Solid State 'AC30' clone

[James UK]

Probably a little low, but still high enough that it should work OK. Lower drive levels have been used successfully in commercial amps. Rod Elliot suggests around 30-60mA RMS for 8Ω tanks.

Okay, it would be good to increase the drive current a bit then, hopefully then requiring less recovery gain. If I decrease R21 to, say 10 Ohm, does that not risk damaging the op-amp with excess current?

There's still a load (R32) with the tank disconnected, so the current will stay the same (unless the driver circuit clips, which doesn't happen in this case).

I thought that having the impedance of the tank in parallel with the R32 would change the current as the load has changed.

I noticed that you're using an NE5534 for the recovery stage. That chip isn't unity gain stable without an external frequency compensation capacitor between pins 5 and 8. I don't see any external compensation, so that stage may actually be oscillating because of C26.

I thought that the NE5534 didn't require the capacitor for gains >5 so I omitted it. What I seem to have missed, from what you've said, is that C26 lowers the gain to unity at high frequencies and therefore the cap is required. Is that correct? I suppose I can solder one directly to the pins on the IC.

1.For several reasons the reverb return amp should incorporate a highpass with a cutoff frequency in the region of 500Hz. With R35=220R und C25=10u you get fc about 70Hz. Reduce C25 to 1.5uF.

This seems like a good idea. Is one of the reasons to reduce low frequency feedback as this would make a lot of sense!

Thanks,
James

 

[bucks bunny]

btw thumbs up for your perfect case working

 

[bmc0]

If I decrease R21 to, say 10 Ohm, does that not risk damaging the op-amp with excess current?

Not sure what you mean. The load on the op amp is (mostly) determined by R22 and R23. If you decide to decrease R21, it may possibly be a good idea to increase C18. With 33Ω and 47µF, the highpass frequency is 103 Hz. With R21=10Ω, the frequency is 339Hz.

I thought that having the impedance of the tank in parallel with the R32 would change the current as the load has changed.

Since it's set up for current drive, the output current (all of which goes through R21) is proportional to the input voltage unless the circuit is driven into clipping. It doesn't matter what the load impedance is. R32 serves to set a maximum load impedance (or you could think of it as a finite output impedance—equal to the resistor value—from the perspective of the tank) so the output voltage doesn't go crazy at high frequencies.

What I seem to have missed, from what you've said, is that C26 lowers the gain to unity at high frequencies and therefore the cap is required. Is that correct?

Yes.

Definitely not.

No, it's definitely current drive (ignoring R32). Note how the feedback is connected. The voltage across R21 is proportional to the current through the load.

 

[bucks bunny]

Definitely not. To achieve this, the inputs of the opamp must be swapped at least.

My apologies, I was wrong. Should not respond before the first coffee

 

[James UK]

Not sure what you mean. The load on the op amp is (mostly) determined by R22 and R23. If you decide to decrease R21, it may possibly be a good idea to increase C18. With 33Ω and 47µF, the highpass frequency is 103 Hz. With R21=10Ω, the frequency is 339Hz.

This might not be a bad thing. I was considering that the cutoff was too low anyway and might benefit from being less muddy.

Okay, hopefully I'll get time to make some changes tonight. I'll start by scoping the output of the NE5534 to see if it's oscillating. I'll add the compensation cap regardelss but it will be useful to see if this is indeed the case. I'll also try reducing the value of R21 to see if the drive current increases noticeably. I'll also reduce C25 to increase the cutoff frequency of the recovery amp.

I'll report back when I've made the changes.

Thanks,
James

 

[James UK]

Hi,

I've had a chance to work on the amp this evening and made the following changes.

-33pf compensation capacitor added to U5 (It wasn't oscillating but it seemed a good idea regardless.)
-C27 changed to 47n
-C25 changed to 2u2
-R21 changed to 10R
-R19 changed to 10k
-C15 changed to 100n
-R37 changed to 39k

-Tank wrapped in foam, then the tolex bag, then placed on three blocks of foam. I haven't secured it to the bottom yet so I'll probably do that tomorrow. It does raise the tank quite far off the base of the amp which is annoying but it does do a good job of isolating the tank from feedback.

The electrical changes got it most of the way there and the tank dampening finished it off. The reverb now works acceptably! Thank you for all the suggestions!

Thanks,
James

 

[Blue Steel]

Dude those clips sound awesome.

 

[James UK]

Dude those clips sound awesome.

Thanks! I'm going to make some proper demos in the next couple of weeks, hopefully. I used it for a gig last night and it sounded good, definitely 'voxy'. As is my experience with other solid state guitar amps, the heatsink barely got above room temperature. I probably could have got away with just mounting the transistors to the chassis, like a lot of guitar amps do.

I'm surprised at how quiet the amp is, noise wise. It's the lowest noise guitar amp I've used, not a trace of hum or hiss even turned right up. I'm not sure of the reasons why commercial solid state guitar amps seems noisier. When fiddling with the reverb, I was amazed at how much difference fitting the back panel, with metal tape on the inside, made. With it removed, the amp was picking up French radio, replaced, absolutely silent.

Thanks,
James