7W DIY clean amp for kid's Christmas present

[Grubb]

My oldest two kids (10 and 12) have started playing guitar this year and are picking it up quick. I am designing them a clean pedal platform amp for Christmas that's based on diy projects or schematics I've seen. Some experienced eyes to look over the schematics before I layout a PCB would be appreciated as there are some parts I'm not confident about.



In particular I'm not certain about the speaker protection circuit, which is a 555 IC operating a delay. I copied a circuit from eevblog which I think is doing what I want - muting the outputs of the power amp via switching a relay for a short period of time on start up to keep thuds and pops out of the speaker. This circuit is set up for 5 seconds delay, and I'd prefer 3 so I may need to tweak it. I'm still at the copying recipes while trying to learn stage of electronics.

Any feedback or suggestions for improvement would be most welcome 🙏

[Steben]

What a great idea and nice present!
A complete full option amp is far from the goal here as I see it.
Yet reverb is already present.
What I would add is some clipping protection. Just some basic diodes to ground in order not to let the chips clip.

[Grubb]

Thanks Steben.

Yes, it only needs to be basic, I have lots of DIY pedals the kids can use for drives or other effects. But I wanted reverb in the amp design because a guitar without any reverb sounds pretty terrible to me unless there's a lot of distortion.

Where would you put these clipping protection diodes?

[Steben]

Thanks Steben.

Yes, it only needs to be basic, I have lots of DIY pedals the kids can use for drives or other effects. But I wanted reverb in the amp design because a guitar without any reverb sounds pretty terrible to me unless there's a lot of distortion.

Where would you put these clipping protection diodes?

Depends on the sensitivity of the power amp chip.
Gain is about 26 afaik. A silicon diode pair at the input won't cut it. 0.6 * 26 = 15.6 Volts. Far beyond the supply/2 = 4,5V.
Given the preamp has a possible gain of 101 .... yikes.
Master at full will always give the power chip going nuts.

The following might be handy:
- perhaps thinking about higher supply voltage in order to make a 0.6V diode pair at the power amp input useful.
- lowering max gain in the preamp. Enough to recover the EQ, but 101 is too high. If pedals are used the input signal will be more than high enough. And high gain in a clean neutral circuit is not wanted.

[ElectricDruid]

A few comments

R7/10M at the input seems too big. 10M is going to be noisy. 1M would be enough, 2M2 if you must. Not sure why they went for such a huge value.

Changing the time constant for that 555 is simple enough - change C7/10u for something smaller to reduce the time. 4u7 would roughly halve it, so maybe 6u8?

Personally, I'd connect the reverb stage directly to the Pre-amp output and leave the switching out. After all, with the reverb control at zero, the dry signal goes straight through anyway. You could drop the extra buffer IC5.1, since it'd be fed directly from IC2.2, and you could drop the extra buffer IC3.1 in front of the power amp, since it'd be fed from IC5.2. That saves you two op-amps, which is one whole chip!

HTH

[Steben]

A few comments

R7/10M at the input seems too big. 10M is going to be noisy. 1M would be enough, 2M2 if you must. Not sure why they went for such a huge value.

Perhaps because high higher highest input impedance is one of the most spread mantras in guitar circuit design. And low impedance a taboo.

[ElectricDruid]

A few comments

R7/10M at the input seems too big. 10M is going to be noisy. 1M would be enough, 2M2 if you must. Not sure why they went for such a huge value.

Perhaps because high higher highest input impedance is one of the most spread mantras in guitar circuit design. And low impedance a taboo.

Perhaps. But isn't noise bad too?!?

[Grubb]

Thanks for comments and suggestions. I can see now the dry tone path in the reverb circuit. I will probably simplify the circuit as suggested.

Changing input resistor R11/4K7 to 10K in the preamp gives me a gain of 50, is that a better level? I'm not sure what I should be aiming at here.

I did notice that 10M on the input of the preamp is a lot higher value than most of the pedal designs I've looked at. I'll happily put a standard 1M resistor in its place if that will improve it.

[Grubb]

I'm working those changes now. In removing the superfluous buffer stages, can I remove decoupling caps like C21/100N? I have removed C14 already as C17 makes it unnecessary once the power amp buffer is removed.

