3 Band tube eq

[Heikki]

At one point my brother and I were going to build a small tube mixer but we abandoned the project. Here's the EQ we were going to use.

It's based on the Jeffery high gain phase splitter to get very high gain from a pentode. Wireless World, August, 1947, page 274
http://www.americanradiohistory.com/Archive-Wireless-World/40s/Wireless-World-1947-08.pdf
Cathode follower is used instead of the phase splitter and negative feedback tone control is added. It's not really different from any 1 op amp 3 band eq plan that that Google can find, except it uses tubes. It can give approximately 15dB boost or cut at 50 Hz, 1 kHz and 10 kHz

It does need to be driven from fairly low source impedance (less than 2k) or there's some loss of highs. Distortion is very low even at very high output levels, assuming output load is 100k or higher.

I haven't built this but I thought it might be useful for someone who must use tubes for everything. Here's a version that can handle 600 ohm loads http://www.kolumbus.fi/~kp5188/EQ/3bandeqwcf.svg

 

[abbey road d enfer]

Here's a version that can handle 600 ohm loads http://www.kolumbus.fi/~kp5188/EQ/3bandeqwcf.svg

This version uses the SRPP-derived cath-follower (I don't know what name it's been given); it's an example of the common mistake of thinking that, because its output impedance is low, it is capable of driving a low impedance load.
The current drive capability is actually limited by the operating current, as always.
Indeed, the overall NFB helps but I don't think the ECC82 is happy to drive 600r directly, even in this configuration. For HiFi/domestic, it is probably acceptable, but not for pro use, where +20dBu headroom is the norm.

 

[Heikki]

I think it'll drive +18dBu cleanly to 600 ohm load before it clips horribly.

 

[gyraf]

I wonder if the now very-cheap ECF82-type (pentode-triode) tubes would do?

 

[abbey road d enfer]

I think it'll drive +18dBu cleanly to 600 ohm load before it clips horribly.

+18dBu into 600r needs absolute minimum of 17mA idle current. I can't see the component values in the schemo, but with 300V B+, so I don't know what operating current has been designed in, but that would be dangerously approaching the dissipation limits.

 

[ruffrecords]

Here's a version that can handle 600 ohm loads http://www.kolumbus.fi/~kp5188/EQ/3bandeqwcf.svg

This version uses the SRPP-derived cath-follower (I don't know what name it's been given);

It is a White follower named after the man who patented the circuit (attached). Like the SRPP is has the advantage that its peak to peak current swing is 4*Iq rather than 2*Iq in the case of the cathode follower. Although its output impedance is no better than a cathode follower it can output 4 times the power of a cathode follower running at the same quiescent current. Even so, I would not be tempted to ask it to drive +18dBu into 600 ohms without a step down transformer.

Cheers

Ian

 

[Heikki]

For anyone who can't see the component values in the schematics. If you're opening the svg files in your web browser, you can zoom in and the picture quality won't degrade. You can zoom in at least in Firefox and Chrome and probably in other browser too.

With White cathode follower there if I got the top plate resistor right with 10mA idle current I thought it could get fairly close to +20dBu to 600 ohm load. Maybe I'm too optimistic about the capabilities off the White cathode follower. 

Another place where that high gain circuit could be used would be active mixing circuit. The cheap ECF82 Gyraf mentioned would be a great choice. Cheaper and less heater power to use just one bottle. Of course these high feedback amps probably won't have any "tube sound".

 

[abbey road d enfer]

For anyone who can't see the component values in the schematics. If you're opening the svg files in your web browser, you can zoom in and the picture quality won't degrade. You can zoom in at least in Firefox and Chrome and probably in other browser too.

Using Firefox here, I can't find how to zoom in - G.. knows I've tried...

 

[Heikki]

Ctrl + to zoom in, ctrl - to zoom out. Or you can press ctrl and use the mouse wheel to zoom in and out.

 

[abbey road d enfer]

Thanks for the tip, never seen it documented...

 

[PRR]

If you have Menus visible, it is under View, Zoom, mouse or keystroke.

Recent Windows/FireFox defaults to hiding the dang menus. Tap Alt to get a peek. Try Right-Click over on the right of the space where you would expect menu to be, you get a drop-list of things including Menu Bar.

 

[abbey road d enfer]

If you have Menus visible, it is under View, Zoom, mouse or keystroke.

Recent Windows/FireFox defaults to hiding the dang menus. Tap Alt to get a peek. Try Right-Click over on the right of the space where you would expect menu to be, you get a drop-list of things including Menu Bar.

OK, I finally found how to have the Menu visible. Thanks.

 

[Heikki]

I did a test circuit of just the White cathode follower from this schematic http://kenotron.org/mixer/mixerwcf.svg except I used 6SN7 instead of 12AU7. I used Edcor XSM 600:600 transformer with 560 ohm resistor across the secondary. I got +18dBu with 2.5% THD and that was without the advantage of high feedback from the high gain pentode circuit.

http://tube.kenotron.org/wcftest.png

 

[abbey road d enfer]

I did a test circuit of just the White cathode follower from this schematic http://kenotron.org/mixer/mixerwcf.svg except I used 6SN7 instead of 12AU7. I used Edcor XSM 600:600 transformer with 560 ohm resistor across the secondary. I got +18dBu with 2.5% THD and that was without the advantage of high feedback from the high gain pentode circuit.

http://tube.kenotron.org/wcftest.png

What's the quiescent current then? 6SN7 is a different animal than ECC82, with almost twice rated plate power.

 

[Heikki]

Quiescent current is 10mA. ECC82 is similar enough to 6SN7 that I would expect similar performance.