Vox Cambridge 30 Twin

[SinfulXHope]

Hi, forgive me if i'm posting this in the wrong spot.

Just got an old vox 30 that I cosmetically restored recently, the controls seem to work fine it's just that the louder sounds causes it to break up/crackle. It's annoying and I wanted someone's opinions on the matter. What I'm aiming for is a clean sound when loud on channel 1. Thanks.

[J M Fahey]

Give it the Enzo Whack [tm] , as in hit it hard with your tightly closed fist (just don't break your hand) .

If you hear annoying crackle, something is lose inside causing a poor contact, playing loud just self-shakes the amp.
If so, post it for suggestions.

If not, maybe 30W is not enough for you (not surprising considering how loud can some drummers be) and above certain level amps start to distort..

[SinfulXHope]

Will do, also upon looking at photos of when I cleaned it up I just noticed that filter cap looks a little torn on top and leaking on the bottom. Should I replace the one or all PS caps? It'll be a while before I can buy replacements.

[J M Fahey]

Thanks for posting the pictures which clear a lot of doubts.
Sorry but:
1) that's an average quality ~25W SS amp, driving a couple not too efficient speakers ... you don't simply have "muscle" there to sound loud and clean.

Those are Celastion and frames are painted blue ... but magnets are the smallest available.
Not a bad sounding speaker at all ... but not an AC30 with "Celestion Alnico Blue" either.
Notice that even the Tube version has a beautiful gold painted frame, a gorgeous back cover ... and a very small magnet.


2) the power amp is a TDA2050 chipamp, what's used all over the world for around 25W amps, very good integrated circuit but it gives what it gives and no more


3) those caps look fine.
Tops are notched so *if*  for any reason (plugging 110V into 220V mains) they explode, tops break open at the seams and don't cause that much damege.
They are not leaking but that's some glue applied to damp vibration.

4) now I believe your amp is fine, simply not loud enough.
And overdriving a chipamp is not nice as overdriving power tubes but quite harsh ... what you hear and don't like.

Personally, i'd use it as is, it's a fine sounding amp (just not *loud*) , and mike it up at larger venues.

[Roly]

Enlighten me - what on earth is Q1 and R17 doing in series with LED4?   :loco

[SinfulXHope]

Even at less than 12 o'clock on the volume, maybe around 9-10 it's crackling up. Is that normal then?

[J M Fahey]

Maybe, only way to be certain is actually measuring power output.
If it reaches at least 20/25W "it's not broken" , just doing what I suspect and described above; if it puts out, say, 5 or 10W then there is some kind of problem.

There is different ways to measure power, being in a faraway Country I have tried them all, one time or another, if I didn't have access to a proper Lab.

different options:

1) If you have access to an oscilloscope, that's the Pro way to do it.

If not, but you have a PC, you can download a software scope which uses the PC sound card (line or mic)  inputs and displays the waveform on screen ... which to boot you can upload here.

You'll need to build a simple adapter so speaker out properly matches PC input without damage.

2) or you can build a simple peak to peak detector (2 diodes and 2 capacitors) which gives us a reasonably accurate power measure ( shows 10/20% higher than real power, because it counts supply ripple as "good" signal but no big deal)

3) or you can inject 400Hz clean sinewave and *sloooowwwlllyyyyy* rise volume until you hear it start buzzing, then lower slowly a little until it disappears, repeating up down 2 or 3 times you can catch the edge of distortion almost as well as onscreen, no kidding.

To boot, the sharp chipamp distortion combined to buzzy speakers (I suspect those are Celestion 10G25 , google them, very bright and piercing) makes it esier to catch the clipping point.

Then you measure actual power.

Just to save some time, pklease measure +V and -V rail voltages , you can ***carefully*** measure on TDA2050 pins 3 and 5  but be aware that if multimeter test pin slips and touches 2 pins at once it will destroy the chipamp.

Or to be safer, measure at the ends of D3 and D4, the points called +MV and -MV , I expect around 20V there

[SinfulXHope]

Thanks, Can you give me procedures for number 2?
Never had to measure power output on an amp before. I have a lot of components lying around and a digital multimeter.

Few questions to help me get started. What does the peak to peak circuit look like? I know the single diode single cap circuit.
Am I connecting the output wires that go to the speakers to the circuit? Once the caps have charged is it just a matter of reading the voltage across them?
Lastly I assume I use P = V^2/R, R being my "warm" speaker resistance?

[J M Fahey]

This is the basic circuit, was very popular with VTVM (vacuum tube voltmeters) , I used it for ages before I got my first scope and still kludge it for rough tests away from home/lab when some problem catches me by surprise:

those values were for RF/TV testing, in fact many weird pulses (sync and similar) were stated on the schematic (SAMS Photofacts, etc.) as "xxx VPP" , period.

For amp testing you can use 1uF x 100V electrolytic caps, 2 x 1N4002 to 1N4007 diodes.
The right cap will charge to output peak to peak value, Vpp which you measure with your voltmeter; RMS is Vpp/2.83 .

Inject a 400Hz tone and rise it until it starts to sound funny , it will be buzzy and you'll also hear some hum mixed with the signal , which is the ripple riding on the rail voltages.

This is what you are measuring: Vpp and calculating Vrms based on that.


Why not measure plain AC voltage and just be done with it?

Because we are not good judges of distortion (although that's the base of Method 3) and when clipping, meter indication will rise more and more ... which is true because Average or RMS voltage increases with distortion, cheating us, while peak voltage will not, even if it distorts, because that's the essence of clipping, not being able to go beyond rail voltage 

An amp just reaching clipping and a heavily overdriven one will show the exact same peak to peak voltage, so accurate detection of clipping is not needed.

Beautiful, isn't it?  :dbtu:

Only problem is that the detection probe circuit is so brutally honest, that it will also measure ripple voltage which is "riding" on top of the audio wave.

So in practice, if you measure, say, 38Vpp , you should substract, say, 4 V before converting to RMS.

And I suggest 400Hz because it's easily available as an MP3 and speakers are close to rated impedance around there, so measurement is meaningful.
http://www.mediacollege.com/audio/tone/download/
Set MP3 player (or Winamp)  to Loop/Repeat so tone is continuous.

If you measure by playing a guitar *loud* , narrow peaks (pick attack) will also give you higher than normal values, because amps can give brief bursts of high power, but we want a more realistic test.

Good luck

[SinfulXHope]

Well I checked and I was getting around 12Vpp. I've been doing some pedal works and even replaced the preamp tube just for a try, I did notice that only the low frequencies were causing it to crackle. I could hit high notes all day, but when I play an on E it would do it.

-BH