Request for design guidance with Fixed Bias Output Tube
#16 Re: Request for design guidance with Fixed Bias Output Tube
For clarity I always (to date) set operating points so that I'm well into A1 operation with plenty of headroom. For example my input valve is the 26, biased at roughly 11v. Balanced output from my DAC is 4.4vrms, so 6.2vpk, into an input transformer which puts out SE to the 26 grid, which should swing from 7.9v to 14.1v. Thinking about that I have no grid resistor on the 26, only the transformer output which gives any stray grid current a path to ground.
Acknowledged that grid current won't be zero, but it should be negligible ? how would one go about measuring grid current ?
Acknowledged that grid current won't be zero, but it should be negligible ? how would one go about measuring grid current ?
#17 Re: Request for design guidance with Fixed Bias Output Tube
Just look at the circuit for a original Ongaku or some of Andy Groves designs for WAD
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#18 Re: Request for design guidance with Fixed Bias Output Tube
And it does that by reducing the impedance of the driving stage so grid current can go somewhere.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#19 Re: Request for design guidance with Fixed Bias Output Tube
Doesn't take much current to generate a couple of volts across 100k.Acknowledged that grid current won't be zero, but it should be negligible
The manufactures numbers are based on what will keep an old valve stable, not what is needed for a new and shiny one. Grid current tends to increase with use.
I can think of two different amps that I have probably repaired 80% of the ones in the UK and in both cases the makers didn't ready (or ignored) the minimum grid resistance value.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#20 Re: Request for design guidance with Fixed Bias Output Tube
That's fascinating - as it happens in going down from a PP design to SE I have twice the number of valve bases in the wood that I need. I'd considered implementing cathode followers but had decided to try the fixed bias option first.
However what you've opened my eyes to there is the use of balanced rails driving cathode followers which also set the bias for the following valve. Genius - and thanks for that. Actually a little more complex than that, it's combination bias.
Much thinking to be done.
#21 Re: Request for design guidance with Fixed Bias Output Tube
I'll admit to being confused by that ECC82 though...
The plates are at +195v. The cathodes are at -35v. The grids are at -40v.
Vak is 230v. Vgk is -5v. 33K cathode resistors with 160V across them means just under 5mA through each triode section.
I can't see how to get to 5mA from the 12AU7 plate curves with those values, could be because I haven't designed a cathode follower before. Good time to find out.
The plates are at +195v. The cathodes are at -35v. The grids are at -40v.
Vak is 230v. Vgk is -5v. 33K cathode resistors with 160V across them means just under 5mA through each triode section.
I can't see how to get to 5mA from the 12AU7 plate curves with those values, could be because I haven't designed a cathode follower before. Good time to find out.
- izzy wizzy
- Old Hand
- Posts: 1496
- Joined: Fri Nov 02, 2007 7:02 pm
- Location: Auckland NZ
- Contact:
#22 Re: Request for design guidance with Fixed Bias Output Tube
The cathode resistors will set the current more or less and the cathode/grid will find it's own voltage in the end. Maybe the Vgk is a bit off from the datasheet. Could depend on when the measurements were taken i.e maybe not exactly at the same moment.
A 10V error in that measurement only varies the current by 300uA.
A 10V error in that measurement only varies the current by 300uA.
- Paul Barker
- Social Sevices have been notified
- Posts: 8863
- Joined: Mon May 21, 2007 9:42 pm
#23 Re: Request for design guidance with Fixed Bias Output Tube
Yes nothing to worry about, build that stage alone and tweak it to get you’re desired 300b bias.
300b does conduct grid current from -1v but to an ECC82 at 5mA easy.
5687 as used in Ongaku better where grid current more in the choice of output valve. But also a preferred sound over the ECC82 to most people whose opinion is proved worthwhile.
300b does conduct grid current from -1v but to an ECC82 at 5mA easy.
5687 as used in Ongaku better where grid current more in the choice of output valve. But also a preferred sound over the ECC82 to most people whose opinion is proved worthwhile.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
- izzy wizzy
- Old Hand
- Posts: 1496
- Joined: Fri Nov 02, 2007 7:02 pm
- Location: Auckland NZ
- Contact:
#24 Re: Request for design guidance with Fixed Bias Output Tube
+1 for the 5687 over ECC82Paul Barker wrote: ↑Wed Jun 03, 2020 5:37 pm 5687 as used in Ongaku better where grid current more in the choice of output valve. But also a preferred sound over the ECC82 to most people whose opinion is proved worthwhile.
