TS808, working... but with problems

[FIŠ]

Hi all, first of all as this is my first post I would like to introduce my self.
I am a 20 years old EE student from Croatia who is into building DIY stuf, I have been reading this forum for some time now but one unfortunate event actually made me register here and write this post 

So, I built this TS 808 clone, not the first stompbox I made but most complex I would say.
Decided to go with the schematic with 3PDT switch from general guitar gadgets (http://www.generalguitargadgets.com/pdf/ggg_its8_lo_no_clipper_sw.pdf?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a). I was building it as a prototype (to see if i like the sound/overall difficulty and schematic itself) on a standard pre-drilled PCB

pictures:

http://a7.sphotos.ak.fbcdn.net/hphotos-ak-ash3/s720x720/522142_3879741187785_66534907_n.jpg

http://a3.sphotos.ak.fbcdn.net/hphotos-ak-ash4/s720x720/392947_3879738947729_1935675317_n.jpg

Turned it on and it worked, but as i played it i started to notice some noise commin outta it, even with the guitar turned down, note that that "noise" wouldnt change with turning up volume on guitar, instead it is constant "crackling" equally loud whatever I do besides the obvious - turning up pedal and amp volumes

The noise i hear is almost identical to the noise made by blowing air into a microphone/telephone and it is not constant nor peropdical, It appears just like that for no appearant reason and dissapears moments after 

I have measured all the voltages on the IC's and transistors and they are acceptable

IC pins :                           Q1,Q2:                 
p1, p2, p3 ~4,5 V                          C~9V
p 4=0V                                         B~4V
p5,p6,p7~4,5 V                             E~3V
p8~9 V



I tried to record the sound of it but it is kinda quiet but at the beginning and the end of the recording you can hear that noise well but because my lousy microphone you hear it as a crackling sound instead of "blowing air" sound but trust me

record: http://vocaroo.com/i/s0Wfgu988f9s

I have replaced input cap, all the semiconductors but "D5" diode (according to schematics linked above) and electrolytic caps. Nothing has changed 

So please help me out here as I will be building another one soon , this time on a self-etched design and i don't wanna have this thing bothering me if it is gonna turn out good or bad.

[Paul Marossy]

Look for a bad solder joint. You might have a cold solder joint somewhere that's causing an intermittent poor connection.

[FIŠ]

I believe i have resoldered everything already :/, but ok ill do it again, any other advice?

[CurtisWCole]

I would definitely make an audio probe and see if you can find exactly where this noise is coming from.

Curtis

[chromesphere]

Also check around the pot wires on the PCB end.  Make sure there are no stray strands shorting out against each other.
Paul

[Paul Marossy]

I would definitely make an audio probe and see if you can find exactly where this noise is coming from.

Yeah, that might be the only way to figure out this one.

[R.G.]

You have a real advantage in being an EE student. You probably have access to the laboratories and therefore oscilloscopes in them, as well as perhaps lab instructors who might be willing to help you. This is a real world problem that will be an advantage to your learning to solve.  The EE curriculums no longer teach much practical work, and this will amount to a super self-guided learning project.

As a hint, intermittent problems always involve some part, component, joint, or wire which suddenly changes state to cause it to start/stop. Almost the entire trick in solving intermittents is to find out how to do something - anything! - that makes it happen. Then you know what to look at for the real cause.

I have had intermittent problems I caused myself by bending leads too sharply and cracking the place where the lead joins the body of the part. That took me a looooong time to find.

[Paul Marossy]

I have had intermittent problems I caused myself by bending leads too sharply and cracking the place where the lead joins the body of the part. That took me a looooong time to find.

That's a good point. That can happen easily on capacitors where you have to bend the legs outwards to get it to fit in the holes they need to go thru. If it were a tube amp, this could cause a loud popping noise because it arcs. Been there done that (a repair of an old Fender Twin Reverb).

[PRR]

> advantage in being an EE student

I'll take a different position. Though actually agreeing with " EE curriculums no longer teach much practical work".

There's stuff they don't teach EEs. Soldering is one. It's a real Craft. Teaching soldering takes away time for teaching important EE concepts. Professional EEs will have technicians to do the real work. And some of the EE students I knew should not be trusted with a soldering iron.

> I believe i have resoldered everything already :/

Don't "re-solder". Solder it right!

As mentioned elsewhere: I had a sick air-compresser. I put new valves in it. Air came OUT the IN hole! So I took it apart and put the valve in again. Air didn't go at all. So I actually *looked* at where air was supposed to go in and out and through. Now it works good.

Glopping solder on a tarnished lead don't give a good steady contact. Re-glopping the solder isn't the best answer. You want clean leads, you want to *see* the solder "wet" the lead metal, and the pad or other lead metal, and finally merge like the waters of the Danube and the Sava.

Of course the problem may NOT be your solder joints. Plugs and jacks are very prone to tarnish and poor contact. (However re-re-re-inserting the plug often changes the problem, better or worse.) But practice in making solder joints RIGHT the first time is a valuable skill. (But lower-pay than a mature EE.)

