Studiomaster 16-4-8 latest on 14 October 2008

Gonna get on with the mixer tomorrow
Im just gonna plug in one channel to stip, but do i still need to connect the monitor channel also which has a ground strip to earth?  (this would usually be the last channel in the chain of multiple strip)

thanks paul

20 comments

  1. Now when i say noise, i mean white noise(like the old TVs when you didnt tune the channel properly) when i turn the gain the noise gets louder gradually, so the gain pot is working fine(now, it didnt befiore servisol)
    thanks paul

  2. Er..Well it would, wouldn’t it? It’s the transistor output after amplification.

    Have you applied some heat/extra solder to the joint? If not, do so, then test again. It’s a good test to listen as you solder. It may give you a few more clues.

    I notice that there is some board marking or something so it looks like work has been done in that area previously. Just south east of your mark. It’s the thing you get when you spot and then clean up shorts between tracks.
    Hold it up to a strong light. IF no connection is supposed to be made, there should be no trace of metal in that space.

    Check the gain resistors and the blocking capacitor next to the gain control. It could be, that all you are seeing is the effect of EXCESS amplification. The gain is related to resistor ratios, so if one is duff it’ll muck the gain up. The capacitor may be passing DC and as such, it will affect the transistor bias current which in turn is actually the thing that sets the gain.

    Is the gain control changing the gain smoothly when you turn it?

    Look at all that above BEFORE pulling things from the board. You can’t do this too much without ruining the tracks and then it’s hard work… the reason is that in my experience, transistors usually work or they don’t.
    They are so easy to break. So if it’s half-working, it’s an effect from somewhere else.

  3. Thats madness, at least you tracked em down and exposed em wow, that earpiece thing is the dogs bollocks i founf where the noise for both channel 15 & 16 is originating from it seems to be the collector of the first transistor on the line in(see pics, ive included schematic and transistor specs)) does that mean i have to replace them(obviously) with same or better todays equivalent?
    thanks
    paul

  4. cool awesome, thanks
    got my hacksaw ready, just kidding
    that makes a lot of sense 1/2 then 1/2
    im gonna have a go this afternoon
    ill report my findings later
    thanks again
    paul

  5. Good idea, felt tip pens! Whatever works. But see my last post as well to speed up your work. You need to identify where you are on the board with where it is in the circuit. Try and test a mid way point to see if the problem is there. You automatically can discard testing half the circuit then! Repeat on the half left until you hit then dud. Then fix it.

    Imagine the circuit is a long insulated wire but the wire is broken
    somewhere inside the insulation.
    Now someone says, “find and fix the break using a meter, soldering iron and some solder, but cut the wire as little as possible”
    So you cut it in half and see which half conducts okay – therefore the
    break is in the other half. And so forth.
    That’s the way to fix the circuit!!! (Obviously, you don’t want to cut
    the circuit – it’s the principle)

    You do the same thing but listening out for good or bad sounds!

  6. so, i basically check every component and solder joint from point a – z in series?

    this is my first attempt an a visual aid for tracing signal red is hot blue is cold yellow are components thinking now though i should have different colors for resistors, capacitors, transistors and op amps also i should probably color earth green what do you think, does it seem right?

  7. You’ve followed one of the line inputs in red. It’s balanced so the other part goes to the other transistor. (that’s why “tape” is a double-poleswitch) What happens is, if you use a unbalanced jack plug, it shorts the other input to ground at the socket. Look closely on the socket and you’ll see as you push the jack in and out. The switch just connects in a resistor network that effectively changes the gain of the two transistor amp.
    There’s a name for the back-to-back transistor configuration, but I’ve forgotten it.

    First, is it hissy on the line input? If yes, then leave the mic part for now. Looking at your red line, in between the jack socket and the base electrode of the transistor there are about 10-12 solder joints! Remember, each resistor/capacitor component has two solder connections to the board.
    You’ve also got solder connections for the tape switch and where the jack enters the board.

    So just trace through and check the sound at each point. I take it you’ve cleaned the switch?
    After the sound has been amplified by the two input transistors it goes to the +ve (non-inverting) input of the op-amp. This has a bit of negative feedback into the two transistor amp. After that, it’s an unbalanced signal onwards to the EQ stages.

    Good luck. There are millions of joints! You could do with a little sticky tab to mark where you’re up to on the board.
    Trust me. This is very straightforward and it’s a standard design. It just needs good spatial awareness and patience.

    Oh. And lots of spare time. Ha Ha!

