• 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.

300B SE with a DHT driver

Or, with your system, it may be worth exploring some thing around the DAC's linear stage.
Removing the factory opamp IV and replacing it with a decent valve stage: differential (2 triodes) and a high-quality differential-to-SE transformer (Sowter) will probably suit you very well! You should be able to increase the DAC output voltage while you are doing it.
 
Low-µ end-stages are best: leave the gain to the driving stage.

Hello Rod. What are the reasons why low mu end stages are better? Anything specific? e.g. what would be the drawback of using a 4P1L end-stage (mu=11). Or a higher mu=20 end stage? One thing about the valves I have in mind is that the bias voltages are quite low, like 10 to 20, if that comes into it.

Indeed, I've always thought the solution is in the DAC linear stage. I have very little experience of solid state, alas, which means a lot of reading up when I try to do anything. My understanding of digital circuits is virtually nil. I've been building with tubes for 15 years or so and nothing else!
 
> What are the reasons why low mu end stages are better? Anything specific?

hello Andy - µ is defined as gm x ra, and both of these are better to be low for an open-loop speaker driver.

Low gm is found in big DHTs with wide spacing between grid → anode, and filament. This makes for lower open-loop distortion in all cases, and especially reactive loads that spread the load lines into elliptical areas.

Low ra allows lower transformer impedance, and a more favourable damping factor.
 
Thanks Rod - I get it. I've been listening to the EL12n as an output for a day or so, and mu there is around 18. I don't have the triode curves - does anyone? I believe anode impedance is 1.2K and gm is 15. My present operating point is 230v, 70mA. I need to increase the a-k voltage.

But anyway, the bass is very good to my ears. Quite tight and punchy. The top end is a little grainy, and if I could fix that it would be a very good output tube - it's very detailed. My OPT is LL1664/70 at 3K and that really should be double that. At least 5K or even 7K. I doesn't have a 4 ohm tap but I have other 3.5K OPTs that do, which I could try.
 
Yes, with DHT driver, you will need step-up, to make the output from your preamp or DAC enough to drive 300B to the full.

OTOH, If you can afford buy the EML20/30 series driver, and have a high-output preamp (4-5V zero→peak) you don't need the IT.
 
It depends on your system and your speaker sensitivity and what voltage you're putting into the amplifier, 2v or more or less. In very ballpark figures my threshold with 2v in, Alpair 10M speakers 89db and a smallish room not playing music loud, is 26 driving PSE 4P1L. That's mu of around 9 driving mu of around 11 in theory though in practice could be different. That's my absolute minimum. So I would be OK with a mu of around 10 driving a mu of around 18-20 or vice versa. But your system may be quite different.
 
Ale uses a 1:8 stepup in his 46 input stage if you look that up.

Even the 1:4 is a compromise at line level (phono SUTs - for example 1:10 - working in few mV input region). HF behaviour is very critical.

It is no coincidence that many manufacturer do not produce such stepups, only 1:1 line input transformers.

My Sowter 9063 has upper -3dB point
as 1:2 at 100kHz;
as 1:4 at 62kHz.
 
Remember, the load on the input of a step up transformer is calculated to be 1/the Square of the step up turns ratio.

1:1 input impedance/output impedance = 1/1
1:2 input impedance/output impedance = 1/4
1:4 input impedance/output impedance = 1/16
1:8 input impedance/output impedance = 1/64

The circuit that has to drive an 8 times step up transformer will see 1/64 of the secondary's load.
Suppose the load is a 50k g1 resistor.
50k/64 = 781 Ohms
The driver has to drive 781 Ohms.
And 60pF Miller Effect Capacitance of g1 is 60pF x 64 = 3,840pF load on the driver.

There is no free lunch.
Cost
Weight
Magnetic hum pickup (more on some than others)
Real Estate space required
Load on the Driver (better with a 1:1; or a 0.5:1 step down), step up is more difficult

I have designed amps with interstage transformers.
I do not design interstage transformers into my amplifiers any more.
That is my Personal preference.

Your mileage may vary.
 
Last edited:
That's extremely useful! Many thanks. I still have the EL12n outputs in my system. It's really good on rock and popular music. Very detailed with punchy bass. I still have reservations about classical music and voice - I'd like it to be a bit smoother and more like a 300b. But I'm really enjoying Steely Dan with it. Must try different OPTs - I'm using a LL1664/70, 3K, at present.
 
Yes, there is no free lunch. I have a zobel in the secondary and no other grid load resistor.

Yet the performance is excellent and sounds extremely good. probably the best driver I tried so far
This is 200Vpp

47-LL7903-driver-03.png


However you need to drive it with a low impedance source and my DAC has no issues. I also have a slagle AVC at the input

Alternatively, you can drive it with a cathode follower (or SLCF which I prefer)

Thanks
Ale