Help with El84 SE design

[then1xon5]

I am pretty new to building tube amps. I built a small 60fx5 hifi amp last year and I would like to build a SE low power guitar amp this year. I have the transformers from an old Hoffman console that should work great but I have some concerns about my design. I spent a lot of time looking at EL84 schematics and took what I thought were the best but easiest pieces from each and combined them. I would like to ask the gurus if there are huge flaws or improvements I should make.

There are a few areas that I am concerned about. One, I’m only using ½ of the second 12AX7 for the tone control. I didn’t see any reason that I needed it since the first gain stage is using the first 12AX7 and I would like to experiment a bit with the gain control. Second, the tone control seems very basic and concerns me a bit. Third, the transformer has a 5V secondary that I don’t really need. What do I do with it?

Thanks in advance.

Scott

[mooreamps]

There are a few areas that I am concerned about. One, I’m only using ½ of the second 12AX7 for the tone control. I didn’t see any reason that I needed it since the first gain stage is using the first 12AX7 and I would like to experiment a bit with the gain control. Second, the tone control seems very basic and concerns me a bit. Third, the transformer has a 5V secondary that I don’t really need. What do I do with it?

Thanks in advance.

Scott

only a few things that I can see. Tell you want I would do with this, if you really wanted to use everything.

1. Use the remaining 12ax7 gain stage as a cathode follower to drive your tone stack.

2. Use the 5 v secondary for a rectifier tube. Then make it into a full wave bridge using the tube and diodes. This will get you full voltage for the amp, and the benefit of a slow HV power up ramp.

3. For the EL-84, change the cathode resistor to 270 ohms, 3 watt.

4. For the EL-84, change the screen grid resistor to 2.7 K ohms, 3 watt.

5. For the EL-84, install a series grid resistor of at least 22K ohms, 1/2 watt. {without it, your amp will sound like an icepick}. Values up to 56K for a warmer Tweeder tone...

6. For the EL-84, install a parallel grid leak resistor of 100K.

7. On the second and third gain stages of the pre-amp, install parallel grid leak resistors of 900 K, 1/4 watt.

8. Optional : Go with DC filaments..

-g

[then1xon5]

I assume that I would add another 900k parrallel leak resistor on the forth gain stage when added? I have made all the chnages to the drawing but the rectifier. I have to draw a object for that one.

Is this correct?

Thanks very much for sharing your wizdom with me.

Scott

[pdf64]

Rectifier - do you understand the difference between normal full wave (2 diodes and centre tapped winding) and full wave bridge (4 diodes and no centre tap)? Your design has combined both and I fear there will be a big problem.
Rectifer - 500V? Do you understand the peak inverse voltage that a rectifier diode has to withstand? If 500V is the rectifier max voltage spec it is way too low.
el84 - It's good to have more B+ filtering between the plate and screen supply, as the screen is a grid which amplifies any ac hum from the B+ that gets input to it. So put an extra filter stage in - 1k / >47uF between the plate supply and pre amp supply.
Tone stack - 'the tone control seems very basic and concerns me a bit' most every guitar amp ever uses the same configuration, so you need have no concerns. However, that 750pF feeding the treble control is a bit large and will make it more of a upper mid and treble control and the perceived volume will change according to it's setting. 250pF is more normal.
Also putting the tone stack between the 2nd and 3rd stage doesn't make sense to me. If lots of gain is wanted then put it after the 3rd stage, or if best control of overdrive texture is wanted then put it between 1st and 2nd stage.
I would also advise grid stopper resistors on all tubes. They will prevent oscillation and help reduce blocking distortion if appropriately sized (10k - 220k).
Peter.

[then1xon5]

I have never used a center tapped secondary. My assumtion was that I would have a full wave coming from both outputs of the secondary and using a full bridge would rectify the full wave to positive DC voltage. Some of the schamatics that I looked at had the switch on the center tap for standby. Am I thinking about this the wrong way?

The other stuff makes sense. I posted the updated drawing with the grid resistors previously.

Thanks for the advice.

[then1xon5]

oh, I forgot two things...

I just put the 500V in for the bridge as something to put in. I currently have 700V diode bridges so that is most likey what I will use.

Secondly, if I move the tone control between the 1st and 2nd stages and add the fourth stage as recommened earlier, would I place the gain control between the 3rd and 4th stages? My thought is that it would control gain and still allow the 4th stage to drive for the EL84.

