gemini86 said:
Ian, I found your test paper on the 6sn7 mu follower. Good stuff! I plan on using a 6sn7 myself, mostly because it's what I have on hand. Curious though, what is it that's stopping the circuit from driving a 600 ohm load? The output impedance should be really low. Is it the circuit or the tube? I also have some 6bl7 sitting around. I've heard of people subbing those in 6sn7 circuits with some success.
The reason is standing current and ultimately plate dissipation. My circuit runs the 6CG7 at close to 5mA standing current. If we assume the mu follower is a perfect push pull circuit the most peak current it could pump into a load is 10mA or 7mA rms which across 600 ohms is:
V = IR = 7mA * 0.6K = 4.24V rms = +15dBu which means we have very little headroom if our normal operating level is +4dBu
In practice, due to other factors, it is less than this and distortion is also a lot higher.
Into 10K however it can in theory do 7mA * 10K = 70V rms = +39dBu.
In practice it can do about +34dBu into 10K.
So the obvious answer is to increase the current. There are two problems with this. The voltage drop across the 10K resistor between the tubes increases which reduces the voltage across each tube and thus the maximum output voltage. We could reduce the value of resistor but then distortion goes up and the gain goes down (and if you reduce it to zero you effectively end up with an SSRP). The only real solution is to raise the HT and it is already at 300V. Raising it to 400V would probably allow is to run the idle at 10mA but even this would only give us a 6dB improvement in maximum output level.
An alternative is to use a 10K:600 transformer with the existing circuit. The output level (and overall gain) would then drop by 12dB but we could then achieve +22dBu into 600 ohms (34 - 12), except the tube is running at a level 12dB higher than the output which means the distortion is 12dB higher than it would otherwise be and we still need to find 12dB of gain from somewhere.
The bottom line is the mu follower is not really suitable for driving power into a 600 ohm load. Remember that
+26dBu into 600 ohms is nearly half a watt of power. To do that you might think along the lines of using an SRPP and a tube able to handle higher currents. An ECC99 could be a good choice which,running at 20mA in an SRPP, should be able to put 40mA peak or 28mA rms into 600 ohms which is:
V = IR = 28 * 0.6K = 16.8V rms = +26dBu.
Not surprisingly, the ECC99 in an SRPP topology has been used for headphone amplifiers.
A mu follower gives us a gain close to mu which in the case of the 6CG7 is 20 or 26dB which is just enough with a little margin for the gain make up we need for our EQ. However, an SRPP has a gain close to half mu, which in the case of the ECC99 would be 22/2 = 11 = 21dB. This is barely enough to make up the losses in the EQ circuit. It also produces a lot more distortion at a given output level than a mu follower. So, on balance, the mu follower provides us with everything we want - high headroom, low distortion, good gain - but we have to compromise on a 10K load capability.
Of course, you are free to try and ECC99 SRPP if you wish, but you will need four times the HT current and twice the heater current which may make it a little more expensive.
Cheers
Ian
This could be then used as a gain stage in other projects too ( for example console channels trip, no?). Put 8 EQs side by side in one 4u case etc. Separate psu pcb.
