6aq5 Screen resistor?

[ThSpeakerDude88]

I finally have my modified harmony up and running. I found a mistake in the original circuit too, pin 6 of the 12au6 had its 2.1meg resistor going to ground instead of B+.

Anyway, I am having some trouble with the output stage. The 6aq5 is running on 180v, into a 5k transformer. I have tried different values of resistors , but my undertsanding is that you can just tie the screen to the power supply side of the OPT. When I do this though, my plate voltage is at 180v, while my screen voltage is 183v. Even with a 2.2k resistor and 47uf cap, it still is too high. I thought it was bad to have your screen voltage higher than your plate? The amp sounds and runs fine otherwise. My bias resistor is currently fixed untill I can get a pot, it is fixed @ 470 ohms and bypassed with a 22uf cap. Cathode voltage is 11v, which is also close to datasheet spec.

[Miles Prower]

Originally posted by ThSpeakerDude88
Anyway, I am having some trouble with the output stage. The 6aq5 is running on 180v, into a 5k transformer.

That's a standard Q-Point.

I have tried different values of resistors , but my undertsanding is that you can just tie the screen to the power supply side of the OPT.

Yes.

When I do this though, my plate voltage is at 180v, while my screen voltage is 183v. Even with a 2.2k resistor and 47uf cap, it still is too high. I thought it was bad to have your screen voltage higher than your plate?

Not necessarily. It's not often done for audio apps, but it is done otherwise. (Saw a video amp design in an old TV that had 18Vdc on the plate and 120Vdc on the screen. Not so sure how linear that was, but the time domain response is more important for video anyway.) So long as you don't bust the Psg ratings, and keep the loadline out of the tetrode kink area, it's OK. Besides, 183 volts v. 180V isn't going to make any difference. If you need a resistor, add it as a screen stopper only.

The amp sounds and runs fine otherwise. My bias resistor is currently fixed untill I can get a pot, it is fixed @ 470 ohms and bypassed with a 22uf cap. Cathode voltage is 11v, which is also close to datasheet spec.

Then there's nothing here to worry about.

[ThSpeakerDude88]

Hey thanks for the info!

I just thought that I read somewhere that if the grid was higher it would cause it to draw excessive current.

However, my power supply/filter section is a bit primitive. It only uses a 47uf cap /2.2k /47uf cap. I am going to further improve upon it by adding another filter section. If the hum does not go away then then I will be back...


Thanks again!

~tsd88

[ThSpeakerDude88]

Eh, another question I had:

My other amp, single ended EL84 with triode/ pentode selection, is quite a bit cleaner at higher volumes than this amp. This is when run in triode mode, which limits the EL84 to about 1.9 watts @ 9% thd.

My 6aq5 is stated on the datasheet as having an output of 2 watts @ 8% thd on my given voltages/bias. So, it should be about the same volume shouldn't it? Both tests were done into the same speaker in the same cab.

Here is my current modified schematic

[Miles Prower]

Originally posted by ThSpeakerDude88
Eh, another question I had:

My other amp, single ended EL84 with triode/ pentode selection, is quite a bit cleaner at higher volumes than this amp. This is when run in triode mode, which limits the EL84 to about 1.9 watts @ 9% thd.

My 6aq5 is stated on the datasheet as having an output of 2 watts @ 8% thd on my given voltages/bias. So, it should be about the same volume shouldn't it? Both tests were done into the same speaker in the same cab.

Here is my current modified schematic

1.9W vs. 2.0W is not audible at all. Cleanliness at high volumes is determined by clipping behaviour, amoung other things. Since this 6AQ5 amp is running with no feedback at all, you have lots of pentode nastiness in the form of higher order harmonics. Also, the damping factor for the woofer is going to be considerably less as a result of the higher r(p) of the pentode. The regulation of that PS isn't all that great, you have cathode bias, which doesn't perform as well as fixed bias during overdrive and clipping. Combine PS sag, bias that shifts toward a more nonlinear part of the characteristic, raw pentode nastiness, and woofer resonance, and it won't sound as clean during clipping as would a triode.

[ThSpeakerDude88]

well, its not during clipping that im worried about, it has its own tone there. Its the amount of clean headroom I have. Both amps do not use nfb, and both are cathode biased. Both, regardless of triode/pentode mode, run into a 5k transformer. This might explain why my triode mode does have more clean headroom though, as a higher plate load ( 5k is normal for an el84, but in triode mode it is cut to about 3k) might add to the more hi-fi cleanlieness of the signal. I could try increasing my load on the output , however I still have to be able to run the 6aq5 triode strapped as well, and it already is working into a load almost twice what is reccomended for triode strapped.

How could I improve upon my current bias setup? I mean, its single ended and class A, and about the simplest you can get. Is it really worth trying to use a fixed voltage bias?

[Miles Prower]

Originally posted by ThSpeakerDude88
well, its not during clipping that im worried about, it has its own tone there.[

Actually, you do have to worry about it. Most of the time, music plays at some low average level. However, it also has the occasional transient that rises well above the average (sometimes as much as 20db(v) above average) and that will cause clipping. That's why SS rigs are often built with such huge power ratings: they need to avoid clipping completely since transistors have a very nasty clip behaviour (good for digital, bad for analog) that's made worse by the large amounts of feedback needed to linearize those three terminal fuses.

Its the amount of clean headroom I have. Both amps do not use nfb, and both are cathode biased. Both, regardless of triode/pentode mode, run into a 5k transformer. This might explain why my triode mode does have more clean headroom though, as a higher plate load ( 5k is normal for an el84, but in triode mode it is cut to about 3k) might add to the more hi-fi cleanlieness of the signal. I could try increasing my load on the output , however I still have to be able to run the 6aq5 triode strapped as well, and it already is working into a load almost twice what is reccomended for triode strapped.

How could I improve upon my current bias setup? I mean, its single ended and class A, and about the simplest you can get. Is it really worth trying to use a fixed voltage bias?

About the only thing you could try is a variation on the design of Sy's "Red Light District" amp, and substitute LEDs for the cathode bias resistor. This will give fixed bias and reduce the AC impedance in the cathode return path. This might help improve the clipping behaviour. Since this is such a simple rig, as you pointed out, making elaborate mods wouldn't make much sense.

[ThSpeakerDude88]

umm...... Sorry I didn't make it clear to begin with, although I would have thought the schem would...


this is a guitar amp, so I want distortion. Thats why I'm not worried about its distortion charactoristics because it has its own 'tone' there. I was just wondering why this 2 watts is not as clean as my El84's 2 watts. I did notice that when my EL84 is in pentode mode, it breaks up with less comparative headroom than in triode mode, while in triode mode it stays comparativly clean untill it reaches its maximum output. Since the 6aq5 is putting out 2 watts, only in pentode mode, does this make sence of my earlier breakup than expected?

[Miles Prower]

Originally posted by ThSpeakerDude88
I did notice that when my EL84 is in pentode mode, it breaks up with less comparative headroom than in triode mode, while in triode mode it stays comparativly clean untill it reaches its maximum output. Since the 6aq5 is putting out 2 watts, only in pentode mode, does this make sence of my earlier breakup than expected?

Yes it would. Pentodes, running as pentodes, tend to have a nastier overdrive response than do triodes.

[ThSpeakerDude88]

Thanks!

Here's my completed schematic.

I added an extra filter section, and filiment reference , which helped a lot. Currently. there is no hum when in pentode mode, you can't tell the amp is on. However, as with my other amp, only a little more noticable, when triode connected as is, there is a hum that is a little noticable. Is there a way to get rid of this?