TDA2050 could work as a guitar amp?

[bool]

If you want to work with unipolar supplies, TDA 2003, TDA2004/TDA2005 and TDA2009 will have a significant edge over TDA2030/TDA2040/TDA2050. Both in terms of output power and cost. They are "almost" indestructible as well, they were built to work in "hostile" environment, like car radios etc.

With bipolar supply, I would try to use a TDA2040 - I think it is able to deliver larger output before clipping compared to 2030/2050 - at lower supply voltages. These will blow up more frequently compared to the "car radio" chips (esp. TDA2030).

I wouldn't even try to use the LM1875 for guitar. This is a Hi-Fi chip of a Gainclone fame and it has I think even lower supply-to-output-power ratio than TDA2030 and doesn't like lower supply voltages as much as TDA chips do. Subjectively, it sounds better, IMHO, but hey...

So, there ...

[Crontox102098]

This is my last "desing"... i guess it could work, i'll see in a couple of days  ... Too i'll try to use TDA2040 instead TDA2050  ...


What do you think about the schematic and those this things?

[bool]

R3 is wired wrong. R8, too. Won't work.
C16 will blow up.

[Crontox102098]

Corrected... i don't see it  dunno what i'm thinking about when i'm drawing.

[bool]

Btw. if I were you, I'd reduce the C2 substantially. 100nF should be plenty. Now you're amplifying all rumble and hum that comes through input. Also, I'd put a small 50-100pF cap across the R4, to shunt the RF component - to prevent the opamp to amplify any stray RF interference. With this in mind, I would also put a 100pF cap across the R2 -  to shunt the incoming RFI to ground.

As a safety measure, I would add a 4k7 resistor in series with the C1, to prevent static surges etc. to damage the opamp.

So good luck with your project.


======================
With that in mind, your R6 is of wrong value (should be 220K or 240K) and your second opamp is wired wrong (swapped + and - inputs)

IMO the Pot 5 would be better at 100K

[Crontox102098]

Btw. if I were you, I'd reduce the C2 substantially. 100nF should be plenty. Now you're amplifying all rumble and hum that comes through input. Also, I'd put a small 50-100pF cap across the R4, to shunt the RF component - to prevent the opamp to amplify any stray RF interference. With this in mind, I would also put a 100pF cap across the R2 -  to shunt the incoming RFI to ground.

As a safety measure, I would add a 4k7 resistor in series with the C1, to prevent static surges etc. to damage the opamp.

So good luck with your project.


======================
With that in mind, your R6 is of wrong value (should be 220K or 240K) and your second opamp is wired wrong (swapped + and - inputs)

IMO the Pot 5 would be better at 100K

I gonna take your suggestions but... a little thing i dont understand... as you say (I'd put a small 50-100pF cap across the R4, to shunt the RF component &  also put a 100pF cap across the R2 -  to shunt) where i should put the cap? i didn't understand, sorry... i speak spanish.

And... this is before or after the C1? (I would add a 4k7 resistor in series with the C1)

I'm just a little beginner 

[R.G.]

Just kibitzing.

You already have bipolar power supplies, in the form of +/-24V, and supply +12V from a regulator to your opamps. Why not use a 7912 regulator from -24V and be able to run your opamps from +/-12V?  This elminates a lot of components in there for biasing and blocking DC levels in the preamp section.

You could eliminate R1, R6, C6, C8, and maybe C10. This has benefits beyond eliminating components, as the biasing you have for IC1A is noisy right now, and the bias string for IC1B is... odd. A 22K and a 240K will bias this thing up near +12V, and it probably won't pass signal much if at all. That is, assuming you flip IC1B so the + input gets the input signal and not the feedback. As drawn, it will not work at all, since the + and - inputs are swapped.

I would also put large diodes (1N4004/5/6/7) at the speaker output, reverse biased in normal operation to protect the output against inductive loads - like speakers.

In general, you ought to check the frequency turnover points of all the caps you do use and see if they make sense or just let junk in.  For instance, in addition to C1 letting in stuff down to F = 1/(2*pi*100n*500K) = 3.18Hz, C2 and R3 will amplify it. In fact, C2 and R3 amplify down to F = 1/(6.28*10E-6*100K) = 0.159Hz. Kicking both of these up to about 70Hz by lowering the capacitances makes sense for guitar, which goes no lower than 82Hz in standard tuning.

I haven't checked it all, just noticed these few things.

[samhay]

The inputs of IC 1B are reversed (I do that a lot too) and it is biased to ~ 11V, which won't work if the op-amp positive supply is 12V, and is just weird if it is 24V. R6 and R7 need to be the same value. As RG said, go for a bipolar supply and then bias the op-amps to ground.

