No, this will be my first use of the tube. Art Loesch, may he rest in peace, designed a Phono Preamp using the tube. I am using 198.75 VDC quiescent, at low current , 4.5 mA. so its only running at 35% of Maximum. All I need to do is swing plus or minus 18.75 VDC so large voltage swing linearity is not much of an issue with 6AH4GT as the Final tube.
I just wanted to build an amp with UNpopular tubes, that easily "takes out" every " linear load line, popular tube choice" SE amp, due to my overall topology, wiring, supply and general execution.
Jeff Medwin
Once lost linearity cannot be regained. If your build is meant to be "for best results", I wouldn't throw any linearity out the window at all. If it's not, then that's another thing.
Finding unusual solutions is an admirable and good pursuit. I myself mainly make amps that most people think are either insane or pointless (if interested, see long discussions on my designs on this board elsewhere). They do however 1) sim 2) measure 3) sound very good.
I've also - mainly for financial reasons - found some tubes that are not on anybody's radar as of yet.
You totally CAN find unpopular / unknown tubes that perform very well.
You CAN NOT find tubes that have bad curves, and perform very well.
Load lines are not a social construct. Linearity is not contingent on popularity. Unless oscillating, no amount of wiring will make even 5% as much difference as a good load line.
We must be looking at different plate curves for the 6GK5. If used with a high plate load (ideally ccs or choke) at moderate currents around its specified operating point it is quite a linear tube: Certainly moreso than the oft specified T7. I can vouch for its sonic merit in phono stages and when used as described above.
Generally badmouthing a tube because you do not consider it suitable for certain applications is pretty poor form. Among others, Poinz from Audiotropic is certainly enamoured of the tube as a driver for his much lauded 6V6pp design.
Generally badmouthing a tube because you do not consider it suitable for certain applications is pretty poor form. Among others, Poinz from Audiotropic is certainly enamoured of the tube as a driver for his much lauded 6V6pp design.
Last edited:
You make a good point about badmouthing a tube. Sorry for that. Just because I could not find a sweet spot with this tube does not mean it does not exist. I have no doubt this tube is good in guitar amplifiers. I could never make it work to my satisfaction in phono, even at its specified operating point.
As you know, this thread is about directly coupled 2a3 amplifiers, so we probably need the input/driver plate at around 130V or maybe higher. The filtered B+ is probably 400V or higher.
I attached a few potential load lines. The first is unfortunately not so linear (Violet Line - 10 mA). It swings -35V to +30V on a 1VAC input signal. In addition, by my calculation, the "Triode Characteristic Factor" is only 2, so you will have to by-pass the cathode. You could of course use an infrared LED instead to bias it or use a 1V NiMH battery or even global feedback.. whatever floats your boat! It won't be linear though...
So, perhaps you might want to use a bigger Rp and bias for lower current flow? As you see with the Rp=100k ohm (Green load line) it's still not linear, swinging from -30V to +25V on our 1VAC input. Also it's not possible to calculate the "Triode Characteristic Factor" since on my GE specs sheet, mu and Rp values aren't shown for 3mA.
Hmmm... getting tough but let's not give up! Perhaps a much higher B+ would help? Indeed with a filtered B+ of 500V and an Rk of around 72K we start to get a more linear response (orange line) - it appears to finally swing from -30V to +30V on a 1VAC input signal. However by my calculations, the "Triode Characteristic Factor" only gets up to about 2.5 (still much less than 4) so we would still need cathode bypass, LED, NiMH Battery or global feedback loop. perhaps I missed something? Probably...
I agree 100% with you though - it is poor form to badmouth a tube
Attachments
Last edited:
The late Arthur Loesch, college professor, in the USA, made a lovely phono preamp using 6GK5s, a well regarded circuit.
Jeff Medwin
Agree 100%, very good point Sy.
The 6AH4GT SE DC amp grid only needs to swing plus and minus 18.75 VAC .
For the majority of the time, it is doing plus or minus 10 VAC. Our conventional conceptualization and design process does not consider much other than full power out !!
Jeff Medwin
Then you don't need a high mu tube. You could use a medium mu tube that has lower Ri and possibly avoid cathode by-pass (like Sheldon has done).
For example, why not 6CG7 ? Nice NOS start around $5 apiece on fleabay. This is assuming you want to stick with noval 9-pin.
Ian
Last edited:
I want to avoid a medium mu triode DCed to a low mu triode like the plague. The basic two stage amp, with a low mu finals, say a 2A3 or a Type 45, needs a HIGH mu driver tube, IF you want the amp to sound lively at low listening levels.
****** With out this extra gain I like to have, the amp will only sound good when pushed hard, played loudly, which is not an acceptable trait for me and my daily hi fi use. *******
A mu of 70 6SL7 becomes the minimum mu I find acceptable with a low mu DHT finals tube like the 2A3 or Type 45.
Even better yet, is a mu of 100 tube, that is linear, like half a 12AX7 DCed into a 2A3 or Type 45.
BUT, you have to implement the 12AX7 properly, which most builders do not ever do. It includes, among other things, full Rk bypassing, multiple film cap bypassing, of the 12AX7, often using 4 to 7 additional expensive high quality film Rk bypass caps in parallel, carefully chosen values and types.
Only other alternative for high-performance results is you use a single REL - Richard Marsh MULTICAP ( Percy ) with perhaps a .033 uF DynamiCap E across it.
Most DIYers and Manufacturers do not employ the money to do it properly. What a shame.
Jeff Medwin
It would appear so. I looked at the RCA curves. There's also curves by GE, which are better.
The RCA curves are simply not for audio use at all. I wonder why GE and RCA have so wildly different curves?
The linear area is still quite small. It's apparent that the rp is not at all constant. I'd run it at 150V 15mA.
What is this T7?
I don't know if you're referring to me or soulmerchant (who does and has spoken for himself), but for myself I don't consider giving an honest opinion 'badmouthing' in any way. Based on the RCA curves, 6GK5 is not at all for serious audio use, and based on GE curves it's got marginal use (small linear area). I would still advise to keep looking.
There are a LOT of unknown russian tubes that have excellent characteristics.
But again; if you're looking for superb results, you simply are not going to find them without superb curves and load lines.
Doesn't need to be. If the plate load is >> rp, then what you care about is the variation in mu with plate voltage.
12AT7. It's a far more linear tube than most people think if it's run correctly. I have a lot of data on it.
If you're looking for an effects box rather than something sonically transparent, that's certainly one way to go about it.
I don't follow, please elaborate.
The 2A3, for example, requires certain voltage swing input to produce certain sound pressure levels on a system.
A lower mu tube, say a 6SN7, has to swing it's plate voltage that amount to get that sound pressure out of your system.
The 6SN7 needs a certain input voltage to do that. It is bigger than that required by 12AX7 to produce the same plate swing, and thus same sound pressure level.
Where in this does the mu come in? Both tubes have to swing the same amount, they just need different input voltages...?
I don't understand the significance of the mu in this?
I would say what matters are
1) curves
2) rp, load lines, plate load.
It's much harder to provide proper load lines for high rp tubes like 12AX7, with resistor plate loading.
Edit: I prefer to design gain structure so, that I can keep my volume pot on as high a setting as possible. I don't like to pay for expensive pots, so I prefer to keep them out of the signal chain as much as possible. (I know, it's still there, but...)
Last edited:
Hello,
You won't find the answer in curves, loadlines, and Rps.
If you built two all-out DC two stage 2A3 amps, one with half a 6SN7 front end tube, and one with half a 12AX7 front end tube, and played them long term at high and low listening levels, you would likely hear what I am suggesting.
Jeff Medwin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.