  8. cool man thanks
    we’ll continue later im off for a walk to clear me head ive attached some files for your eyes email later if you get chance i know its like electronics by numbers but your my only help once i figure out one channel the other 15 will be easy!
    apart from busses, master and headphones thanks paul

  9. The green stuff is varnish, usually shellac or something, that protects the copper from oxidation. You shouldn’t need to break it. Just test at the solder joints or on the component legs which are usually nickel plated or nickel. So either top or bottom of the board.
    You’ll have to keep flipping.

    Jack plug tips are called ‘hot’ and the ring is ‘cold’. This means signal and earth to normal people like me! Trace it through USING THE CIRCUIT DIAGRAM AS REFERENCE! You need good spatial awareness for this – not many women can do it! It’s true, blokes are better. The reason is you are comparing an abstract diagram to a physical reality. Also, sometimes you will be on the top of the board and sometimes the bottom, so it does your head in and you can easily get confused.
    You’ll also need proper pin-out diagrams for ALL your semiconductors – chips, transistors and diodes, that exist on your board. In other words:

    The phrase, ‘ if I follow the same tip signal it goes to the right hand side of the second transistor is that right?’ – is meaningless for me. I’m at home now, miles from anywhere, I don’t have the board in front of me, I don’t know the transistor type, I don’t have the circuit diagram of the board etc. You see what I mean? You need ALL info to hand because you will be constantly referring to it.

    However, saying all that, if you can’t find the pin-out for a transistor, with a multi-meter measure the voltage across each pair of the tree pins.Now, assuming the part is working correctly (and this isn’t guaranteed), two of the pins will have about 0.5V to 0.6V across them, either +ve or -ve depending on which way you have the probes. These are the base and emitter which means the third pin is the collector.

    But really you need the pin-outs. They are widely available on the web and the Maplin catalogue used to have them all in the back. How else would I remember that pins 7 & 4 are the power supplies on a 741? I looked it up enough times!

    You’ll also need good eyes. Some of the lettering on the components is really small…

    Make sure you know the resistor colour codes. This is really helpful in identifying where you are as once you know the codes, you can read the circuit and match it to reality a lot easier (except if you’re colourblind!) No hunting for little letters…From memory they went black, brown, red, orange, yellow, green etc so a
    10k resistor = brown-black-red
    47k = yellow-violet-orange
    220 = red-red-brown

  10. cooll, tankhs fro tath
    lol
    now on the back of the pcb, its lightly covered in a thin layer of green stuff will the signal sill go through this?
    or do i have to push probe through it?
    Next hurdle
    tracing the signal?
    i think the signal comes off the tip
    managed to trace that on the pcb from pot, resistor, cap, resistor,line switch, ?
    now after the tape switch if i follow the same tip signal it goes to the right hand side of the second transistor is that right?

  11. Studiotoe!

    I wouldn’t have bothered buying a probe. I made my own from a bit of stiff (say 2mm) wire, insulated. Then I bent the exposed end a bit to get round awkward corners! More like a dentist’s probe I guess. Gaffer tape (the musician’s friend!) for insulation and a handgrip…
    Anyway, snip the earpiece jack off and replace it with a dinky crocodile clip and your probe on the two wires. The joins will need insulating and supporting as the earpiece wire is really thin. So loop it back so you have a bit of slack and gaffer tape it to the handle and itself for mechanical support.

    What you now have is a highly sensitive, AC testing device for audio frequencies! What’s more, the testing device (your ears) knows exactly what to look for! It doesn’t need relative stuff from graphs or anything. You already know what your problem is and what it should sound like if there isn’t a problem. Now go for it!
    Hopefully, you got the crystal one. This is high impedance (>1 mega Ohm) so will not cause a short across PN junctions (this is bad, very bad!). You could test it’s impedance with your multi-meter to be sure as physically, there’s not much difference between the two types.

    Just get your two connections and prod them where you want! Be careful not to short out any components or wires with your probe ends. I usually just left 1 or 2 mm exposed on the end of my “custom” probe!

    Because it’s an op-amp based circuit, the crocodile you can attach to earth or 0 volts. The signal swings around this point. Then touch the probe on the back of the channel input jack and you should hear your signal (sound) source. I usually used a tape for the source as it’s a bit draining on the ears listening to a 1kHz test tone for longer than a minute…
    Now trace the sound as it goes through the channel stage, all the way to the output. You’ll spot any oddities. As you are not testing for continuity this time, it’s the character of the sound that you want to be listening to.The volume will change quite a bit at various points. (Testing for continuity is the same, but you are looking for the point in the signal path where it breaks).

    If you can’t find anything, take the input source out and turn the gain right up (or you’ll go deaf). Now trace the same way as before. At some point after the gain control, it’ll go loud (obviously!) and hissy. It’s the quantity and character you are after.
    Depending what the problem is, you have several options afterwards. It may be dry joint or duff capacitor or leaky resistor or noisy transistor.