Thanks again!

Scott

[Bruce / Mission Amps]

Scott this amp, as drawn is going to be almost unusable, if it doesn't catch fire ha ha.
As pointed out the bridge rectifier is drawn wrong in one and you really only need a full rectifier (only two diodes) with this PT anyhow.... but that's assuming from your artwork, this is a 270v-0-270v high voltage secondary.
With a FW rectifier this will give you about +350vdc under a light tube load and very good for a SE EL84.
Using the full wave bridge... don't ground the center tap.... will give you about 700vdc!!
If you use a FWB and ground the center tap you'll probably blow the FWB rectifier and all the 450v-500v filter caps.
The standby switch in the center tap is OK for a FW rectifier.
Since it is drawn as VCT PT you can dump the solid state diodes and use the 5v line to power a 5Y3GT or GZ34 and no changes.

Secondly there is so much gain in the preamp that it will drive the poor SE EL84 to clipping with the slightest breath of signal on the input.
It only takes about 12-15vac of signal on the grid to smack an EL854 into submission and the way you have your preamp drawn, it could probably swing 45vac-50vac.

Gary's comments about using a cathode follower are a good place to start if you just want to use up one of the extra triodes... it will eliminate one gain stage but also offer a desirable, very low impedance to the tone stack making it work very well.
All you really need to get started is one 12AX7 with the tone and volume controls in between the two triodes.
Like a black face Champ.
If you want a gain monster... use all four triodes with the third triode being the cathode follower (zero gain) and the output of the treble pot going to the fourth triode in a lower gain setup up.
The fourth triode would be driving the single EL84.
Look at any 50 watt Marshall and pretend like you are going to use only one triode of the phase inverter tube (tube right before the two power tubes).
That layout does work... I've built a few... but the gain is still too high and you'll need to tweak the gain structure and the plate voltages so the three cascading amplifying triodes don't turn the EL84 inside out.

[mooreamps]

You missed the safety bais resistor on the fourth pre-amp gain stage. Otherwise, it looks pretty good.

-g

[mr.jetski]

hi all,

correct me if i'm wrong but isn,t there is a missing coupling cap between v1b and v2a. 200+ volts onto grid of v2a can't be good.

[Bruce / Mission Amps]

You missed the safety bais resistor on the fourth pre-amp gain stage. Otherwise, it looks pretty good.

-g

Hmmm... I guess it wouldn't hurt to have a large value resistor to ground on the wiper or that grid in case the pot goes bad.

[Bruce / Mission Amps]

hi all,

correct me if i'm wrong but isn,t there is a missing coupling cap between v1b and v2a. 200+ volts onto grid of v2a can't be good.

No, it's correct as drawn.
It is DC coupled... without getting in too deep, that triode section is sort of like boot strapped and fixed bias with the cathode of V2 a being held up high at 100K and the plate at full B+... it's a fairly common way to do a vacuum tube, cathode follower.

[mooreamps]

Hmmm... I guess it wouldn't hurt to have a large value resistor to ground on the wiper or that grid in case the pot goes bad.

You'd be surprised....

-g

[mooreamps]

No, it's correct as drawn.
It is DC coupled... without getting in too deep, that triode section is sort of like boot strapped and fixed bias with the cathode of V2 a being held up high at 100K and the plate at full B+... it's a fairly common way to do a vacuum tube, cathode follower.

As long as we are doing this, why not draw it up as a boot-straped CF. Except, I don't use 1 meg for the grid-to-cathode resistor value. The input impedance does not have to be "that high". 100K works just fine, and that value helps abate the instability problems I see just so many others seem to struggle with.....

-g

[then1xon5]

This is all good information. Thanks everyone for the help. i do have a couple questions regarding the gain stages though. Based on what I am seeing from Bruce's changes to my original "fireball" design, I still see four preamp stages. Since i have never built a guitar amp, this is really new to me. Can one of you tell me if this statment is correct regarding the stages?

First gain stage drives the original signal a bit. The 1M pot controls the amount of gain headed to the second stage. The second stage adds more gain based off how much signal is being let through the 1M pot. The fouth stage king of levels things out a bit using the 68k resistor and the pushes it a bit into the EL84.

