audio-talk

Whahey! A new section to waffle away about amps and geetars n stuff. Ta Nick!

Right, we've (Jon (step son) and me) have been using this amp for about 12 years. Mostly it sits there just doing its thing in the study room but at one time Jon gave it a hard life gigging with bands when he was young before he'd bought himself a bigger (horrible! IMHO) SS Marshall combo.

This tiddler has a VALVE in it. Note use of the word 'A' as opposed to 'lots'.

Anyway, faithful old AVT20 started playing up a while back, intermittent jack socket, noisy controls, odd bursts of Radio Baghdad coming through, usual sort of stuff guitar amps do when they want to retire.

I opened it up for a service and a fettle and whilst doing that found R53 fried. http://www.diyaudio.com/forums/attachme ... _60_02.pdf



R53 is a 10R (1/2 watt) resistor is series with a .22uF 63 volt cap. I think this is a Zobel network to prevent the amp oscillating as it's connected across the o/p of the amp to grnd.


The amp was hammered regularly (I'm guilty too, it's not just Jon!!!) but little did we know there were 2 dry-joints on the terminal connections that feed the internal 10" speaker. So, I reckon that at some time the amp has had no load connected and fried the resistor due to the laod disappearing because of the dry joints (the 2 terminal pins were so dry and wobbly they actually fell out the through-holes in the PCB when i disconnected the internal speaker!!!

Here's a thing though. Search around on teh web and others have found this too wrt to AVT20's, thats how I found the schematic. But, no-one else has mentioned dry joints on the spreaker terminals, maybe they weren't looking perhaps. Either way, the conundrum is.... whats causing the resistor to burn up? Maybe the amp is oscillating above AF and taking out the Zobel resistor?

I dunno. To be fair Marshall have fitted a fair amount of ferrite around the (obligatory) ECC83. There's eben a tiny CMC across the i/p jack I'm guessing to prevent RF entering the 1st stages of amplification. I've never seen that before but TBH I haven't worked on that many guitar amps, I'd like to though.

Any ideas on the resistor bonfire?

DTB

If it's a hotty I'm wondering why it's not 'stood-off' the board like R34 and R35. The little cap sitting next to it is probably shaking with feeeear and could have made a mess in it's pants.

edit: I have thought for a couple of seconds longer, silly me, it's not supposed to be a hotty...it was provoked!

furthermore....did you play your bass through that??? I have a mic pre with the (obligatory?) ecc83 and I wonder what it actually brings to the party...I've always suspected not a lot.

I converted to ss on stage in the 80s after a chum dropped his HH 100watt head a couple of gigs in succession(after gig frollicks, you know how it is) and it still worked. Really havn't looked back.........

edit2: contentious bit removed....wrong side of the bed monday morning etc etc

Would you not upgrade it to a 2W resistor ?

I've got a couple of 2W 10R 5% metal films if you need them, can post today.

no but I don't understand the circuit. They have strange ways of doing things in the guitar world.

both sides of the ECC83 haven't got grid stoppers. I do see the one ferrite.

All the opamps amp chips and the fet or mosfet driving the stone tack ~I haven't a clue about.

Not much help really.

I reckon, most likely, it's something to do with your technique Dave

@ our-Ed, nah, I don't play bass through it, its just a likkle Guitar practise amp but I do soooo like it. I use an SWR Super RedHead hybrid for gigging, valve front end and SS power amp. Big and heavy all-valve Mesa-Boogie 'rig' was sold a good few years ago.

pre65 wrote:Would you not upgrade it to a 2W resistor ?

I've got a couple of 2W 10R 5% metal films if you need them, can post today.

Ta for the offer as always Phil, Sure, I could just drop in a heavier duty replacement but I'm feeling there's summat else going on here. Fitting a 2W'er would (probably!) stop the 'R' from from igniting but I reckon it's more benefit to get deeper into why it's happening. The network is in // with the speaker, whether internal or external, soooooooooooo... the cogs in the brain are thinking "ang on, thats a CR network, ergo it'll have a resonant frequency, ergo at that frequency it's gonna pass AC....and looking at that burn-baby-burn resistor there was a lot of current to do that", thats why I'm a-thinking the amp might be being pushed in oscillation. Having the speaker load disconnected probably doesn't help either due to the (now repaired) dry joints on the PCB next to the resistor! Those terminal pins are where the internal speaker connects.