As said, a cap across R4 and especially R9 is a good idea.

Nice to see an op-amp-driven tone stack with the resistors scaled down to sensible values. Did you design this, or does it come from somewhere else?

Edit - modified once I re-read the voltage regulator value.

[Crontox102098]

You have biased the input of IC1B to +2.5V. I suspect you meant this to be at 4.5V or 5V, so either remove R2 (probably the better choice) or replace the VC connection with on to VA.

The inputs of IC 1B are reversed (I do that a lot too) and I have no idea why you have it biased quite like this, although it will get you to ~4.5V.

As said, a cap across R4 and especially R9 is a good idea.

Nice to see an op-amp-driven tone stack with the resistors scaled down to sensible values. Did you design this, or does it come from somewhere else?

The IC are reversed, yeah, just a drawing error. The tone stack comes from a Google image with some modifications, i tested it a long time ago and i put here because i think that it could work well with this...

I'm gonna take all your suggestions and gonna try to fix all

[Crontox102098]

Look at this 

Should breadboard it tomorrow? 

[samhay]

I forgot to add these, but it is a good idea from RG:

'I would also put large diodes (1N4004/5/6/7) at the speaker output, reverse biased in normal operation to protect the output against inductive loads - like speakers.'

[bool]

Yep diodes .. TDA2030 needs them, but the other two should work without them ..

It's always better to be safe, so here's how:



I also think that the C12 (from your schematic) is a bit too big, I would make it 100nF

Time will tell if C9 is also a little too big?


And the "other" thing:

Make the R9 as a "gain" potentiometer, imho 100k or 200k should be ok, and add a pair of clipping LEDs across it a-la tubescreamer, and make those LEDs switchable (so you can have two different distortion flavors on the cheap).

[Crontox102098]

Yep diodes .. TDA2030 needs them, but the other two should work without them ..

It's always better to be safe, so here's how:



I also think that the C12 (from your schematic) is a bit too big, I would make it 100nF

Time will tell if C9 is also a little too big?


And the "other" thing:

Make the R9 as a "gain" potentiometer, imho 100k or 200k should be ok, and add a pair of clipping LEDs across it a-la tubescreamer, and make those LEDs switchable (so you can have two different distortion flavors on the cheap).

TDA2050 requires C12 (.47uF) i think is well... SURE! i'm gonna put the gain, great idea! 

[bool]

If you stick with the values you listed (C12, 0,47uF), then use at least 2W part for the R17.

Distorted guitar signal contains more "square wave" and fast transients at higher amplitude (compared to f.e. listening to music), so the zobel network will suffer more onslaught at the values listed and quite probably heat more.

Lowering the C12 value to the least possible value that would maintain chip stability would reduce the (averaged) current through this network and so reduce the dissipation, but would require some testing and close inspection that everything works as it should.

So, to recap - if you want to stay on the safe side, use a fat 2-or-3-watter for the R17.

[Keeb]

then use at least 2W part for the R17.

I built a tda2050 amp with the datasheet schematic. I fried that resistor almost seconds after trying it out. Replaced it with a 2W resistor and the amp works great!
So, in my experience you don't have  a choice - that needs to be at least 2W.

[Crontox102098]

Another thing guys, and if i want to put headphones? 

[bool]

Simply use a resistive divider at the amp output (so between "out" and "ground").

I would try 150 (or 180) ohm into a 68 ohm resistor (both at least 1-watters). Take the phones out at both resistors' junction. You may want to add a 47n - 100n capacitor in parallel with the 68 ohm resistor to act as a simple lowpass filter to reduce fizzy frequencies in your phones.

[Crontox102098]

One last thing ... I will buy the transformer... how many amperes should have the transformer to run this amp?

I can use this scheme to buy a transformer without TAP?

[bool]

I'd shoot for 3Amp trafo at least. The schematic you posted will work, but not as good as a proper tapped transformer. It will sag - which is not as good with silicon chip amps as it is with tube amps, but hey...

A 9V 30W toroid will get you there with your schematic. Be careful with rectifiers - use at least 1N5404 (any num. to 1N5407). 1N540x are 3Amp parts. Or use something stronger.

If you can't find more powerful parts, parallelling, say four, five 1N4007's or two 1N5407's in your schem will yield a little bit sturdier PSU, but it isn't as reliable as a proper hi-current rectifier diode. It's quick and dirty, but can work ...

[Seljer]

Making a bipolar power supply without the center tap means that the two diode bridge only works as a half wave rectifier. If you had the center tap you could use the full four diode bridge for a bipolar supply with a full wave rectifier, which means you could get away with a much much smaller amount of filtering (smaller electrolytic caps) for the same amount of voltage ripple on the output.