Thanks Steve. I bought a Kenwood 20mhz dual trace scope and a couple of probes so I have the right scope at least. I think I'll go ahead and pick up the book you recommended.
Honestly I thought problems would be easier to detect. It seems finding them may be even harder than solving them in many cases.
Oh well, anyone else have any tips that might make things simpler. As always, I really appreciate the sharing of knowledge! This forum is awesome.
Shane
[quote author="SmG"]Just to expand slightly on what owel said:
It's worth considering what oscillation will do to a test signal, because that will dictate what the scope shows you. If it's a low level of constant signal, then you may get a thickened trace, and very likely some triggering instability. Also, if you turn the timebase control to its shortest time/cm setting (usually fully clockwise), you may see what is like a solid block of waveform that you can't necessarily trigger on at all - but this depends rather on the spec of your scope.
Other times, you will get little 'bursts' of oscillation showing up, just on parts of the waveform, and you will be able to see these at more normal timebase settings.
So basically, the thing to do is to be bold with the timebase control - wind it right around. It's worth noting that not all oscillations are fast - you can get LF instability too, which will become rather more obvious at lower speeds. This is usually indicative of power supply problems. You should also experiment with the triggering options - sometimes altering these will let the scope triger on things other than your wanted signal, which can be quite useful.
Also, there are a couple of other things to note in terms of connection technique. If you have a really high frequency oscillation going on, then just the application of a scope probe to the output of the opamp that's causing this, especially if its set to x1, can kill it temporarily for you. This can make some problems rather hard to spot. Fortunately though, a small coil of copper wire (half a dozen turns) connected between the probe and its ground connection will still pick this up if it's waved near your circuit. This is because the oscillation will be radiated as RF, and that coil will have a rather higher impedance at these frequencies than the dead short you might have thought it would be. Most scopes with a bandwidth of 20MHz and reasonable Y sensitivity will be able to show you this.
If you think that you've got this sort of problem, then another thing you can do is to get a AM tranny radio tuned between stations and put it next to your circuit whilst you touch the scope probe on and off. If there is a change in the level of hash in the background, then there's a good chance that your amp is an oscillator at higher frequencies.
And if you are going to get into this, then the other thing I would recommend you to get hold of is a copy of Bob Pease's 'Troubleshooting Analog Circuits' - there is a wealth of good stuff in that book.
Oh, and the foregoing all applies to an analog scope - these cheap sampling digital scope kits you bolt onto your PC really aren't any good for this sort of thing, because they don't have anything like the bandwidth required unless you pay a fortune for them. For this sort of troubleshooting, ANALOG RULES.[/quote]
