Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 05:12:35 AM

Login with username, password and session length

Recent Posts

 

Old SS Project Rebuild

Started by Littlewyan, March 18, 2021, 05:35:49 AM

Previous topic - Next topic

Littlewyan

Hello all!

It's been a rather long time since I posted here, but recently I decided to get back into building Solid State amps. The last one I built was......well no other way to describe it, a death trap! So I've gutted it and am starting again applying all the lessons I've learnt building various valve amps, mainly in safety! I will post project photos etc to show progress but I need some help please with applying current feedback to my design.

You'll see in the attached screenshot that I'm basically using the standard TDA2050 (Ignore the fact the model says TDA2030) circuit from the datasheet, but am trying to add current feedback to it. As I'm using it in a Non Inverting configuration I'd have thought it should be applied to the voltage feedback loop, but this gives me positive feedback? Is this right?

Littlewyan

#1
I may have actually just solved this.....possibly. Found an old post from J M Fahey on another forum where he suggested this to someone. It seems to have the desired affect. The only issue I have now is working out what wattage resistor I need to use for the 0.33...

Loudthud

I think you have it right. The only problem I see is the diodes from the output to the rails should be 1N400x.

Littlewyan

Ah sorry, ignore those, rather embarrassingly I don't have those diodes as models in LTSpice yet....I'll get around to adding them one day :D. So yeah they'll be 1N4001 in the final build.

edvard

Quote from: Littlewyan on March 18, 2021, 06:37:05 AM
I may have actually just solved this.....possibly. Found an old post from J M Fahey on another forum where he suggested this to someone. It seems to have the desired affect. The only issue I have now is working out what wattage resistor I need to use for the 0.33...

Another reference: https://www.current-drive.info/projects/17

His is based on a TDA2040 with a 0.56Ω gain resistor at 2 watts.  For a 0.33Ω resistor in a slightly higher-powered chipamp powered by more supply volts, I'd go 3 watts at least, 5 watts for peace of mind. 

In LTSpice, you should be able to simply put your pointer over the resistor symbol and click when the cursor turns to the clamp-meter symbol to read Amperage.  If you do a Transient analysis, you'll see Amperage on one side, and click on the wire above it when the cursor looks like a meter lead to get Volts to show on the other side, then do the math.

I'd find a TDA2050 model first to get any kind of accuracy, and understand that Spice modeling assumes ideal conditions; err on the side of more watts.

Littlewyan

I decided to drop the Mixed Mode Feedback circuit in the end and just stuck with the standard Voltage Feedback circuit from the datasheet. I read a few articles (Written by teemuk I think...) whereby various feedback circuits were compared and I wasn't confident that I'd have room on my board for the appropriate sized resistors. Perhaps in a future project I'll attempt this again!

The amp is currently half built. The PSU and Power Amp boards are complete and working, I have successfully breadboarded the Pre Amp and have just started to put the board together for it. Pictures will be uploaded soon!

Littlewyan

Here are photos of the before. This was an amp that guys on this forum helped me build years ago. I knew practically nothing about electronics at the time, I barely knew ohms law, yet everyone was still very patient with me :). One in particular was Roly who I sadly understand is no longer with us. I would have loved to have shown him that his patience at the time was not wasted. I went back through those posts (I think back from 2013) and now with a better understanding was able to use the various equations and advice that he gave to me at the time to rebuild this amplifier into something better (And safer).

Littlewyan

Well the amp is built, I'm waiting for a new power switch to arrive, but for now I'm using a terminal block on the power wires. Unfortunately, I'm having weird oscillation issues! I've attached the schematic. Now, let me tell you a story!

First of all, I have tested the power amp by itself absolutely fine and then tested again later with the preamp on a breadboard. It worked ok, but had a lot of noise (expected as it's on a breadboard). I built the pre amp board, fitted it into the chassis and wired up, on first switch on the amp oscillated with the volume on 0. I measured frequencies all over the place, from as low as 500Khz to 2/3Mhz. I moved the volume pot ground from the preamp board to the power amp board and it went away, however when the volume was above 7 it would come back (This is without any signal on the input). On my scope this appeared as an almost perfect sine wave?! I could also see this oscillation on the ground points around the boards, however resistance between these points was no higher than 0.1ohms. Odd.

It gets weirder, when testing this morning, without changing anything the oscillation started happening all the time again. It's the power amp board, even isolated (Pre completely disconnected and PA input grounded) it oscillates and is no longer a perfect sine wave. I scoped the power rail and the oscillation signal is present and the same on the power rail as it is the speaker output. What on earth could this be?! The only change I made between breadboard testing and now is the feedback resistor on the PA board, I changed it from 22K to 18K, however it's still above 24dB gain so should be ok. I'm not sure that would cause the oscillation to be present on the power rail though. Any ideas anyone? I'm at a loss.

Littlewyan


Littlewyan

And here are scope pics of the oscillation I measured. Test12 is the speaker output and Test13 is the power rail.

Littlewyan

Quick update, turns out the TDA2050 has blown! Hence the funky stuff I found on the power rail. I'm going to order an LM1875 to replace it and see what happens. I don't think the chip blowing was the issue, I think it blew because of the issue.