General Question - Chip Amp as a Clean Guitar Amp

[darklife]

You may want to check out my thread on the same topic I made not long before you...
http://www.diyaudio.com/forums/chip-amps/269056-has-anyone-used-lm3875-lm3886-guitar-amplifier.html

Working on a similar project, only I am also building my own preamp using discrete low noise transistors. Using a single NPN stage at the input jack to the 3 band tone stack based on the old familiar Fender circuit, then from there going to another NPN stage to bring it to line level and proper phase to feed a LM3875.

So far my project is coming along nicely. I may have to add another stage of gain at the front to add in a second channel, that way I have clean and fuzz inputs switchable.

So far the biggest issue I am having is stabilizing the LM3875.
The chip works great by itself with line level input, but when adding a sensitive preamp to it to get guitar level up to line-level that it wants I have some issues with the project oscillating/squealing on full volume and high tone settings, but I almost have the kinks worked out and will update that thread in the future.

If you just want to run line output from a pedal to your chipamp it should work flawless.
You may need to add some RFI filtering so that powerful radio stations and such don't get into your amp either through your guitar cable or from the line-in to your chipamp.

If you just want to take guitar output and run it to your chipamp you are not going to get a lot of amplification, depending on how loud you can crank your "level" output on your pedal.

All that will take experimenting.
Best of luck!

 

[cpsmusic]

Would the following be suitable as a very simple, clean preamp for a 3xTDA7293 module?

http://www.muzique.com/schem/mosfet.htm

Cheers,

Chris

 

[bentsnake]

So far the biggest issue I am having is stabilizing the LM3875.

The chip works great by itself with line level input, but when adding a sensitive preamp to it to get guitar level up to line-level that it wants I have some issues with the project oscillating/squealing on full volume and high tone settings...

An amp is stable until a preamp is plugged in, then oscillation occurs. This doesn't exactly scream "unstable LM3875." On its face, it sounds more like the preamp is oscillating--or some connection is causing oscillation--and sending the signal on to the LM3875.

In any case, be aware that oscillation can occur in the ultrasonic range, so you can't hear it. Then something changes and you can hear it.

The check for this, lacking a $2,000 scope, is to put your finger on different components. Be careful, along with the danger of electrocution, an oscillating component can get seriously hot.

It's possible for a bad solder (or other) joint to act as a rectifier, as in the cat's whisker radio days, making high or radio frequencies audible and/or causing oscillation. Solution: for solder joints, use additional flux, and don't let components move until the solder has cooled. For screw connections use a serrated star washer.

Some components in the circuit I've posted below (which is otherwise of no significance whatsoever) might be of interest:

R1 and C1 act together to form a low-pass filter that shunts very high frequencies to ground. This can help prevent oscillation.

C3 acts to shunt high frequencies around R3. The result is more feedback for these frequencies, which means less amplification, which can help prevent oscillation.

Strictly speaking C3 should be matched to R3, so the shunt effect comes in around 100kHz. This sounds like a high frequency, but it's only a bit more than 2 dectaves (as opposed to octaves) above audible frequencies. The idea is to preserve overtones, which can be felt even when they're not heard (assuming very expensive speakers).

Lacking the facilities to match C3 to R3, 10pF is a generally workable number.

Just some guesswork.
.

 

[darklife]

An amp is stable until a preamp is plugged in, then oscillation occurs. This doesn't exactly scream "unstable LM3875." On its face, it sounds more like the preamp is oscillating--or some connection is causing oscillation--and sending the signal on to the LM3875.

In any case, be aware that oscillation can occur in the ultrasonic range, so you can't hear it. Then something changes and you can hear it.

The check for this, lacking a $2,000 scope, is to put your finger on different components. Be careful, along with the danger of electrocution, an oscillating component can get seriously hot.

It's possible for a bad solder (or other) joint to act as a rectifier, as in the cat's whisker radio days, making high or radio frequencies audible and/or causing oscillation. Solution: for solder joints, use additional flux, and don't let components move until the solder has cooled. For screw connections use a serrated star washer.

Some components in the circuit I've posted below (which is otherwise of no significance whatsoever) might be of interest:

R1 and C1 act together to form a low-pass filter that shunts very high frequencies to ground. This can help prevent oscillation.

C3 acts to shunt high frequencies around R3. The result is more feedback for these frequencies, which means less amplification, which can help prevent oscillation.

Strictly speaking C3 should be matched to R3, so the shunt effect comes in around 100kHz. This sounds like a high frequency, but it's only a bit more than 2 dectaves (as opposed to octaves) above audible frequencies. The idea is to preserve overtones, which can be felt even when they're not heard (assuming very expensive speakers).

Lacking the facilities to match C3 to R3, 10pF is a generally workable number.

Just some guesswork.
.

In my case, or design rather, it was just a matter of adding some negative feedback to the transistor preamp stages using a capacitor across the collector and base of the transistors.
It seems to have solved most of the issues. Obviously shielding helps a ton also.
The larger issue is with the actual speaker output being near the preamp inputs since they are on the same board. I isolated them by using some shielding and an RF choke on the speaker output.
Oh and that 2k$ scope... lol.. seriously? Any decent 20$ scope will do just fine for most audio work. I have an old Eico 460 tube scope that has a top of 5MHz bandwidth and it is perfectly fine for audio work.
Not to be an *** but when it comes to audio work you don't really need a great scope, you just need one that works with a proper visual grid that is aligned 😀

BTW, a cat is a good ultrasonic detector. I learned early on that if I touch something and it makes a funny noise but not loud, yet my cat freaks out and runs out of the room, it is probably above my hearing and I need to connect that scope to check.