Now the question. The third triode stage confusses me. How can signal come through the grid to the cathode of V3? I have never seen anything like that before. Second question, I still need a 900k resistor to ground before the fouth stage correct?

thanks again everyone.

Scott

[Bruce / Mission Amps]

This is all good information. Thanks everyone for the help. i do have a couple questions regarding the gain stages though. Based on what I am seeing from Bruce's changes to my original "fireball" design, I still see four preamp stages. Since i have never built a guitar amp, this is really new to me. Can one of you tell me if this statment is correct regarding the stages?

1., First gain stage drives the original signal a bit. The 1M pot controls the amount of gain headed to the second stage.
2., The second stage adds more gain based off how much signal is being let through the 1M pot.
3., The fourth stage king of levels things out a bit using the 68k resistor and the pushes it a bit into the EL84.

Now the question. The third triode stage confusses me.

4., How can signal come through the grid to the cathode of V3? I have never seen anything like that before.
5., Second question, I still need a 900k resistor to ground before the fouth stage correct?

thanks again everyone.

Scott

There are 5 items to address.

1. As drawn, this 1st stage (and others) can have HUGE amounts of gain and all you can do is attenuate it with series resistors, lack of cathode bypass caps, negative feedback or the volume pot.
With a small 100mvac signal from your guitar pickups into the grid of stage 1, you might have as much as 3.5vac to 5vac at the plate, driving that volume control. That is a gain of between 35:1 and 50:1!!
I suggest this stage be tamed down quite a bit before getting to far along with the project. You could also use 5965, 12AY7, 12AV7 tubes instead.
Another choice would be to reconfigure the plate load resistors and use a 12AU7, 5963... or even a 12AT7 which is still pretty hot compared to the others.

2. The second stage also has a huge amount of built in gain, probably around 20:1 or more... none of which will be lost before driving the cathode follower.
This stage will need tweaking too.

3. The fourth stage also has plenty of gain but it is after the tone controls and the master volume, which do lower the amount of available signal.
I'd guess this stage will probably be in the +20:1 to 30:1 range also.
The 68K resistor is not to reduce anything except to change the bandpass response of the tone controls, the master volume and the grid of the driver stage. This will slightly inhance the tone and frequency roll off curve of the low end with respect to mids and high end.

4.The 100K cathode resistor is not bypassed to ground so what ever appears on the grid will also appear on the cathode. It has <unity gain... which means it's output will be slightly less then 1:1. But it is extremely low impedance so the effects of the tone controls (as a lumped RC load) is very very low.
This makes the tone control quite effective because what ever signals are left over after you have steered unwanted signals to ground (so it sounds good to you) will not have much effect on the cathode because the tone stack's impedance is much higher then the cathode follower.

5. No, you do not need the resistor... it was only mentioned by Gary as a safety device in case the pot fails where the junction of the wiper and the carbon pile in the pot goes bad and there is no longer a reference to ground on the grid.
If the pot fails in that manner and the grid's resistance to ground is lost, the tube will become under biased (zero bias) and draw maximum idle current, possibly damaging the triode... if not the amp will be horribly noisy too.
CAVEAT: In my MANY MANY years of working on tube amps, I have seen this failure happen so very few times that I don't even think of it as an issue, but Gary brought it up and there isn't anything wrong with using a large value resistance here at this stage since there will be too much gain anyhow.

OK getting back to all this gain.... remember what I said at first that a Class A amp with a single EL84 only needs a few volts of signal to turn on full blast and beyond. Typically that is around 10vac to 15vac of grid drive signal.

Unless you are looking for a buzz box that sounds like a paper sack full of angry bees, ....
with a small 100mvac signal from your pickups, the preamp as drawn will have much more gain then is needed or quite possibly, be used.
You should note that every stage except the cathode follower has a gain potential of +25:1.
That will take the drive level at the power tube up to a level that will swing from zero v to full B+ rail voltages... which of course can't happen but suffice to say it still will be high enough to drive the power tube into insane overdrive, cutoff, clipping and saturation.... all of which will sound terrible after the first few minutes of "It's alive" bliss.
So a little gain restructuring and signal reduction here and there will make the amp as a whole much more usable.... read that as "final tweaking".
That's also why I said two triodes of one 12AX7 is all you really need to have a rockin' dog amp... but you asked what to do with the other two triodes of the second 12AX7.