Adapted 5F2A with EZ80 and EL84

[stephenhartley]

I've swapped out the el84 for another and some of the noise has gone - I uncoupled the negative feedback resistor and it was slightly harsher and brighter so I put it back. The same problem remains however that if I turn it above 2, it's breaking up into a horrible brittle distortion. I'll try some of the above suggestions tomorrow - maybe the 5F2A circuit and the 190-0-190 80mA transformer that I'm using just aren't suited to an EL84?

[g1]

Your heater voltage is holding up? I'm only asking because this rectifier runs on 6.3V heater.

[stephenhartley]

The heater circuit (6.3v) runs from the pilot light to the rectifier to the el84 to the ECC83. It doesn't have a centre tap - I've earthed it to the cathode resistor earth as recommended on here by one of the American experts. I don't think the heaters are the problem.

[Chuck H]

You're only biased at 7W idling (total dissipation***) with 221 working volts on the plate. You should be entirely safe reducing the cathode resistance to get closer to center bias. Try a 100R cathode. Maybe even lower. Ideal would be about 12w or 13W at idle. Maybe even 14W because the bias does get colder as the tube conducts signal. There are even market examples of amps running the el84/s at that level with higher plate voltage. ***One thing to consider is that the 11W max plate dissipation doesn't include the screen dissipation, which is part of the total tube dissipation as measured at the cathode resistor. Right now you're at 7W. That means something more like 5.5W or at most 6W of plate dissipation for a class A amp that should be at 100% plate dissipation for center bias.

I'd try a 100r cathode resistor and remeasure. You may even need to go lower.

[stephenhartley]

Thanks - I'll give it a try. I'll try the 5000 ohm tap again now that I've replaced the faulty el84.

[Dave H]

Right now you're at 7W. That means something more like 5.5W or at most 6W of plate dissipation for a class A amp that should be at 100% plate dissipation for center bias.

I'd try a 100r cathode resistor and remeasure. You may even need to go lower.

Chuck, With the plate voltage and OT impedance it has I don't think it can be centre biased at 100% plate dissipation. It's close to centre bias now at 7W, 30mA on the 5k tap. It would need a lower impedance OT to be centre biased at 100%. Having said that it's a guitar amp, it doesn't have to be centre biased. If it sounds better with a lower cathode resistor then go for it.

SE class A amps are usually designed to run centre biased at the rated Pd max of the tube for maximum power output but the don't have to be, they could be designed to be centre biased at a lower dissipation for longer tube life.

[Chuck H]

Chuck, With the plate voltage and OT impedance it has I don't think it can be centre biased at 100% plate dissipation. It's close to centre bias now at 7W, 30mA on the 5k tap. It would need a lower impedance OT to be centre biased at 100%. Having said that it's a guitar amp, it doesn't have to be centre biased. If it sounds better with a lower cathode resistor then go for it.

SE class A amps are usually designed to run centre biased at the rated Pd max of the tube for maximum power output but the don't have to be, they could be designed to be centre biased at a lower dissipation for longer tube life.

Thanks Dave. I've never been able to wrap my head around load lines. I'm working from bench experience and example circuits. In which case 150R for the cathode with 220-ish working volts on the plate would be cold for a guitar amp. I'm also considering that the bias will cool considerably when the amp is clipping hard. On the bench what I've seen in that case is a higher spike on the positive swing. When biased so that the clipped wave form is more symmetrical that spike is reduced. YMMV.

[stephenhartley]

Thanks for the expert advice everyone. It's much appreciated.

I'll get back on it next week when I have time, and report back - one change at a time to see what works - I'll measure the voltages at each stage:
1. OT back to the 5k tap from the 8k tap
2. lower the cathode resistor
3. Double up the feedback loop resistor
4. screen grid resistor
5. dropping HT with an extra filter stage (like AC4) - that one's a bit beyond my novice scope

[Tom Phillips]

...5. dropping HT with an extra filter stage (like AC4) - that one's a bit beyond my novice scope

That change will also lower the hum that exists in many small single ended guitar amps. This is because the extra filter stage will reduce the ripple in the output stage plate supply. For a given level of ripple a single ended power amp will hum more than a Push Pull output stage because the single ended amp does not benefit from the hum cancelling that is inherent in the PP power amp circuit.

