Valve amp input circuits

[Ian - G4JQT]

All valve amp circuits appear to have the input grid resistor of between 100k and 2M for use with older equipment like crystal mics and pick-ups, etc. But what about when using modern low impedance sources, specifically the headphone output of MP3 players and similar that expect a load of 100 ohms or less?

Is there another reason for such a high-value grid resistor?

I'm building (well, almost finished) the Gerry Wells WADAR amp, and he designed it with two volume controls for use as a PA amplifier I believe. (See part of circuit below.) The WADAR amp is probably overkill for what I need, but it's nice to keep this design alive.

For my purposes I've changed the original 220k VR1 for a 1k preset pot, as the amp's input will be the headphone output of a modern TV or MP3 player. The 1k is till a bit high, but a better match than 220k. It appears to work well like that, so a matching transformer is unnecessary. (If there's too much gain I expect I can remove C2 to add negative feedback to V1.)

The point is, no matter how high a value the grid resistor is, in all designs where it's variable the grid is connected to the wiper, so V1 grid may have only a few ohms to ground if the input signal is large in any case. I guess I could even put a 100 ohm fixed resistor across the input so the headphone output is happy.

It feels wrong to me, but I don't have enough theoretical knowledge to know if it is, and I haven't found any obvious explanation on the web - neither do I have the classic valve amp design books by Blencowe or Jones!

Any thoughts?

Thanks.

Ian

[cmjones01]

Don't worry about it. Modern transistorised electronics generally doesn't need to be loaded with a particular impedance. Its source impedance is effectively zero, or very nearly, and there's a lowest load impedance below which it won't be able to deliver enough current, but that's about it. A headphone output intended for 32 ohm headphones, or indeed a loudspeaker output, will be quite happy operating into an open circuit, or a 220k pot. You'll find that the input impedance of most transistor amplifiers is at least several tens of kilohms, or even more.

The time impedance matching becomes important is when the source has a non-zero impedance, such as a magnetic cartridge, or a significant length of transmission line such as a coax cable. Then it's important to load it correctly so that the available energy from the source is absorbed in the way the source is expecting.

Chris

[Herald1360]

From the valve grid's point of view, the grid resistor is only there to tie the grid to a fixed dc level- either a bias voltage or 0v if cthode bias is used. In an all valve circuit, as high a value as possible is used to avoid loading the previous stage's highish output impedance. If low impedance sources are used it can be whatever the recommended load for the source is, the valve cares not!

Small signal valves can tolerate higher resistances- a megohm or more, but output valves usually specify a max value depending on whether they're fixed or cathode biassed.

[Ian - G4JQT]

OK, thanks chaps for the replies. Seems I was on the right lines, but had forgotten I'm not dealing with matched impedances in domestic audio, and indeed the valve doesn't really care as long as there is some DC grid connection to ground.

I put a fixed 1k resistor at the input in place of VR1, and I removed C2 then did some tests. Only a slight drop in gain with C2 removed - still plenty of audio with input from cheap MP3 player output set for normal headphone listening - and a very slight improvement in harmonic distortion, particularly below 200Hz.

Ian