How to add gain adjustment of first triode

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
ruffrecords said:
I have not checked that but there will be some increase in noise. If you see my other post you will see the 10K pot is already causing the CF problems so adding a low value pot divider after the CF output cap will only make that worse.

Yep I got it, thanks Ian. 
So maybe, 1 X 6072 split into the two gain stages, and 2 X 12BH7 Whites? 
 
Just checked the dissipation in the CF triode is under 300mW. We could double its quiescent current for another 6dB of headroom.

Cheers

Ian
 
ruffrecords said:
Just checked the dissipation in the CF triode is under 300mW. We could double its quiescent current for another 6dB of headroom.

And that might just do it.
If not, re-jig the 6072's to eliminate that 2nd follower and utilise it with another half for a White.  Lots of ways we can skin this here cat  ;D
 
Winston O'Boogie said:
And that might just do it.
If not, re-jig the 6072's to eliminate that 2nd follower and utilise it with another half for a White.  Lots of ways we can skin this here cat  ;D

Just bumped the current up to nearly 3mA (400mW dissipation in CF triode) Not we can get +20dBu into 10K with 0.1% distortion at 1KHz.

Gone midnight now. Off to bed.

Cheers

Ian
 
BluegrassDan said:
Okay. My next course of action is to do as you suggest. 1/2 6072 ---> 12BH7 WCF ----> 1/2 6072 ----> 12BH7 WCF

If it's easy cool. If scoping the circuit is easier, do that.
I still think Ian's initial suggestion regarding the 1st stage cathode value/biasing might be worth looking at.  How does that stage clip, etc.?  symetrically or? 
I.O.W.  Find the weak link, the stage that craps out first.  I don't doubt Ian's info that the cathode follower is under biased but, is that all?
Just keep at it until you're happy dude

Easy test: with your rebiased follower, how is it with that 1st gain stage cathode cap removed as well?
 
Morning all.

In my sim the first stage looks OK. Plate is around 160V, current 0.5mA and the plate will swing easy past +30dBu. The same is true of the unloaded CF, but as soon as you load it with 10K it starts to clip much earlier.

The way to change the CF bias is to change the 2K2 resistor. I changed it to 1K5 and got 2 point something mA. I also changed the 56K to 25K in series with  3K and used that junction for the NFB.

Simple rule for CF bias. 10V rms into 10K is 1mA rms current. This means you need a peak to peak current of  2.828mA. Since the CF can swing down  from its quiescent to 0mA and up as much as you like, the largest peak to peak current it will supply without clipping is 2 x quiescent. So you need a minimum quiescent current of about 1.5mA. In practice you need somewhat more - 2.5mA would be better which is about what you get with a 1K5 bias resistor.

Lastly, about noise and the feedback resistor. The 150 ohm source reflected to the secondary of the mic transformer is 15K. The secondary dc resistance is typically about 2K5. The toral noise resistance at the grid is therefore normally 17K5. With the NFB (25K and 3K) the total becomes 20K5. This will worsen the noise by less than 1dB.

Cheers

ian
 
Morning Ian,
sounds good, Dan should be pleased. 
One thing that doesn't compute for me is your noise analysis but I think you just used a 1:10 transformer rather than Dan's1:5
Still, with a 3k lower FB resistor it won't be a terrible difference.
 
Winston O'Boogie said:
Morning Ian,
sounds good, Dan should be pleased. 
One thing that doesn't compute for me is your noise analysis but I think you just used a 1:10 transformer rather than Dan's1:5
Still, with a 3k lower FB resistor it won't be a terrible difference.

Ah, I forgot he changed to 1:5. So 150 is reflected as 3750 ohms. Assume secondary has half the dc resistance because it has half the turns so go for 1250. Total is 5000. Adding 3000 ohms makes it 8K  which means noise will be 2.04dB worse.

I think the real problem is the ability to drive the 10K pot. The first stage on its own seems fine. Increasing the current of the CF will allow it to drive 10K  to over +20dBu.

At the moment the overall gain from mic to pot is 14dB (transformer) plus 29dB (tube) = 43dB. This means the maximum input before hitting the CF ceiling is -23dBu. An input pad would make this -3dBu which is why it works so well. To achieve the same effect without a pad means reducing the gain by 20dB somehow. This means we really need to be able to drop the input amp gain to 9dB if we can. We can get close to this with the NFB method by making the pot divider 18K and 10K.

The noise will be 10log(1500/5000) = 4.77dB worse but the signal will be 20dB higher so the S/N will improve by  15dB. Or put another way, the noise at the input is up by 5dB but the gain is down by 20dB so the noise at the output is down by 15dB.

We could probably have a pot to change the gain over this range.

Cheers

Ian
 
In the most basic sense, changing the CF bias resistor to 1.5k will clear up all of the mess? The NF is optional?

I assume I should do the same thing for the second gain stage CF as well? It is identical to the first, although feeding the White follower.
 
BluegrassDan said:
In the most basic sense, changing the CF bias resistor to 1.5k will clear up all of the mess? The NF is optional?

I assume I should do the same thing for the second gain stage CF as well? It is identical to the first, although feeding the White follower.

Hey Dan just try it.  I apologise  for sending you off on a wrong turn last night with only changing the 56K resistor, I had a slightly different schematic in my head (straight, directly coupled follower) and didn't refer back to your actual scheme.  My bad.
 
BluegrassDan said:
In the most basic sense, changing the CF bias resistor to 1.5k will clear up all of the mess? The NF is optional?

I assume I should do the same thing for the second gain stage CF as well? It is identical to the first, although feeding the White follower.

Yes.  Just change the 2K2 to 1K5 and the 56K to 27K and give it a go. The second CF does not need changing because it is not driving 10K but a much higher impedance.

Cheers

Ian
 

Latest posts

Back
Top