18Watt.com

I know this has been hashed over a lot here,but it needs to be addressed again to keep us current.
Most guys use a 125ohm cathode resistor and some even go to a 130 or larger.Then there is the cathode bypass cap hovering around 50-100uf.Well an AC30 uses 100ohm cathode resistors and a reissue 1974x uses a 91ohm resistor.Then we say to reduce boominess we should use a smaller cathode bypass resistor around 100uf instead of the 500uf original.
There are lots of other EL84 amps out there that also use 100ohm cathode resistors.They have comparable B+ to the 18 watt.I have been using my latest TMB as a test mule for these values and I have come across some interesting results when substituting values.When I use the common 100uf/125ohm pair,the distortion is so-so and the amp is a little flubby when cranked.I went from 125ohm all the way to 200ohm and it got progressively worse sounding.I also went from 50uf all the way up to 1000uf and it got better up to 1000uf.It is much more defined at 1000uf and has tons of focused bottom end.
I have also been experimenting with B+ from 320v to 360v and it sounds better up at 360v.
All of the EL84 amps I have mentioned use screen resistors of 100ohms as well.
When you get to that creamy distortion sound,you know what I mean,it sounds right.
I am sold on the 1000uf cathode resistor for my amp right now and when I get the parts I will be upping that in stages and also lowering the resistor to 100ohms to see what happens.

Coo-er. What was your plate dissipation at 360V with a 100r cathode resistor?

Time for a short theory lesson. Here's the thing with cathode biased amps, especially Class AB ones. You get a lot of fluctuations in the bias voltage across the cathode resistor, as the amp is played, especially on bass notes. At the peaks of these fluctuations, the amp can be temporarily biased very cold and suffer from crossover distortion. This can make the tone of a cathode-biased amp sound somewhat ratty.

The colder the amp is biased at idle (larger cathode resistor), the more likely it is for the tubes to get too cold on loud transients and suffer crossover distortion. Also running at higher voltage means you have to bias colder not to burn the tubes out.

Some cures for this issue are:
- Negative feedback
- Bias warmer (lower cathode cap) - risks burning out tubes
- Stabilize bias voltage

The warmer biasing approach used in vintage Marshall 18W and many Vox amps is an interesting one. The amplifier is idle-biased to say 80% or 90% of max dissipation, unlike a fixed biased amp that would be biased between 50% to 75% of max dissipation. Normally in a fixed biased amp, 80% to 90% would result in the tubes running too hot when cranked. However, with cathode-biased amps, the surges in bias voltage across the cathode temporarily bias the tube colder while cranking, and help keep things under a degree of control.

The cathode voltage can also be stabilized by means of a very large cathode cap and/or zeners. Our man at Marshall, who sometimes posts on this board under the name "english jim" originally pointed this out to me, that a very large cathode cap of 1000uF to 2000uF can really help smooth the sound of a cathode-biased amp. A cap this big charges and discharges very slowly and therefore the cathode voltage doesn't fluctuate with brief transients - however, it will gradually rise in fall with how hard the amp is being played and help provide some compression. You can then also idle bias more safely into the 75% region, and still have the amp sound good. Overall the amp will have more punch and a solid low end, somewhat more like a fixed bias amp - not farty or boomy as many would expect.

So these days I tend use 180 ohm or 220 ohm cathode resistors with 1000uF caps. I also always use the Lite IIb arrangement of a single shared 1k screen resistor, with individual 100 ohms at each tube socket. I'm not taking any risks with my NOS tubes.

The cathode voltage can also be stabilized by means of a very large cathode cap......

Isn't the fluctuation in the cathode voltage, one of the properties that add to the "18 watt sound"? Won't to large of a cathode cap make the amp sound too stiff?

Thankyou,
Keith

No, and for two reasons. First you will still get nice gentle fluctuations of the cathode voltage which will provide some compression. Secondly, you will still get rectifier sag, which IMO is more significant anyway. Sustained notes will still be able to "bloom". You will just lose some buzziness, and gain more solid low end response.

In reference to the cap, would that same 1000uf value work as well for a 6v6 setup?

Keith

Absolutely. In fact I first tried this with 6V6s, before using it with EL84s.

