6AQ5 vs. 6BQ5/EL84

[Malcolm Irving]

Sometimes people find EL84s to be ‘too crunchy’ etc. and prefer 6V6 etc. I have no problem with that of course, but I’d just like to throw in a technical factor which I think is relevant. (All the following figures are just ‘ball park’ – it depends on B+ value etc.)

In rough terms the EL84 is about twice as sensitive as a 6V6 (and 6AQ5). The mutual conductance of an EL84 is around 11 mA/V, while the 6V6 and 6AQ5 are around 4 mA/V. It takes about 15 V swing from grid-limiting to cut-off for an EL84, but about 30 V for 6V6 and 6AQ5.

A typical PI can deliver around 100 V of swing. The upshot of this is that it is ‘too easy’ to massively overdrive EL84s compared to 6V6. Power stage overdrive is generally regarded as good, but my personal opinion is that too much of it is not to most people’s taste.

To test this theory out would require a PI that can deliver a swing of 100 x 30 / 15 = 200 V and then see how those 6V6s sound!

The larger bottle size of a 6V6 is a good thing for heat dissipation and robustness, which are other factors.

[rifraf50]

PI,s

Sometimes people find EL84s to be ‘too crunchy’ etc. and prefer 6V6 etc. I have no problem with that of course, but I’d just like to throw in a technical factor which I think is relevant. (All the following figures are just ‘ball park’ – it depends on B+ value etc.)

In rough terms the EL84 is about twice as sensitive as a 6V6 (and 6AQ5). The mutual conductance of an EL84 is around 11 mA/V, while the 6V6 and 6AQ5 are around 4 mA/V. It takes about -14 V to put an EL84 into cut-off, but about -30 V for 6V6 and 6AQ5.

A typical LTP PI can only deliver around -50 V to -70 V of negative swing on one of its outputs (more on the other output). The upshot of this is that it is ‘too easy’ to massively overdrive EL84s compared to 6V6. Power stage overdrive is generally regarded as good, but my personal opinion is that too much of it is not to most people’s taste.

To test this theory out would require a PI that can deliver a negative swing of -70 x 30 / 14 = -150 V and then see how those 6V6s sound!

The larger bottle size of a 6V6 is a good thing for heat dissipation and robustness, which are other factors.

one must also consider the advantages of perhaps trying a bootstrapped cathodyne PI instead of LTP PI......LTP PI can suffer from imbalance once the dual triode tube has passed its used by date where as cathodyne PI is almost perfect balance because it uses only one triode to phase invert without gain......the gain itself comes from the other half of the dual triode feeding the

I found this style of PI usefull with using 6av6 and 6at6 tubes.......6av6 was the driver and 6at6 was the PI......got some nice drive from this config but not over the top crazy distortion
NOS 6av6 and 6at6 and 6aq6 are 7 pin singe triodes and can be bought for coffee money.....keep the amp as a full 7 pin lineup !

just google bootstrapped cathodyne phase inverter and you will get the picture......here's one with it in it .....not bootstrapped though >>>>>

[Dave H]

To test this theory out would require a PI that can deliver a swing of 100 x 30 / 15 = 200 V and then see how those 6V6s sound!

It would be more 'doable' to attenuate the output of the EL84 amp's PI by a factor of 2 (replace the 220k grid leaks with a 110k:110k divider) and then compare it with 6V6s under heavy overdrive.

EDIT: If simulation is to be believed it removes the crossover distortion.

[Malcolm Irving]

It would be more 'doable' to attenuate the output of the EL84 amp's PI by a factor of 2 (replace the 220k grid leaks with a 110k:110k divider) and then compare it with 6V6s under heavy overdrive.

Yes, excellent point! I have tried something very like that and found it took away the (what to my ears is) 'too much crunch'.

[Dave H]

Perhaps the 'crunch' was crossover distortion I've added a couple of sim plots above.

[Malcolm Irving]

Yes, you get a lot of cross-over distortion with typical cathode-biased push-pull EL84s. Are your sim plots for fixed-bias, though?

[Dave H]

Are your sim plots for fixed-bias, though?

It's cathode biased with two separate 270R cathode resistors driven by an LTP similar to the AC15 i.e. no NFB.

[printer2]

As a gigging amp I would not use 6AQ5's unless you pick up a mess of them cheap. And before that try a design with them first to make sure you like it. 10W is an optimistic output from them, maybe 8W for a pair in P-P. Mind you not too much difference between 8W and 10W soundwise. Mind you I think EL84's are more like 14W tubes. I have run them at 280V, they most likely will take a little more, I think the place you want to take care is the screens, especially in heavy clipping.

I used a pair of 12AB5's and would recommend them over the 6AQ5's if you have 12V available for the heaters. They have the same voltage rating as the 6V6, 315V plate. From memory I have a pair with about 410V at idle on them, the transformer that I have does sag a bit when at full output, think around 360V, might help them survive. I had the pair run into a Fender Special Design 12", think it came out of a Blues Jr or Hot Rod Deluxe. Been away from live music for a while and I forgot how a guitar can sound REALLY LOUD. I first found out they have done high voltage service for Harmond Kardon 12AB5, mind you with hifi and probably biased colder. On the schematic it has a crazy 495V plate and 490V screens. Can believe their 20W if this is true.

harman kardon a120 | Service Manuals

Either way they are a more robust tube than the 6AQ5. For your viewing pleasure.




410V@30mA per tube at idle. At 410V*0.030A = 12.3W. At edge of breakup the voltage drops to 370V, down to 355V with the amp feeding back and gain at a point where more really does not make much difference. Was using a 120V:120/240V 22W Hammond isolation transformer for the PT. 6600 ohm output transformer. 12V transformer primary winding for a choke. The pentode up front is a little too microphonic, will change it to a 12AX7 some day.