Page 3 of 3

6L6s

PostPosted: Wed Dec 29, 2004 8:40 am
by dhuebert
I changed the cathode resistors on the 6L6s to 820R and was able to drop the current to 45mA, is it my imagination or does the amp sound better? Cathode voltage is now ~28 V. Since my last post the unit has about 20 hours on it and seems to be sounding nicer as a result of that as well.

350 V is low for 6L6, however the tube is not designed for ultralinear and I wouldn't want to risk meltdown by running the screen grid too hot. I'm thinking about a tightly regulated supply ~250V for G2, this way I can give the 12AX7s a solid supply and run the 6L6s at 450V without worrying about fireworks. Lighthouse Electric has a switcher @ 250V. Eliminating UL will increase distortion and if I don't like it, can run EL34s, which are designed for UL. Eventually I would like to try KT88s, but I think I will wait for the Eiclone and fixed bias. I like the look of the 6L6s, and KT88s are even more impressive, they give the amp an air of authority that 6BQ5s lack.

I have read a couple of times now of attatching G2 to B+ through a dropping resistor, 1500R seems a common value. Since G2 draws current, this resistor gives it the lower voltage it needs to survive. The drawback is that power supply sag will cause a substantial drop in power with a corresponding increase in distortion. The question is: Has anyone tried using a resistor between the UL connection and G2? Any ideas what effect this might have on power and distortion?

AFN Don[/i]

PostPosted: Wed Dec 29, 2004 10:15 am
by EWBrown
Using a resistor between G2 and the UL tap is a fairly common practice. The usual R value is 1K, but more or less (within reason) should work fine.

As you are running the 6L6s at 350 VDC, I would SWAG that applying more than 250VDC to G2 won't cause any problems, in my ZenClone SV83 SET Amps, the G2 of the SV83s are connected to the plate through a 1K 1W resistor, and I haven't had any problems. There are all sorts of dire warnings about not exceeding the SV83 (6P15P) G2 200V DC voltage, but these plug along at a hefty 2WPC nice and happy. The G2s draw a couple of mA at most, the voltage drop is only a couple of volts.

Let's see, 820 ohms, 45 mA per tube should give about 36.9V at the cathode E=IR. E = .045 X 820 = 36.9V which is about right for the 6L6 family. I have some Russian 6L6GC (6P3C-EV) equivalents that may get connected to my next rev D project.

If you don't mind a couple of extra tubes which glow pretty colors, a VR105 and a VR150 (or their equivalents) in series will give a regulated 255VDC, just feed them off the end of the RC string in the power supply through an appropriate value resistor. Hint: don't put a capacitor across the VR tubes, it will make them behave as relaxation oscillators.

KT88s and 6550s are definitely Eiclone tubage, they would be overkill for a modified ST35.

There are lots of old 6L6 UL and other circuits in this old Acrosound Transformers Catalog PDF file:

http://www.the-planet.org/pdf/acrosound.pdf

/ed B in NH

Bias settings

PostPosted: Mon Jan 03, 2005 8:16 am
by dhuebert
OK, I took some measurements yesterday and found: test points are ~0.45V, cathode resistor ~700 ohms for a cathode voltage of ~31.5V.

Don

Gain Boosting

PostPosted: Fri Jan 07, 2005 8:49 am
by dhuebert
I haven't done anything with my amp for a couple of weeks as I am dismantling my TR6. However, last night I was listening to some Police and thinking about the amp when I realized that the caps I put across R5 and R6 would cancel the NFB from both the output and the phase inverter. I was wondering if you have thought of anything to keep the feedback and boosting the gain. Maybe swapping R30 and R31 with 200K?

Don

Re: Gain Boosting

PostPosted: Fri Jan 07, 2005 12:55 pm
by Shannon Parks
dhuebert wrote:I haven't done anything with my amp for a couple of weeks as I am dismantling my TR6. However, last night I was listening to some Police and thinking about the amp when I realized that the caps I put across R5 and R6 would cancel the NFB from both the output and the phase inverter. I was wondering if you have thought of anything to keep the feedback and boosting the gain. Maybe swapping R30 and R31 with 200K?

Don


Hi Don,

You are correct. As per my earlier post in this thread, the feedback network from V2 (C22 & C23) are accomplishing the same thing as a bypassed cathode. I would remove the bypass caps you put it and leave that spot stock. Certainly don't use both.

If you do choose to remove C22 and C23, you'll need to use a 100 ohm resistor (to ground) and 1200 ohm resistor with the electrolytic strapping the 1200 ohm one (like the Ike). Then inject the NFB at the divider. But this requires a whole re-tweaking of your negative feedback network, too.

What is your source, Don? The amp should have at least 30dB of gain at this point. I would remove the cathode bypass caps AND the negative feedback connection. Listen to the amp like this a few times, then you can add a little NFB with 200K resistors.

Shannon

Boosting feedback

PostPosted: Mon Jan 10, 2005 8:26 am
by dhuebert
I removed the bypass caps from the input section and inserted 150K resistors in series with R30 and R31, this is when I noticed they were only 50K instead of 100K as the schematic called for. I don't recall how this substitution happened... Gain is now right, but it still sounds like crap. I'm wondering now if it has something to do with the leads I had to use to adapt for 6L6s.

Don

Re: Boosting feedback

PostPosted: Mon Jan 10, 2005 2:47 pm
by Shannon Parks
dhuebert wrote:Gain is now right, but it still sounds like crap. I'm wondering now if it has something to do with the leads I had to use to adapt for 6L6s.


Don,

Did you add grid stoppers? Put in 1K grid stoppers right at the grid socket leg. You may have high frequency 'stuff' muddying it up. Seems the 6BQ5s let us get away without them.

Shannon

Grid-stoppers

PostPosted: Tue Jan 11, 2005 8:10 am
by dhuebert
Yes there are 1K grid-stoppers soldered right to the sockets.

Don