    BTW, be very careful with your spellings and stuff on your replies as sometimes I have a job figuring out what you mean. I could give you some totally wrong info in future if I don’t understand or interpret incorrectly, which, now you are getting stuck in, could be a bit dodgy. Apart from that,

    Happy Prodding!

  12. hey man, picked up that crystal earpiece, a croc clip and this probe https://www.maplin.co.uk/?U=strat15&C=SO&doy=15m10&ModuleNo=5664&
    as the small ones are now discontinued
    this will still work though. right?
    i just cut the wire and solder it to the earpiece cable and the croc to the other then?
    croc goes to………….
    probe follows line input + along trace to each component until noise appears, right?
    earpiece goes in ear
    i jump up and down with excitement cause i found the problem, or something like that
    thanks paul

  13. Yeah. It’s like a car connection with the blade just pushing through
    the cable insulation.
    Get a teeny screwdriver and try and push all the wires in every plastic connection down (or up and then down) a bit to try and make a good metal to metal connection. The distance the wires will move will be fractions of a mm but may be enough to make that good connection
    you need. It’s important as some of these are signal wires, and you know what I’ve said about the diode like properties of oxide connections!
    You won’t be able to see any change visually I reckon, but when you’re done you may notice something aurally.

  14. heres a closer pic
    yeah there just pushed into some plastic stuff which has a meatl connectoe that then goes over the pins on each board i though this wasnt too good myself ive tried to look for another way of making it more robust with thicker wire but havent come up with anything yet im off to bed now as i have an interview tomorrow so ill check your return email about 2pm tomorrow thanks and have a good evening paul

  15. No it doesn’t. If all you need is an earth then just run one from the board to the PSU 0V connection or somewhere with a wire and two crocodile clips or something.
    I thought those green wires were part of the test – I didn’t realise that they are part of the gear. They look like an afterthought to me…
    Also, I think you’ll find that the multi-cable has a couple of OV in it anyway.
    Also, you should be able to test any module in a thing like this. It’s designed for that sort of servicing. So I reckon you should be able to disconnect all boards except one from the PSU/busses and maintain through voltage and earth connections back to the PSU. There’s no point in having nice busses and connectors otherwise!

    Also, looking at the looms, how are the wires attached to the plastic board connectors? Are they pushed into a V shaped bit of metal like a phone connection? If so, you’ll probably find some bother there. I would have hoped to see some solder but I can’t in the pix!

    Strangely Perfect

  16. the problem is, that the channel has to be connected to the busses/master/headphone to complete earth(see incl pics)
    pic) bus wire config 2 – top wires are carrying various power and signals
    pic) 2534 – if you look at the top left, has 2 green wires connected to chassis, one goes to the headphone channel and the other connects to the earth on power switch for PSU. so i would imagine they have to be connctec to channel 16 to complete the circuit and for safety reasons this thwarts my idea of just being able to connect 1 channel outside of mixer with just multi cable?
    any ideas?
    thanks
    paul

  17. Have you got a signal tracing earpiece yet? If so, you should be able to put a line level into the input and then, at unity gain, trace the signal through the channel. If it’s as bad as you say, you should be able to find where the hiss starts. Remember, the circuit is the key, but it’s a topological map, like the London Underground map, so the
    connections are correct but the actual distances between points aren’t real representations of either the board layout or relative distances or positioning.
    When you find where the hiss starts, you’re nearly there!
    Rees

  18. hey man, so put in channel 16, had to connect to 4 busses, masters and phones channel beacause thats the only way of having a ground connected the servisol has definetely stopped the problem with gain pot and high shelf BUT Im still getting white noise across the whole frequency band, im also getting a peak at 50hz, 150hz, 250hz, 450hz, 550hz and 650 hz a descending gain though, really interesting any ideas?
    thanks paul

  19. I’m not sure. What I’d do is eliminate as many variables as possible and expand the checking from there. It depends on the circuit and how it’s wired together. Remember, even a pre-amp stage is actually several transistor stages in series – but it’s a bit hard to separate them as they are all on the one circuit board. It’s possible to test a transistor at a time but you’d need to split the tracks and then re-solder, which is obviously impractical and self-defeating and would
    lead to more errors than you started with.
    So just do one channel at a time without any following stages connected if possible.
    Same goes for the output stages. You could do them first! Check them
    without anything connected to their fronts except your signal generator (of whatever type). If the output stages check out okay, you can then use them to check the outputs from the inputs because you know they are good. Do the inputs one at a time, checking, fixing or ditching as you go.

Comments are closed.