• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

cascode tubes

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey Everyone,

Which tubes make good cascode tubes? MJ writes that the only tubes designed to work with the cascode are the 6922/6DJ8. I have a 12AU7/12AU7 long tail phase splitter in the amp now and at high volume there is too much distortion. Its a Dynaco Mark VI with four KT88 outputs. I could use a 6922 in the preamp position and keep the 12AU7 in the phase splitter position. Or use 6922's in both ala Joe Curcio's Mark III design. But would the 6922 as phase splitter have enough drive for four KT88's? And if I use the 6922/12AU7 combination could both tubes be run with the same plate voltage? I often see the 6922 run at low voltages, around 90 or so.

Kevin
 
Last edited:
That is a good for cascodes tube too, also ECC88, PCC88 and all the variants. Yes, the ECC2000 is slightly better and less noisy then the 6DJ8. It was the last tube that Valvo developed in the late 60th so it is fairly recent. It is a bit rare though and may have gotten expensive.
How it sounds against the 6DJ8 is hard to tell. The circuit is important and also the particular manufacturing, older ones are usually better. I posted my reply because the ECC2000 is optimized for cascodes from the beginning. When you study the specs you will notice that the two systems inside are slightly different. One system belongs in the bottom and one system belongs in the top.
 
Other cascode valves are ECC84/6CW7 and 6BQ7. Most cascode valves, having been designed for VHF receiver RF stages, are likely to have some degree of remote cutoff characteristics so not suitable for audio except for very small signals such as phono preamps. ECC88 is an exception, as there is a separate remote cutoff version - the ECC189.

Cascode valves are designed to work well at low anode voltages, and in some cases this may mean that they don't like high anode voltages.

The 12AU7 has low mu, so should only be used as a LTP phase splitter if the tail is a CCS.
 
Has anyone tried the 6N14P in Cascode.

The data sheet always shows it with section 1 on the bottom and section 2 on top which may indicate it was designed for Cascode operation.

It is an HF tube so no audio related data present.

Likewise, 6N23P-EB shows Cascode operation but it is supposed to be an equivalent of the 6DJ8/ECC88 so that would be consistent with that tube.
 
Been there; done that. I used cascoded 6BQ7s (this type doesn't specify audio in its spec sheet) as an LTP phase splitter/gain stage. With an active tail load (cascoded BJTs) it performed quite well. Excellent phase balance and very low distortion.

This was one of those cases where there was very little information, though audio cascodes are quite common in solid state practice. It was a case of try and listen while hoping it wasn't common because it was strange, and not because it had been tried and sounded horrible. The extra gain eliminated another gain stage while allowing for good input sensitivity even with gNFB applied. There are a few examples "in the wild", the "Hedge amp" described in a 1965 issue of Wireless World, but it didn't catch on commercially.

The 6BQ7 was one of those types intended for VHF small signal amplification out to 300MHz, however it still works just fine as an audio triode though good loadlines aren't so easy to find, and there is something definitely squirrelly about the plate characteristic (undocumented variable-u characteristic?) You need to get that VPK up above ~120V. You also have to watch out for the cheaper versions that have series connected heaters: these ring like bells. The ones with parallel connected heaters don't seem to have that problem.

The 6BQ7 also has a u-factor that falls nicely between that of types like the 12AU7 or 6SN7 and the 12AT7.
 
there is something definitely squirrelly about the plate characteristic (undocumented variable-u characteristic?) You need to get that VPK up above ~120V

Not entirely undocumented. look at the curves on the 4th page of the GE data sheet for the 6BQ7A. Mu varies from 30 to 40 as the plate current goes from 1 to 5 mA. The effect diminishes as the plate voltage is increased.

6BK7 and 6BZ7 are similar "cascode tubes" but the varying Mu is worse.
 
So I'm just a young buck with nothing better to do but shouldn't you be able to use about any tube for a cascode stage? Best I can tell it's going to be dependent on the available Vg swing per a particular tube but if you design out what you need and go from there.....I'm failing to see why any one particular tube would be preferred for a cascode stage. I do quite like the 6N1's I've got as an input stage for an OTL headphone amp at the moment.
 
Well... Look at what would be optimum... low mu tubes are fine.. as long as the gm is strong... The top tube is essentially acting as a voltage regulator to lock in the plate voltage of the lower tube, since we want the gm to remain constant "linear" as possible on the lower tube..so high gm is good for tighter regulation... The top tube also acts as a trans-impedance amp..
The lower tube is best set to be biased at it's best gm .... just like operating a BJT at the best part of the Beta curve.. I prefer to look at the gm curves.... The higher the gm of the lower tube, then the higher the cascode gain stage will be..
Essentially making a fake "pentode" from two triodes.. since TOP cathode is forcing lower PLATE to behave as a "screen" ...
I just make sure to float the heaters with DC so safe CASCODE operation can occur without cathode arc over to heater.. I believe the CASCODE designed tubes have these optimum features already.
 
Last edited:
jlangholzj said:
So I'm just a young buck with nothing better to do but shouldn't you be able to use about any tube for a cascode stage?
Yes, provided that you have sufficiently high supply rail voltage. An ordinary stage can already be a compromise regarding the anode voltage. A cascode doubles this problem. Hence cascodes often use lowish mu valves, or those intended for low anode voltage operation.
 
The only distinguishing characteristics are particularly high Gm, and high Ia at low Va (to make the direct series connection worthwhile, otherwise you need annoying parafeed networks).

Examples are 12AT7, 6BK7, 6BQ7, 6DJ8, etc.

The only application under consideration being VHF amplifiers. Alternatives being the grounded grid triode and the plain old pentode. The former suffers from poor isolation and perhaps low gain, while the latter suffers from excess noise. At much higher frequencies (UHF-EHF), there's no time for extra electrodes (literally) and GG triodes dominate (the cream of the crop being planar triodes like 7211 having high gain, low capacitance, and extremely precise electrode spacings). At lower frequencies, pentodes dominate (the extra gain being greater than the excess noise).

Down at audio, where noise is of nearly zero concern, gain is limited by device characteristics rather than impedances and bandwidths, and impedances aren't even matched to begin with anyway -- absolutely none of this matters, and you can do whatever you want (which, I think, the audiophile community has illustrated ably).

It is quite fortunate that 6DJ8's curves are so linear, though. And I'd say 12AT7 demonstrates how linearity isn't a requirement for the intended application...

Tim
 
Sch3mat1c said:
Down at audio, where noise is of nearly zero concern
Is it?

Cascodes need to be used carefully for audio. They have poor PSRR and poor linearity. These pull in opposite directions: poor linearity means cascode can only be used for small signal inputs, but then poor PSRR means you need a very clean supply rail. They also have very high output impedance so stray capacitance (or Miller capacitance in the next stage) can limit HF.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.