EL84 / Champ Design and build

[J M Fahey]

Test #1!

Hello everybody,
First of all, I'd like to thank everyone who has provided advice so far. It has really helped fill in the gaps. Soooo I got the amp soldered up and turned it on, and it worked! Well, kinda. It did amplify sound from the input!

Cool !!! Congratulations !!!!
More than what what many of us achieved on first build

Upon taking some real world measurements, as you guys said, the voltages were low. I updated my schematic below with the voltages I got in bold. The one I really don't understand is the 96v at the plate of the first stage, when the second has 174v? I plan on lowering my dropping resistor setup to around 1500R in order to get 300+ at the first takeoff point.

Cathode and plate current is the same.
In the second stage it roughly matches, difference is attributable to measuring error; but first stage values are impossible as shown.

So probably:

1) stage 1 resistors are wrong value (either 100k is actually larger or 1k5 is lower)
or
2) miswired tone control which is AC isolated (all paths have a series cap somewhere) is actually pulling DC from plate 1, lowering voltages
or
3) somehow power formtube 1 is taken from tube 2 plate so it will work, but we start with much lower voltage than expected
or
4) 1 or more of above

The tone stack did not work as expected. The volume and treble seem to do ok, but the bass doesn't seem to do much at all.

In Fendery/Marshally tone controls, Bass does not do that much, unless Treble is lowered.
To boot in classic Fender it's set too low in frequency so if you are testing with a relatively small, bass shy speaker, it appears to have very little efect.
Set Treble to 0 and move Bass from 0 to 10 ... any change?.

I may follow Tony's advice and omit the tone controls for now. But at the same time I would like to understand why it did not work. Maybe it is because I'm not using the special pots they used in the Champ tone circuit, maybe I just don't have it wired correctly(probably).

No special pots, they are plain vanilla ones.

Also, there is a prevalent 60hz hum that gets louder and quieter with volume adjustment. I think referencing the filament circuit to ground will get this under control, I hope.

Welcome to grounding problems
Present everywhere and hard to explain.
Often you ground here and have lots of hum, move there, 1 inch away and you do not.
Oh well.

I look forward to get everybody's thoughts on the results of my first test run!

[Chuck H]

The triodes have different plate voltages because they have different cathode resistors. The first is 1k5 and the second 2k2.

Is the bass pot wired correctly? You didn't by any chance wire to the cw and ccw terminals but forget to connect the wiper to the cw terminal? That would set it to full bass with no adjustment and amplify any hum picked up by the first stage.

Ah. Ok then. Reado on my part. But the first triode voltage is still too low.

I agree with Juan that, especially in small amps, the bass control can seem anemic. This is exacerbated by the 15k fixed mid resistor and low voltages. Still, since the amp is humming and demonstrating low voltage on the stage the tone stack is supposed to couple, I suspect there is a purely resistive path to ground for the HV through a miswire in the tone stack. Just my first guess.

This should be easy to detect. Just pull the tubes and see if there's still a voltage drop on that stages plate.

[Dave H]

You and g-one are right about the first triode. I missed that. I should have done the calculations.

[rickyl]

Thanks!

Yes, after reading the posts I did a closer examination and found I had inadvertently used the post on the terminal strip that is connected to the chassis, effectively shunting the first stage plate to ground through the 100K resistor in the tone stack.
As far as the PS goes, I got a 1500 ohm 10 watter I'm going to switch into the Pi and see where that gets me.
Thanks again! I'm gonna get the soldering iron warmed up now....

[rickyl]

Test #2!

Ok! Here we go, the second test run.
Attached below is the updated schematic and voltages.
As you can see I lowered the resistance at R1 and re-biased the power tube. I got the tone stack wired correctly and jumpered the bass pot. I'll have to say it sounds pretty decent now, even through my old Magnavox test speaker. The bass control still seems rather anemic though, and I'm getting some motorboating when the volume is turned up 90% or so.
I probably have to get my good speaker hooked up to really start addressing tone.
As far as the motorboating, I was thinking I could increase the value of R3 to try and de-couple the preamp and/or add negative feedback from the OT secondary to the cathode in stage 2.
I look forward to hearing all of your thoughts and opinions! Thanks again!

[Chuck H]

Is the hum gone?
The amp shouldn't motorboat. In fact, adding NFB to an amp can cause motor boating. The only path for positive feedback (what causes motor boating) in your current design would be:

1) Inductive. This is related to lead dress and component proximity.

2) Non schematic coupling. This would be the combining of signals due to small, non ideal resistances like grounding leads and the ESR/impedance of power supply capacitors.

This should be solved for before adding NFB. Adding NFB could complicate the issue by introducing different anomalous affects like phase/gain shifts and additional leads/circuits.

It's still possible there are additional wiring errors. Check to see that all your power supply capacitors are wired correctly and properly grounded and that each tube is indeed supplied from the correct node.

Not sure what you mean by "jumpered the bass pot". Jumpering the bass pot should short all the bass through the mid resistor. I would expect the control to be anemic. In fact I would expect it to not work AND be anemic.

I'm surprised your screen voltage is higher than your plates. As mentioned, check to see that the power supply is wired correctly.

[rickyl]

Yes, the hum was completely eliminated by referencing the filaments to ground through two 100r resistors. The bass pot I wired the wiper to the cw terminal.

[Chuck H]

Ok. I see that now. Hmmm. Maybe cross that bridge when you have a proper cabinet as you mentioned. No sense evaluating and adjusting tone too much before you have all the sound components in place, right?

Another thing I just noticed is that you seem to be biased a bit cold. About nine watts. Most class A EL84 designs bias closer to 14W!!! That's for the total tube though and doesn't subtract for screen current. I'd shoot for a total tube current of 12.5W to 13W You'll probably need to go with a 220r or lower cathode resistor to get there.

Otherwise... All the other stuff I mentioned.

[eschertron]

What he said ^^^

I'm thinking that increasing R3 will kill the motorboating. As high as 10K is traditional, try different values to see how it affects the headroom in the preamp. I've gone as high as 68k to see what it sounds like. At that point all that happens is the preamp voltage is being controlled. It has no additional effect on stability.

Experiment with the EL84's cathode resistor. Alternately, experiment with the value of R1, to raise the plate voltage. That's another way to control dissipation.

[eschertron]

I'm surprised your screen voltage is higher than your plates. As mentioned, check to see that the power supply is wired correctly.

I've seen that the (cheap?) OT's that I've used drop more than a couple volts. More DC resistance than I expected, lowering the plate voltage below my design. I'm not too troubled by plates at or below screen voltage (that's what the screens are there for, right?) But if this amp's power supply is wired correctly, I might be tempted to make R1 smaller (as mentioned above), and R2 bigger in order to drop screen voltage and control it somewhat.

Aside... Even single-ended 100% class A amps can exhibit sag with a big enough screen node resistor. When the screen dissipation goes from 5ma to 25ma a 5k resistor will drop (25-5)ma * 5kR = 100volts. That's plenty to be heard if the preamp voltages all drop by this amount.