Yes, in theory an oscilloscope is a great tool. Some problems: in simple audio circuits, a "grissscht sound" may appear at ALL points in the circuit, and probably bigger "away" from the actual problem (because of amplification). Also problems which happen all the time on stage often "fix themselves" when brought to the 'scope: the 'scope scares the circuit into correct operation (or the jostling of bring the device to the lab and 'scope wiggles a bad connection into a temporarily good position). (There is also supersonic oscillation which quits when the scope loads the circuit.)

One last thought: cell-phones make some very strange interference. An idle cellfone makes irregular checks of all the cells around. The handshake is digital which sounds like random noise. I just ran into that while checking a wired phone by calling to/from my cellfone. When close together, the wired phone had very strange irregular static sounds.

[Paul Marossy]

But practice in making solder joints RIGHT the first time is a valuable skill.

Good advice. It's always better to get it right the first time.

When I go back to my early builds I am amazed at how crappy my soldering looks. Usually always worked OK, but it didn't look good at all.

[FIŠ]

Thank you all for suggestions, I have decided to make a completely new PCB, properly cleaned and prepared this time, using phototransfer method and put it all back together, meanwhile i made one fuzz face and dod 250 on a pre drilled board and they work fine, it will probably take me more time and nerves to find a bad spot on ts808 than simply rebuilding it.

About that EE student advantages... I am currently in my hometown; it is possible though to knock on the doors of my old highschool electronics teacher (yup, been learning about these things for almost 8 years now) and also as soon as I get my paycheck this month I'm getting my own oscylloscope and that is another topic I wanted to discuss - what kind of oscylloscope to get? meaning old analog one or newer digital, and if digital - are these new USB plug and play oscylloscopes any good? (using screen of your PC for a display) or I should be getting "portable" one?

Don't mind me making a new topic about that issue  if noone replys:icon_mrgreen:

Izidor

edit : also in general.. as I am etching and designing my own PCB (first time designing my own layout) and I really like to minimize things as far as i can so wondered : are there any rules to obey as far as positioning is concerned? I mean... to avoide interference of components if there is such thing in stompboxs, thx in advance.

[R O Tiree]

I've got a USB digital oscilloscope and it's excellent. It can be quite expensive for a decent one, but worth its weight in gold when debugging. It has saved me countless hours in that respect. Without one, it's a wild stab in the dark, sometimes, trying to figure out why your latest oeuvre is not working like you thought it ought to. Is it a short/open circuit? Is it a fundamental design problem? Is it a blown component?

I can heartily recommend the 2204 model on this page: Picoscope Entry-level scopes

Clearly, we're talking low-voltage stuff, here. Max is +/-20V on these entry level ones, with over-voltage protection up to +/-100V (DOH!) You want something for higher voltages, you need to start thinking much more than £159. What's particularly good about these PicoScopes is that software upgrades are totally free and backwards-compatible. Mine is quite old (it's a PS2202, which is not made any more) but the newest version of the software works perfectly and has all the Gucci toys (maths, masking, etc) which have been developed over the last few years since I bought mine.

You can get the cheapest, 1-channel one for £125 (PS2104), for sure, but I'd go with a 2-channel + waveform generator - the extra channel and the AWG cost only an extra £24! This makes it dead easy to probe 2 places and see how the waveform has changed between 2 points in the circuit. Mine (old as I said) does not have the waveform generator, so I built my own (pink noise and square/triangle/sine, 0 to 3.2V p-p, 80Hz-3kHz). That means I can attach one probe to the input and the other to somewhere else. This makes it dead easy to view gain, phase-shift, distortion etc. Of course, these new ones allow you to display the internally-generated waveform, so you won't need to keep one probe on the circuit input like I have to. It's still very useful to view 2 places in the circuit at once, though, and it's something I do often.

As to PCB design... huge subject. A couple of years ago R.G. (he posted at #7 on this thread) wrote a book about stomp-box PCB design. It might be worth asking him if there are any copies left? I started this game before R.G. wrote his book and I found it useful to look at the board layouts at places like BYOC (Build Your Own Clone) and similar sites for clues to good design principles. If there are no copies of R.G.'s book around, then start a new thread and people will, I'm sure, be happy to chip in with their experiences.

[PRR]

> old analog one or newer digital, and if digital - are these new USB plug and play oscylloscopes any good? (using screen of your PC for a display)

An _oscilloscope_ is handy but a lot of good work can be done without one.

A listening box and a DC voltmeter can tell you a lot.

If you wonder about an oscillator's amplitude and waveform, make the caps 10X or 100X bigger so it runs 10X or 100X slower, and watch the DC voltmeter through a couple cycles.

For audio work, there are two main specs: speed (MHz) and channels.

Since 5+MHz 'scopes became common, everybody seems to think they need 10, 20, 30MHz. I disagree. I started with 450KHz (0.45MHz!), have re-built a 19KHz 'scope, have owned 15MHz and 50MHz 'scopes, and now own a different 450KHz 'scope. The 19KHz was not enough for some trouble-shooting (it was built for music class, deliberately cutting-out supersonics (and using 19 cent transistors)). The 450KHz jobs show me everything "useful" for audio amplifiers, including 48KHz/96KHz clocking on digital audio and most supersonic oscillations in sick audio amplifiers. I did once have higher-frequency oscillation in a headphone amp, only with audio signal, and the 15MHz 'scope obviously could not lock onto both 2KHz audio and ~~5MHz parasitics to give a useful display. IMHO, something over 100KHz is a plus but anything over a MHz is a waste.