[PRR]

7W ..amp ...... the speaker protection circuit....

Leo never put "protection" on his 6W Champ. Really on any of his amps. (Except the 300; which shut-down sections when there were not enough speakers plugged in.)

Use a "15 Watt guitar speaker". It will totally shrug-off the really minor thump of that bridge chip (ideally no thump). It will also survive hours of GROSS DISTORTED FEEDBACK, which is the usual death of over-worked guitar amp speakers.

Jensen® MOD®, 5", MOD5-30, 30W,  $36.01
Jensen® Vintage Ceramic, 6", C6V, 20W  $36.43
Jensen® MOD®, 8", MOD8-20, 20W  $38.74
Celestion, 8", Ceramic Eight 15, 15W  $45.00  The Eight 15 is ideal for upgrading your bedroom blaster

[niektb]

Can't you just use a standard N-Channel MOSFET instead of the NE555?

[ElectricDruid]

I'm working those changes now. In removing the superfluous buffer stages, can I remove decoupling caps like C21/100N? I have removed C14 already as C17 makes it unnecessary once the power amp buffer is removed.

Yes, you'll be able to drop quite a few unnecessary coupling caps and bias resistors. Once you redraw it, it'll become obvious. Anywhere where you've got two coupling caps one after the other, one of them can go. And anywhere you've got two bias resistors connected to the same bit of circuit, one of them can go. Post the revised version and we'll double check for you.

[Grubb]

OK, I made some revisions. So far, I have changed the input resistor in the preamp R11 from 4K7 to 10K, halving the gain. I am happy to reduce this more if we think it is needed. Switching has been removed so that the reverb is in the signal path but only audible with the knob turned up.

I've left the decoupling cap C17 and the bias resistor R10 at the front of the power amp circuit and removed the buffer.
On the reverb circuit input, I left the decoupling cap C21, removed the buffer IC and also took out the bias resistor R15 here because of R13 nearby. Please let me know if that's an error.

Things I am considering:

• removing the speaker protection circuit - in a YouTube video I saw with this power amp circuit, Erik Vincent turns his on and there's no noise at all
• adding a charge pump to run at 18v - better headroom, diode protection possible? I dunno. This means I'd possibly need to attach the heatsink on the power amp chip to my enclosure, which I was hoping to avoid for external aesthetic reasons.

I really appreciate the help I've received here, thank you so much. Here's the schem:

[ElectricDruid]

If you look at the output from the pre-amp and the input to the reverb, you've got cap(C11), resistor(R9),cap(C21). That's two caps in series, so one can go. Drop C21. Once that's done, you'll see that you've got two bias resistors R9 and R13, both doing the same job, so drop R9.

Similarly, I *think* you can get rid of R17, R18, R19 and C20 from the output of the reverb circuit, and then connect directly to C17. R10 can also go because the output from IC5.2 won't need biasing (it'll already be biased because there was bias on its inputs).

This process of simplification is pretty common if you build a schematic out of existing circuits. A typical circuit looks like "input stuff - circuit - output stuff", where input and output stuff is coupling caps, anti-pop resistors, biasing resistors, etc etc. If you join two or three of those together you get: "input stuff - circuit - output stuff - input stuff - circuit - output stuff - input stuff - circuit - output stuff". Clearly those "output stuff-input stuff" pairs in the internals of the circuit can all go, so we can generally get rid of quite a lot of stuff.

HTH,
Tom

[anotherjim]

That power amp IC already has a mute function that can help reduce switch-on pop. See figure 5 in the datasheet pdf...
https://www.st.com/resource/en/datasheet/tda7266.pdf

Nice choice of opamp in the preamp. One thing is this design looks hissy to me with a x1 buffer stage first followed by a lossy tone network followed by a boost amp. That boost has to amplify all the circuit noise before it. I would change the first buffer stage to an amplifier with some gain fixed at around x5. This should improve the signal-to-noise ratio since the following circuits won't have to amplify so much. I should want more than x5 in the first stage but with only 9v and a desire for clean headroom, you can't get too greedy. However, you have an opamp that clips without causing too many nasty noises

A charge pump will have to be good to supply more power to the power amp. The usual stompbox charge pump IC's won't do it. A switching DC-DC boost converter can. Small modules can deliver 3A at 18v easy but that power has to come from outside and that would be pulling in 6A at 9v. I would advise against boosting to 18v or you will also be limited to 8ohm speakers - power will be too much with a 4ohm.