Maybe the amp oscillates when the speaker is removed from the cct. Aha! that'd explain it.

Paul Barker wrote:no but I don't understand the circuit. They have strange ways of doing things in the guitar world.

both sides of the ECC83 haven't got grid stoppers. I do see the one ferrite.

All the opamps amp chips and the fet or mosfet driving the stone tack ~I haven't a clue about.

Not much help really.

I know what you mean about the way this amp is done Paul, 'm guessing the source-drive from the mosfet is low impedance current drive into the EQ section? Possblee? It does work well though, I like the range of sounds, the mid control has a lot of 'sculpting' to the sound.

richardcooper2k wrote:I reckon, most likely, it's something to do with your technique Dave

Whaaaaaaaaaaaaaaatttttttttttttttt! You mean its not OK to use RF transmitters as plectrums?!

DTB

Dave the bass wrote:Any ideas on the resistor bonfire?

DTB

It's a power amp chip innit? Also I'm guessing it's the R of a Zobel type RC snubber between output and ground? The little cap next door is wired in series? If so, a note, from memory:

"Do not run the amp without a speaker load connected else it may oscillate at RF (maybe couple of Megahertz), which will burn out resistor R'X'" (instructions, Velleman kit from Maplin, or could be TDA2050 datasheet, whichever)

HTH

PS: if the resistor was missing ~ which it effectively was if open-circuit ~ then yes you would get RF problems since it's that that provides some sort of a load for the amp to work into at high frequencies and thus keep it stable.

Aha! Yep, its a TDA 2050 Mike. Yes, cap and resistor in series betwixt chip amp o/p and 0V rail. Resistor still measures approx 10R, even in its charred state. I have a replacement in the resistor box.

Onwards! To the Riff-Kingdom...

DTB

New resistor fitted (on PCB pins this time to make future replacement easier) and I've been playing silly sausages with the amp and the scope and a battered Les Paul copy. Scope is connected across the resistor in the zobel network.

Wanna see an amp driven from overdrive into oscillation? Click the piccie below


...see how when it bursts into osc it upsets my camera too....yucky.

Took out another resistor, but on purpose this time. Smokey-dokey! At least I know how it happens now. Its the deadly (for the resistor anyway) combination of high gain and the Jons old Geetar as an antenna.

DTB

So Dave does this solid state carp in the output stage need protecting with a component which will render the amp inoperative if it is played?

A single ended 833a 100 watt amp wouldn't need any protection.

Paul Barker wrote: A single ended 833a 100 watt amp wouldn't need any protection.

Not, but others in the room might

You couldn't prove it's ruggedness by throwing it down a flight of stairs as VOX did with the AC30!

Paul Barker wrote:So Dave does this solid state carp in the output stage need protecting with a component which will render the amp inoperative if it is played?

A single ended 833a 100 watt amp wouldn't need any protection.

Carp on the output sounds fishy, arf, see what I did there?

For my next trick I'm going to play around with it and see if I can find out whats pushing it into oscillation. I think the 1st step is to find out if it's the pre or power stage. Also I'll try different leads and axes (geetars) to see if its just a certain combination of axe and pick ups causing a problem. Gut feeling that its not cos there's quite a few other folk out there in Internetland that've had the same problem. Some have blown the TDA2050 too.

Looking at the TDA2050 datasheet <pushes nerdy glasses up sweaty nose and blinks awkwardly...> shows lower values of R and higher values of C than Marshall have used. 2R2 and 0.47uF typically, it specifically states in the literature higher values of R can lead to oscillation..... Ooooooooo.

DTB

In all seriousness if it is only 35 watts is it not just crying out for an adapted output section to push pull KT88 or EL34?

You could sell kits.