Understanding the cathode follower (D-ulator)

[Bob-I]

I haven't been totally satisfied with my built in Dumbleator amp so I checked out the signal flowing through it. Turns out the CF in the D-ulator is clipping the bottom half of the wave form. This happens with both a 12AX7 and 12AT7.

I have a 1M grid, 1.5K to cathode and 10K to ground. I believe the tube is hitting cutoff so I'll need to adjust some of the values. I believe that raising the cathode and ground resistors are the way to go, and a search confirmed that some folks use 1.8K and 27K.

Am I on the right track?

Also, I'm using a single node on the PS. Is this a problem?

[heisthl]

Try and get the cathode follower plate above 260 (300 is ideal) and the recovery plate around 200 or more. It shouldn't have distorted with your old values but the new values are much better for the life of a 12AX7 tube. It's better to have 2 nodes or at least another dropping resistor feeding the cathode follower side at the end of the string so you can get these voltages in the ballpark.

[dogears]

Bob,

Use the higher values. I think you will actually hear some improved clarity.

Let's get together sometime. K?

[Bob-I]

Bob,

Use the higher values. I think you will actually hear some improved clarity.

I'll do that, but beyond blindly putting known good values I'm trying to learn from this.

Let's get together sometime. K?

Sure. Pretty busy the next 2 weekends, let's talk after.

[ayan]

Try and get the cathode follower plate above 260 (300 is ideal) and the recovery plate around 200 or more. It shouldn't have distorted with your old values but the new values are much better for the life of a 12AX7 tube. It's better to have 2 nodes or at least another dropping resistor feeding the cathode follower side at the end of the string so you can get these voltages in the ballpark.

I agree with Henry, top to bottom.

I wonder if you made sure the CF clipped only at the output, or whether you were clipping the input as well? There is a lot more signal than needed from the ODS preamp output, so if you don't turn the amp's mater volume down, you will clip the living daylights out of the input of the CF.

If you will be doing some surgery in your Dumbleator, you might as well replace the first dropping resistor with a choke. You will gain some voltage headroom and decrease the noise floor of the unit.

Gil

[Bob-I]

I wonder if you made sure the CF clipped only at the output, or whether you were clipping the input as well? There is a lot more signal than needed from the ODS preamp output, so if you don't turn the amp's mater volume down, you will clip the living daylights out of the input of the CF.

I adjusted the signal to get a clean sine wave at the input. The signal level drops dramatically and clips.

Something's not right, but honestly I'm a bit discusted with this amp. I've been trying to get it to sound good for 3 years. I'm ready to get a new chassis and build a simple ODS with the parts

[Luthierwnc]

Norm, Scott and I spent some time on this a while back. There are some schematics and resistor values designed for in-amp loops. Here's the link:

https://tubeamparchive.com/viewtopic.ph ... 1&start=15

Lots o info here. Skip

[Bob-I]

Damn, stoopid stoopid stoopid.

I had a 1.5R not a 1.5K on the return side cathode. It was starving the CF because of the current draw and drop across the PS rail.

It's much better now, but still not perfect but I have Skip's thread to go on now. Thx.

[butwhatif]

I found this problem, or, artifact with my latest mod job with built in dlator.
The amp sounds great, but it does clip the cf stage at hi clean and dirty master settings. Hard to know this w/o a scope. Clips one side well before the other hits. Didn't get the chance to shake it down yet, but I'm guessing higher b+ and/or bias might work. Anyone else witness this?

[Bob-I]

I found this problem, or, artifact with my latest mod job with built in dlator.
The amp sounds great, but it does clip the cf stage at hi clean and dirty master settings. Hard to know this w/o a scope. Clips one side well before the other hits. Didn't get the chance to shake it down yet, but I'm guessing higher b+ and/or bias might work. Anyone else witness this?

I don't think you can 100% sure of this without a scope. I was unable to determine where my problem was until I found it with the scope. As it turns out, the clean sides (there's 2 on this amp) sounded great but the OD was terrible. Once I fixed my cathode issue, all channels sound much better.

[butwhatif]

You wouldn't know how unevenly it distorts w/o a scope. I first noticed it on the bench in od mode, but clean clips one side quite a bit sooner also at hi settings, next time I get it on the bench I'll take a scope picture

[glasman]

I ran into the same problem a couple of years ago.

My original fix was to increase the CF rail voltage to about 290 volts. This raised the cathode voltage from about 11 volts to 14 volts. This helps, but I still feel that the higher cathode and tail resistor combination helps even more.

My amps now run between 28 and 32 voltages on the cathode and there is no hint of CF clipping with all controls maxed.

This problem shows up more on lower power amps than say 100 watters, depending on how high you crank the controls. Me, I am cranker, so I found the problem pretty fast .

Gary

[glasman]

FWIW, this is the same type of problem that some folks see with the BM power amp and low power amps. Lack of headroom and clipping from the PI (which sounds like A$$).

[Bob-I]

I ran into the same problem a couple of years ago.

My original fix was to increase the CF rail voltage to about 290 volts. This raised the cathode voltage from about 11 volts to 14 volts. This helps, but I still feel that the higher cathode and tail resistor combination helps even more.

My amps now run between 28 and 32 voltages on the cathode and there is no hint of CF clipping with all controls maxed.

I have the rail at 330V, can't remember the cathode but it's on the high side. With the 12AX7 the return volume is almost off for unity through the loop. I haven't tried the 12AT7 yet.

I'll have to adjust the rail now that I found that screwup. Thing is, I measure every resistor after being burnt by my colorblindness too many times. I guess I saw 1.5 and assumed the K.

[heisthl]

If you using 250KA pots unity is approx. noon and noon on the knobs. Like Gary said the cathode should be at 30 volts with the new values