diyAudio (
-   Tubes / Valves (
-   -   Driving 6AQ5's (

jerluwoo 19th June 2007 07:14 PM

Driving 6AQ5's
2 Attachment(s)
So I got my hands on a pair of nos Sylvania 6aq5's. I have this simple paraphase circuit I want to try. Question is will I get enough drive from my 6j6 the way I have it?

PRNDL 20th June 2007 12:25 AM

I believe 500 mV is Mic/Instrument level, which means you'd need to add a preamp gain stage before the phase invertor.

(I'm also wondering about some of the values in the circuit ... C1=10uF is quite large, R3=100 would make the input resistance fairly low).

jerluwoo 20th June 2007 02:49 AM

The 6j6 with these values put it at the recommended class a1 operating point from the data sheet. I'm using the 10uf because I have a few good quality bipolars to use.

SY 20th June 2007 03:43 AM

With pentode outputs, you'll have a hellacious output impedance without using feedback. You really need more open-loop gain and a feedback arrangement of some sort that includes the output stage.

PRNDL 20th June 2007 01:09 PM

I ended up treating the 6AQ5 like a low-power 6V6, which is what they say it is in the data sheet.

jerluwoo 20th June 2007 06:31 PM

2 Attachment(s)
I see what you mean SY. Would be very sloppy sounding bass. Perhaps this schema I found will work. No feedback in this one either. Any tips to improve it?

Johan Potgieter 21st June 2007 12:07 AM

If I may comment,

Oops, no!

Firstly C3,C4 should be some 10x higher, otherwise poor bass! Then the first triode (6AF4) has a very low amplification - about 9, not sure why something like 1/2 12AT7 was not used (are you going to do stereo?)

Then using a 300-0-300V power transformer is going to leave you with quite about 390V dc, or some 360V at the 300V point. For the quoted dc voltages 250-0-250Vac would be a better choice. Lastly the 6J6 anode resistors are quite low - not primarily a problem, but R5 will need to be a little higher than 10K; better to put a 5K trimpot in series for signal balance or adjust and replace by a fixed value.

One could suggest other improvements like using an UL output transformer, but I am suggesting only necessary changes in order to get things near right, for now.


kevinkr 21st June 2007 02:46 AM

First tube is a 6AB4 which has a mu of about 60, overall it strikes me there might be enough loop gain available for some global feedback to be applied. Take it from the 16 ohm tap and apply to the cathode of the 6ab4, note that some compensation might be required to prevent oscillation. (first stage plate) I'd measure the open loop gain and apply just enough feedback to reduce it by 6 - 10dB.

I'm with Johan on c3, c4, but "one" ;) wonders if this was not deliberately done in order to roll off the sloppy low bass.

First I'd build it as shown and then I would modify it with the addition of feedback and some of the other suggested modifications.

I have heard more than one similar amplifier and despite the rather high source impedance performance was better than I expected, but it is highly speaker dependent.

jerluwoo 21st June 2007 07:23 AM

I was thinking maybe replace the input triode with another pentode for higher loop gain then working out the feedback. I think that may work a little better. Would be lucky to get below 100hz with those .01 feeding 220k,will have to do better, maybe these .68 450v I have will work good there. This will be a test amp for later build of a stereo amp since the 6aq5's arent real pricy like so many output tubes ive looked at.
Was also wondering since i'm in the middle of winding a transformer for this,would it be alright to just turn all the e laminations around the same way and leave the i's out,i want to experament with single ended version rolled off at 500hz for direct connect to my klipsch k400's without a crossover?

Johan Potgieter 21st June 2007 11:23 PM

Oh dear ...dearodearodear :o

What did I look up - a 6AF4. There I went and advised replacement with a 12AT7 - and the 6AB4 is a 12AT7 (half)! Suppose one could not express greater agreement!

Sorry - should sleep in bed after midnight, not on forum. Thanks Kevin.

But Jerluwoo, I would expect that leaving part of the laminations totally out the way you asked, would more likely bring you in the region of 5 - 10 KHz or higher and not 500Hz (depending on the size, but still). The mu of cores is several thousand, and even a small gap like a few millimeter would slice the inductance by lots. I have never done SE, but have seen some nifty designs using a fet as the "load" (high Z) and feeding via a C into an output transformer, thus no dc. Others could better advise here.

Yes, 6AQ5s are quite nice; only slightly less powerful that 6V6s, and low heater current.


All times are GMT. The time now is 10:15 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2