2-channel is often useful, but rarely essential.

The fewer controls, the better. My 19KHz 'scope had few controls: height, 2-range sweep, brightness, and focus.

The _big_ problem with many (not all) PC-connected schemes is that your precious audio signal must share a ground with the very noisy and fragile PC. The PC is full of square-waves at all odd frequencies and sub-products well into the audio band. (Put the tip of a live guitar cord near or inside a PC and listen.) OTOH, any accident in the debugging could possibly blow-up the PC (this is more a risk with tube amps than with battery pedals).

New analog 'scopes are still around. They will vanish in a few years.

When LCD 'scopes came down from $5000 to $300, I got an early one. I like it because it is small (my 450KHz CRT is full 19"/50cm wide and nearly as deep). I know the MHz spec is "poor" and I never noticed. I do hate the multi-function pushbuttons instead of honest knobs.

You can do some work with just a sound-card and software. The problem is that most sound-cards overload at 2V input, and many pedals have 4V to 9V signals inside. Also it is useful to have response to DC but soundcards don't. And soundcards are protected from most "normal" abuse, not the kind of accidents which happen in debugging.

Don't overlook pre-owned 'scopes. Often they are in poor condition and missing probes. Sometimes the guy bought it and never used it, all OK. There is also upgrades: your teacher (or a friend) might be lusting for some newer super 'scope, but can't justify it. If you offer to buy his old 'scope, then he has some money and space in his bench for the new 'scope of his dreams.

[FIŠ]

I see, well I always wanted analog osc but they are expensive and replacement parts won't be aroud for much longer so I think I will go for a digital used one (as they happen to be in MUCH better shape than used analog ones, and there is no maintenance), nice feature that many of them have is that you can save your measurements on a USB stick.
So I want to thank you all because I had some trouble deciding what type to get and what frequency range but its all much clearer now 

[R O Tiree]

I hear everything you say Paul, and agree with a lot of it, but...

The problem with an (expensive?) analogue 'scope with few controls and limited bandwidth is that it's all she wrote. If you need the extra stuff, you're knackered. I can only speak for these PicoScopes, but the default view is the simple one you espouse here (and the most useful, most of the time - stipulated). The more Gucci functions, however, are a mere mouse click or 2 away. Without this 'scope, I'd have given up in disgust on some fundamentally sound circuits. Instead, I was able to diagnose accurately what the snag was and sort it in short order. At least one of those has involved frequencies well above 450kHz (stupid, schoolboy error that I won't make again). Good, old-fashioned knobs are great, agreed, but if all it takes is getting used to a fairly simple GUI to accomplish exactly the same things, then why not? You may never need the (minimum) 10MHz bandwidth of these PicoScopes, but if it's there, at this price point, then why not?

As to interference from PCs, all I can say is that these particular 'scopes appear to be exceptionally well shielded from them. Any interference that I've observed, even zooming right in on the traces, has been inherent in the circuit itself or been picked up in the circuit from some other piece of equipment nearby (SMPS in my soldering station is a favourite "forget me").

Locking on to different, widely separated freqs? Not a snag with these, because it's irrelevant. If it's a large-amplitude signal outwith the bandwidth of your 'scope, you'd never know you had a problem. With an analogue 'scope, a small-amplitude signal at a massively high freq superimposed on your test wave just shows up as a very slight de-focussing of the trace, which you might even miss altogether. A slightly fuzzy trace on a PC screen stands out like a dog's doo-dahs... stop sampling, zoom in... ahh! Now, what's causing that? Simple. An LCD screen has probably got a far lower resolution than a modern PC monitor (mine is 1920x1080) so, again, you'd probably miss this altogether.

You want to find out what's going on in the frequency domain instead of the default time domain? Click... Hmmm... where are those extra harmonics coming from? Simple.

Desk footprint? Even your LCD one takes up space. This little puppy sits on a shelf under my desk, so the only things on the desk are the probes/cables, and only there when I want them to be. Even if I didn't have a handy shelf, it could sit behind my monitor, with the exact same result.

Cost? £159 for 2-channels and an integral waveform generator is a fantastic price with free software upgrades for life, display at PC screen resolution and all the Gucci toys on board...

What's not to like, so far?

Max Input voltage? This is the only drawback to these particular 'scopes, as far as I can see but, for the overwhelming majority of FX pedals, up to and including those operating on 0-18VDC and +/-18VDC, not a snag.

Sometimes old is good and lamentable in its passing. Sometimes, though, for particular applications, new is smaller, better, cheaper and more flexible and should be embraced. I almost bought a second-hand analogue 'scope off eBay when I was hunting around for one - it's what I was used to. Knowing what I do now, however, and using this digital one within its max voltage limitation for exactly what I need, I'm very glad I didn't.

[PRR]

So, Izidor, there's several opinions to choose from.