Serious Blocking Distortion

[stormbringer]

Cool.

Actually the overdrive level might be volume related, have you made an audio probe? it's usually very helpful in finding out where the signal weakens.

The 220k Resistor does lower the input impedance, and might be what lowers the volume.

[Littlewyan]

How would i use an audio probe? Sorry i've not used one before.

[Roly]

One fifth the input resistance won't help at all.

8 or 9 watts is actually pretty damn loud - into a reasonable speaker.  So the obvious question is, what speaker are you using?

What impedance (ohms) is it?  If it's 16 ohms I'd expect it to sound a bit weak.

If you are using a typical Hi-Fi type speaker with an efficiency of around 80dB/W then I would expect it to sound quite gutless, but if you are using a modern guitar speaker in a reasonable enclosure with an efficiency of 95-100dB/W, then you have some sort of problem.

You won't hear the difference between 24 and 28V supply.

If you are using a passive guitar (one without a battery inside) then changing the input resistance from 220k to 1Meg will improve the frequency response and make it sound richer, as well as giving you more raw signal level.

[stormbringer]

that is a simple, but awesome tool. you just connect it through any audio Equipment with a sound input, i usually use it through my Zoom R16 Sound card/portable studio, as it has LED VU-meters which will show me the signal level aswell as letting me hear the sound.

But also, i Think replacing that 220k resistor with a 1M or higher will be a good thing to test.

[Littlewyan]

I'm testing it through a Marshall 2 x 12 8ohm cab so the speakers are fine. I'm definitely not getting 8/9watts out of it at the moment, although i dont think its far off to be honest. I'll change that gate resistor tonight and will then do some proper testing.

[Littlewyan]

Ok gate resistor is now 1M, this has helped however the output still isn't as good as I would like it to be, also the sound wasn't great. So I upped the voltage coming from the PSU to 26 volts and it started to sound better tone wise. Now I measured the output voltage and it reached 8 volts at the most, which divided by 8ohms (speaker cab) gives us 1 amp, multiplied by 8 gives us 8 watts so maybe I should increase the voltage further to get a bit more volume? I know this is a crude way of measuring it but its the best I can do with the tools I have.

Also when I turn down the 100k pot even just a tiny bit from max the sound starts distorting badly, I'm guessing a larger pot is required? Turning down the 10K pot does what it should and just decreases the volume.

Also the distortion is a bit harsh so perhaps I should try different diodes on the BJT? Its also slightly too bassy so I'm going to try decreasing the resistor in series with the input to help with that.

Apart from that the blaady thing works! Can't thank you guys enough for all the help you given me, learnt so much.

[stormbringer]

I Think you got the 100k pot reversed if turning it down creates distortion. You could absolutely try different diodes, (Try leds, diodes in series etc). Nice to see you got it up and running! Good job.

[Littlewyan]

Well no its correct because the volume does decrease, it just seems to distort even more. Very weird. Also I don't think the preamp is driving the power amp hard enough, as after the 1.5K resistor I've only got 18volts, whereas I designed the preamp to be run on 24volts. So may change the collector resistors for slightly lower values. Also I'll probably remove the 22k resistor just above the two diodes as its doing nothing

[Littlewyan]

Also when I say distort its almost unrecognizable, as if a transistor is just being pushed stupidly hard.

[Littlewyan]

Ok firstly I changed the collector resistor on the BJT from 1.8K (all I have at the moment) to 1K and this just caused the same distortion effect I get when I turn down the 100K pot. So I changed it back to 1.8K and all is well again (except for the pot issue). I then changed the collector resistor on the JFET from 27K to 22K and wow what a difference. Plenty of overdrive which sounds great for single notes but not too great for chords. Now I'm getting 12.5 watts out of it which sounds great! Next thing to try is removing the two diodes on the BJT to see what effect this has, may even fix the pot issue I'm guessing?

[stormbringer]

Well, you changed the bias for the jfet, the question is if it clips now adding overdrive, as it is an input stage, i Think you would want it to stay Clean, but if you like the overdriven sound, maybe replace the bjt/diode stage with Another jfet stage.

Problem with overdriving the first stage and removing the clipping diodes will be that the 100k pot will not be a gain pot anymore, but Another volume pot. But as i said, maybe the jfet distortion is the sound you are looking for, and in that case you could just replace the clipping stage with a jfet instead.

You should probably check the level of the sound coming out of the jfet, and listen to it with the audio probe. Maybe the level is way too high, and hence the diodes clip no matter what level you turn the gain pot to? Or maybe the Bias on the BJT is incorrect? what's the DC level of the input/output signal?

[Littlewyan]

You are right with regards to the 100K pot becoming a volume control, however now that I've removed the diodes the overdrive sounds great. Only quarrel I have with it is the amount of bass at lower volumes, so I was thinking of removing the 100K pot and then putting my tone control back in (only have two holes for ports on the case). A gain control would be nicer obviously but can't work out why that 100K pot wouldn't work properly.

[Littlewyan]

Added a tone control in and now it sounds pretty sweet at all volumes, I'll upload an updated schematic another day as its 22:30 here. I'm going to give the amp a proper test at band practice tomorrow night before I say its finished, may not be loud enough yet. I want to use this as my backup amp at gigs

[Roly]

If you have a supply of 24 volts then you have a peak swing of +/-12V or an RMS load voltage of 0.707 * 12 = 8.484V.

P = E²/R

so roughly

8*8 = 64 /8 = 8 watts

Gain of the TDA2050 is about 22k/680r = 22/0.68 = 32 times

If output = 12Vpk then input will be 12/32 = 0.375 or 375mVpk

The main cause of the bassy response is the low value of the tone pot, 5k.  I changed it to 100k 'tho it may not need to go anything like that high, perhaps 20k to 50k, but really should be about 10x the reactance of the tone cap at 500Hz (~3k ohms).

The transistor stage is drawing much more current than it needs to because of the low resistor values.  The Collector and Emitter resistors are increased x10 while the bias resistor is increased to keep the collector voltage about half the supply.  This all has the incidental effect of increasing the input impedance and reducing the loading on the FET stage, increasing signal levels somewhat.

With the current drain of this stage much reduced the preamp supply rail goes much higher.  This allows more headroom and also shifts the bias of the FET stage to a more reasonable point, about 10 volts or again about half the supply.

Reducing the FET Source bypass from 22uF to 1uF has the effect of giving some bass rolloff below about 500Hz (instead of the current bass boost), again brightening the response.

The first stage is still clipping at a fairly low 300mV input, and part of the reason for this is that the first control isn't actually a Gain control but simply another volume attenuator.  If the 4k7 Source resistor was changed to a 5k pot (now surplus from the tone control) and the 1uF connected from wiper to ground, this then really would be a variable gain stage and could be turned down so it didn't clip even with volts of input, or up so it clipped easily.

HTH

[Littlewyan]

Thanks for that Roly, I've currently got the amp working using the design I've attached to this post. It works quite well but obviously I don't have a tone control. I'm taking the amp to band practice with me tonight to test it and will know then if I definitely like the sound. Although I may adopt your design anyway!

With regards to having the 4.7K Source Resistor as a 5K Pot, isn't this bad?

O by the way since I put the tone control back in the bass response as gone down slightly which is nice, but I won't know for sure until tonight when I can actually give the amp a proper going through. Also I used a 1uF Solid Tantulum in the end as the 100N Cap I tried didn't make any difference whatsoever to the tone which is odd, unless its because I'm using a 5K Pot!