up to 400V . 845 curves are close enough , Ra is around the 1.6k mark , mu of 5 . Currently running mine with James 6123HS at 530V/50mA , really these need to be run at around 650V / 55mA to get some decent power . Here's a pic of the Mk1 breadboard with T20 driver , it's totally different now and has LCL DC filament supplies , I would recommend implementing DC fils with all the big Eimacs with spiral filaments such as 100TH , VT127 and 4-65A . Also don't forget the heat dissipating caps !!!!!! "

taken from.

Yes that was me . I have since discovered that 4-65A need to run at >45 watts dissipation . Under that the valve does not behave and the self gettering action fails to mop up any gas . It needs to be run with the anode a dull cherry red colour all over to operate correctly . 650V 75mA does the trick

Nic