[Steben]

Nice choice of opamp in the preamp. One thing is this design looks hissy to me with a x1 buffer stage first followed by a lossy tone network followed by a boost amp. That boost has to amplify all the circuit noise before it. I would change the first buffer stage to an amplifier with some gain fixed at around x5. This should improve the signal-to-noise ratio since the following circuits won't have to amplify so much. I should want more than x5 in the first stage but with only 9v and a desire for clean headroom, you can't get too greedy. However, you have an opamp that clips without causing too many nasty noises

+200%.
That opamp will clip easily. And the OP mentions a clean amp....
It just strikes me now there is a level AND master control doing pretty much the same.
So I concur with the gain in front and dropping the current gain stage.

[antonis]

IMHO, Tone pots are of too big values..
5 times lower values could also be easily driven by IC2.1 and not loaded by R6..
(of course, respective caps values should be proportionally altered ..)

[Grubb]

Awesome! I'm learning a lot from this experience, thank you all so much. I have revised the circuit further.

Added:

• 10UF capacitor in power amp circuit, in line with the datasheet schematic for temporary mute on start
• 47K resistor in parallel to the new cap

Removed:

• All of the mute circuit, as the power amp chip has this functionality inbuilt (thanks anotherjim!)
• Superfluous input/output stuff at circuit transition points (thanks ElectricDruid!)

To be addressed:

• Changing capacitor and pot values in the tonestack - what is the benefit of this? I am happy to give it a go but would like to understand the logic behind this suggestion for my own learning
• Rejigging the preamp section, see below

With the preamp, I am thinking to change the Level control to a Gain control at the first opamp stage. This avoids the redundancy of having Level and Master controlling essentially the same thing and should also reduce potential noise. If this is done, what do I do with the second opamp stage? Do I need a single opamp rather than a dual here?

The schematic as it stands:


Preamp - work in progress:

[ElectricDruid]

Fantastic! Looking a lot tidier!

Changing capacitor and pot values in the tonestack - what is the benefit of this? I am happy to give it a go but would like to understand the logic behind this suggestion for my own learning

It was Antonis' suggestion, so I think he should explain his logic, but broadly, he's talking about the output impedance of the tonestack versus the input impedance of that buffer stage. R6 sets that input impedance (1M) which is a nice high value, but it's not so high in comparison to pot values of hundreds of K.

Preamp - work in progress:


With the preamp, I am thinking to change the Level control to a Gain control at the first opamp stage. This avoids the redundancy of having Level and Master controlling essentially the same thing and should also reduce potential noise. If this is done, what do I do with the second opamp stage? Do I need a single opamp rather than a dual here?

I think what you've got looks good (except for one wire from pin 1 to pin 2 on IC2.1 which looks like a hangover from the previous version of the circuit - get rid of that). I would guess R6 is unnecessary because the tonestack is connected directly to Vref itself, but if you keep R6, you need a capacitor in front of it (100n or so).

You've moved the gain ahead of the tone stack, so the tone stack is being fed a larger signal. Since there are losses in the tone stack, this is a good thing, and it avoids the need to add more amplification afterwards which would boost any noise from the tonestack as well.
Keeping the second op-amp as a buffer like you've done is a good idea in my view because the tonestack is not a low output impedance and like that, the next stage of the circuit (the reverb stage) has a solid drive direct from the output of an op-amp.

[Grubb]

I think what you've got looks good (except for one wire from pin 1 to pin 2 on IC2.1 which looks like a hangover from the previous version of the circuit - get rid of that).

Yikes! That was not supposed to be there, thanks for spotting it.

Is there enough gain in the rejigged opamp section? Currently, it has a gain of around 6. I left it there because anotherjim suggested about 5 in his post. But I wasn't sure if that was a gain of 5 plus the second amplification stage.

Edit: does removing R6 mess up the input impedance of the buffer?