[stephenhartley]

it does have quite a loud buzz which disappears when I plug the guitar in. I've spent a few hours today experimenting with different cathode and grid resistors and no matter what I try, I can't seem to get the dissipation wattage above 7. The battery on my multimeter went so it's time to give it a rest and have a fresh start.

I have two 60s SE EL84 amps - a Selmer little giant and a Broadway, both with awesome sound and I wanted to replicate that but keep the early Fender style build that I've used on my 20w & 30w builds - hence the rational of trying to adapt the 5F2A to an EZ80 and EL84.

[Chuck H]

it does have quite a loud buzz which disappears when I plug the guitar in.

That's almost surely a wiring error.

I've spent a few hours today experimenting with different cathode and grid resistors and no matter what I try, I can't seem to get the dissipation wattage above 7.

I anticipated this could happen since cathode bias is self adjusting. I still think you'll be better with the 100r cathode though because it should reduce bias shifting cooler when the amp is clipping. You duplicated the 150r value from the Selmer design, but notice that the Selmer's plate is some 60V higher than this build.

I wanted to replicate that but keep the early Fender style build that I've used on my 20w & 30w builds - hence the rational of trying to adapt the 5F2A to an EZ80 and EL84.

There's nothing inherently wrong with your tube substitution. The el84 will be more easily driven to clipping by the preamp circuit than a 6v6 would. No great shakes. If it bothers you, you can always reduce drive voltage. Your plate voltage is quite low compared to similar amps. This would reduce output some relative to what you're use to perhaps. I think the combination of the lower output and the el84 clipping at lower drive voltage than a 6v6 may be creating a sort of illusion of low headroom. The amp may actually be making all the watts it can before clipping even though that starts at vol-2. It may also partly explain the harsh tone. Perhaps there's just too much drive voltage on the el84 grid. You can really only clip a single tube so much before reaching a point of diminishing returns where bad things start to happen. I've had this problem in two single ended amps I built. In both cases the amps sounded harsh and started to roast the screen grid resistor. Limiting drive to the grid (and reducing my expectations for clipping level) was the solution.

[Tom Phillips]

it does have quite a loud buzz which disappears when I plug the guitar in...

That's almost surely a wiring error...

I agree and my guess is that the sorting bar on the input jack is not grounded.

[stephenhartley]

I agree and my guess is that the sorting bar on the input jack is not grounded.

Both input sockets are grounded with their own wires to a dedicated ground along with the cathode resistor from the first half of the pre-amp tube. The pots are grounded separately then the filter caps have their own ground. The +ve side of the heater circuit (no centre tap) is grounded to the - ve of the EL84 cathode resistor which too has its own ground. There's obviously a problem with my grounding scheme - it's just finding it.

'Perhaps there's just too much drive voltage on the el84 grid'

The grid voltage has been measuring ~ 170v - how do I alter it?

[Dave H]

'Perhaps there's just too much drive voltage on the el84 grid'

The grid voltage has been measuring ~ 170v - how do I alter it?

Is that 170V DC on the screen grid pin 9?
It's the signal voltage on the control grid pin 2 that's too high.
If the amp vol is below 2 the tone control won't work very well which could be part of the problem. As a test try turning the guitar's vol pot down so the amp's vol pot can be set to 5 and still be clean then adjust the amp's tone pot as required. The amp tone pot will work better with its vol pot set to 5. If that sounds OK the gain of the amp could be reduced to make its vol/tone controls have a more useful range but it will never have much clean volume. I'd guess it's only about 2W because of the low B+

[Tom Phillips]

Both input sockets are grounded with their own wires to a dedicated ground along with the cathode resistor from the first half of the pre-amp tube...

Understood. However, that is a different issue than what I was discussion in post #27. If the buzz you mentioned stops when you plug in the guitar then I don't believe it is caused by the grounding topology that you just described. Are you using shorting jacks for your inputs? If so then the tip shunt switch contact should be grounded so that the input is shorted to ground when you unplug the guitar cord. I'm still wondering what type of input jacks you are using and how they are wired.