Two questions:

What voltage rating do we need for the cathode caps? I measured about 10.5V on my power tube cathodes. I'm guessing 50V should be plenty. Can we get away with something around 25V?

Any recommended cap brand and places to buy for this size of cap?

Thanks

Mark

Great explanation ZP, we neophytes bow to you!

zaphod_phil wrote:Time for a short theory lesson...

Phil, if you could ever find the time, I believe you could write one heck of a book on ampbuilding -- and I'd buy two! Thanks for posting such well-thought "lessons" for us here

jimipage wrote:

zaphod_phil wrote:Time for a short theory lesson...

Phil, if you could ever find the time, I believe you could write one heck of a book on ampbuilding -- and I'd buy two! Thanks for posting such well-thought "lessons" for us here

I second that

Ben

zaphod_phil wrote:
So these days I tend use 180 ohm or 220 ohm cathode resistors with 1000uF caps. I also always use the Lite IIb arrangement of a single shared 1k screen resistor, with individual 100 ohms at each tube socket. I'm not taking any risks with my NOS tubes.

Phil,
What does this look like? Could you show us a picture or a layout diagram?
Thanks,
dustin

Can you post a Schematic for your layout? DR. P? Please?

zaphod_phil wrote:Time for a short theory lesson. Here's the thing with cathode biased amps, especially Class AB ones. You get a lot of fluctuations in the bias voltage across the cathode resistor, as the amp is played, especially on bass notes. At the peaks of these fluctuations, the amp can be temporarily biased very cold and suffer from crossover distortion. This can make the tone of a cathode-biased amp sound somewhat ratty.

The colder the amp is biased at idle (larger cathode resistor), the more likely it is for the tubes to get too cold on loud transients and suffer crossover distortion. Also running at higher voltage means you have to bias colder not to burn the tubes out.

Some cures for this issue are:
- Negative feedback
- Bias warmer (lower cathode cap) - risks burning out tubes
- Stabilize bias voltage

The warmer biasing approach used in vintage Marshall 18W and many Vox amps is an interesting one. The amplifier is idle-biased to say 80% or 90% of max dissipation, unlike a fixed biased amp that would be biased between 50% to 75% of max dissipation. Normally in a fixed biased amp, 80% to 90% would result in the tubes running too hot when cranked. However, with cathode-biased amps, the surges in bias voltage across the cathode temporarily bias the tube colder while cranking, and help keep things under a degree of control.

The cathode voltage can also be stabilized by means of a very large cathode cap and/or zeners. Our man at Marshall, who sometimes posts on this board under the name "english jim" originally pointed this out to me, that a very large cathode cap of 1000uF to 2000uF can really help smooth the sound of a cathode-biased amp. A cap this big charges and discharges very slowly and therefore the cathode voltage doesn't fluctuate with brief transients - however, it will gradually rise in fall with how hard the amp is being played and help provide some compression. You can then also idle bias more safely into the 75% region, and still have the amp sound good. Overall the amp will have more punch and a solid low end, somewhat more like a fixed bias amp - not farty or boomy as many would expect.

So these days I tend use 180 ohm or 220 ohm cathode resistors with 1000uF caps. I also always use the Lite IIb arrangement of a single shared 1k screen resistor, with individual 100 ohms at each tube socket. I'm not taking any risks with my NOS tubes.

Spankenstrat wrote:Can you post a Schematic for your layout? DR. P? Please?

What layout??? You want a schematic for a 1000uF cap? Or if you're asking about the Lite IIb's screen grid arrangement, it's right there in the Downloads section - http://www.18watt.com/storage/18wattLite2b.gif

I wanted to see the grid. To be honest, I don't really understand which parts of the circuit does what. I'll look for a circuit explanation.
Thanks,
dustin

I will try the 1000 mF cap, thanks for the info!

Nooooooo not for the 1000uF cap, you silly goose...
The link is perfect. Thanks Dr. Phil.

zaphod_phil wrote:

Spankenstrat wrote:Can you post a Schematic for your layout? DR. P? Please?

What layout??? You want a schematic for a 1000uF cap? Or if you're asking about the Lite IIb's screen grid arrangement, it's right there in the Downloads section - http://www.18watt.com/storage/18wattLite2b.gif