If you want ECC82 to try though, I got some in a box somewhere you can have.
#25 Re: Request for design guidance with Fixed Bias Output Tube
Very generous of you thanks. I have a cupboard full of valves of various types in all sorts of boxes. I'm trying to avoid having to catalogue themizzy wizzy wrote: ↑Wed Jun 03, 2020 6:30 pm+1 for the 5687 over ECC82Paul Barker wrote: ↑Wed Jun 03, 2020 5:37 pm 5687 as used in Ongaku better where grid current more in the choice of output valve. But also a preferred sound over the ECC82 to most people whose opinion is proved worthwhile.
If you want ECC82 to try though, I got some in a box somewhere you can have.
#26 Re: Request for design guidance with Fixed Bias Output Tube
Would it be a stretch to consider controlling the 300B bias by varying the fixed bias on the cathode follower valve?
#27 Re: Request for design guidance with Fixed Bias Output Tube
No it is a obvious extension. It avoids the problem of meeting the minimum grid resistance requirement. For a ecc82 (for example) for cathode bias, the max grid resistance is 1M.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
- Paul Barker
- Social Sevices have been notified
- Posts: 8863
- Joined: Mon May 21, 2007 9:42 pm
#28 Re: Request for design guidance with Fixed Bias Output Tube
Did anyone read the article on the Andy Grove parallel 300b? First stage; I can’t figure why he’s shunting the cathode resistor and cap instead of bypassing.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
#29 Re: Request for design guidance with Fixed Bias Output Tube
I don't understand what you mean?shunting the cathode resistor and cap instead of bypassing.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#30 Re: Request for design guidance with Fixed Bias Output Tube
So - here's my current thinking. As always very happy to be guided ...
I have an unused secondary winding on the mains transformer which is 150-0-150 @ 20mA which seems to me ripe for providing bias +/- rails. Solid state rectification and filtered not regulated (as the valve B+ supplies aren't regulated). L&R Channels are fully independent. With LCRC filtering and a steady draw current I'm expecting to have say +100 / -100 available on the bias rails.
The amp is SE with balanced inputs : Input transformer to 26, LC coupled to 45, LC coupled to 300B, OPT to speakers (Lowther DX4 in parallel with powered 15" subs).
Tentative plan is to switch to fixed (adjustable) bias : 300B and 45, possibly the 26 too. Cathodes would be grounded. Something like 4 to 8 mA per valve would stay within reason for the transformer (builder confirmed higher mA not a problem but the voltage will drop a little). 5687 doesn't look comfortable at this low current level - either choose a different valve or use a different mains TX.
Along with choice of valve I need to choose cathode follower load to a) be within the max grid resistor for the following valve and b) allow the bias voltage to be varied within the range desired by adjusting the bias on the cathode follower. As the CF bias is varied, the CF current varies and so does the 300B bias.
Just about choosing the valve and load now, right ?
I have an unused secondary winding on the mains transformer which is 150-0-150 @ 20mA which seems to me ripe for providing bias +/- rails. Solid state rectification and filtered not regulated (as the valve B+ supplies aren't regulated). L&R Channels are fully independent. With LCRC filtering and a steady draw current I'm expecting to have say +100 / -100 available on the bias rails.
The amp is SE with balanced inputs : Input transformer to 26, LC coupled to 45, LC coupled to 300B, OPT to speakers (Lowther DX4 in parallel with powered 15" subs).
Tentative plan is to switch to fixed (adjustable) bias : 300B and 45, possibly the 26 too. Cathodes would be grounded. Something like 4 to 8 mA per valve would stay within reason for the transformer (builder confirmed higher mA not a problem but the voltage will drop a little). 5687 doesn't look comfortable at this low current level - either choose a different valve or use a different mains TX.
Along with choice of valve I need to choose cathode follower load to a) be within the max grid resistor for the following valve and b) allow the bias voltage to be varied within the range desired by adjusting the bias on the cathode follower. As the CF bias is varied, the CF current varies and so does the 300B bias.
Just about choosing the valve and load now, right ?