Debugging (success!)

[Paul_C]

So far I've made a number of attempts and suffered a few failures, recently more failures than successes

So while I attempt to debug my latest failure (Hyperion 2) I'd appreciate a few tips.

I haven't as yet made myself an audio probe, but I shall before start.

Do any of you have a preferred sequence to testing a circuit ? (I bought a component tester and used it before my latest build).

Is it possible for a component to be considered within acceptable tolerances but yet not work in a circuit ?

Last edited by Paul_C on Thu Mar 14, 2019 7:51 am, edited 1 time in total.

[BetterOffShred]

I always check voltages at key components like transistors and ICs first .. just to make sure they are getting power. I built myself a tester box with a 4 terminal snap speaker connector.. there's a picture in my build thread.. that eliminates jacks and simplifies it to 4 wires.

Typically then I really eye my board and make sure I don't have something off one row, or a missed jumper or cut (Its one of these 90% of the time).

Then I get out the audio probe..

[Jero]

When I first started, 9/10 times I had an issue, it would be something really simple that I had overlooked...connections reversed, wrong row on the vero, etc.
So I would double check the basics like your jacks, dc, and bypass switch. And if you're using vero, make sure there's nothing shorting rows together, like excess solder, or part of a component lead. Another easy mistake, with vero, is not mirroring the trace cuts. I did this the first time, and they ended up being in the wrong spots.

[crochambeau]

I think everyone else has you well covered here, but I'll throw in as well anyway.

Is it safe to say you haven't purchased an o-scope yet? Those are my debugging tool of choice, but we'll ignore that for now.

My method is visual first pass. I want to admire all the joints/etc. involved. Back lighting the circuit is helpful to spot shorts.
DC measurements throughout can root out problem areas, be it a sort, and open, or mismanaged bias.

I find it beneficial to make a game out of troubleshooting, or at the very least involve myself when my mind is relatively fresh and open to information. It's a benefit of experience that you may have a shopping list of plausible faults in mind when approaching such a problem, but it's far more important to let the circuit speak to you if you want to source the issue (that is to say, don't talk over the circuit, or ignore evidence while fixated on your suspicion). Circuits are often mute though, so you have to find ways to speak for them - it can be a zen like catch 22.

Oblique mysticism aside, I'm trying to say you're better off avoiding certain mindsets while investigating. I think that's the biggest tool.

An audio probe is just a DC blocking capacitor (avoid polarized caps) connected to an audio jack/amplifier somehow. I like the scope because I can watch the DC level as well as the signal, but you can get that information without the expensive thing on the bench.

[Paul_C]

Would this do the job, or would I be better off buying an old analogue one? (don't want to spend a fortune if possible)

https://www.ebay.co.uk/itm/New-Assembled-DSO150-2-4-inch-LCD-Display-Digital-Oscilloscope-with-Probe-/273180181503

[crochambeau]

Would this do the job, or would I be better off buying an old analogue one? (don't want to spend a fortune if possible)

https://www.ebay.co.uk/itm/New-Assembled-DSO150-2-4-inch-LCD-Display-Digital-Oscilloscope-with-Probe-/273180181503

That looks like it would work to let you see what's going on, though the screen looks somewhat congested to me. Perhaps you can toggle the data read out on and off?

I guess before my scope comment is taken as advice I should clarify a bit what I use and how: I've an older four input scope, the multiple channels allow me to dial up a trace or two on the power rails as well. So I'll typically be working with a DC coupled channel depicted inside the power rails so I can see electrically where any given point in a stage is sitting at a glance. It's incredibly helpful in design work...

..and in troubleshooting. BUT, when I was confined to a single channel I was not much ahead of just using a voltmeter and audio probe. Still, that's not a lofty investment and looks like something that would be nice to have in my car or guitar case.

[Paul_C]

I'm not planning to get into design (unless I really get into this) and I'll willingly admit that my electronics knowledge is pretty basic, so at this stage if a simple and cheap Chinese oscilloscope will do the job then that's all I need right now.

That's not to say I wouldn't like to end up with a lot more knowledge eventually, but right now I'm happy to take small steps

[BetterOffShred]

I would build an audio probe personally. Get a scope later when you actually need to look at waves rather than just tell if a sine wave is present.

That's just me

[Paul_C]

I shall build an audio probe in a minute (just finishing off a baked potato) and do a little more research on oscilloscopes.

[Paul_C]

Right, I've made my audio probe, can I confirm that I've understood how to use it ?

The way I read it, I should plug the probe into an amp (I have a very dusty Roland Bass Cube in my workshop for this), connect the crocodile clip to the circuit earth and use the other part to test for sound round the circuit.

I don't have someone to play guitar into it (though I could lean it against the bench and pluck it from time to time), could I use my Drone Thing ?

[BetterOffShred]

Yes. The probe plugs into the amp and then you need an input wave form, anything will really work. You can get little waveform generators off eBay for like 5 bucks and then run a signal into the effect from that.. etc

[Paul_C]

I checked it worked, which it did, so I'll sit down and give the board a full check over tomorrow.

I've got the transistors in sockets which helps, but I have a feeling that every build will start on a breadboard from now on

[BetterOffShred]

Nah it won't man. You just need to get a handful of builds under your belt and you'll be back in the saddle. I've had like 3 builds go south on me in a row. This last week actually.. but don't despair.. think about it on the shitter, study the schematic.. and overcome

[Paul_C]

I had an idea last night that I thought might help - instead of the Hyperion 2 layout I could lose the extra pot and make it a little simpler, so I drew out the layout for the Hyperion and realised that not only was there an extra cut I'd missed, but because of that one of the caps on the board is across 3 rows when it should be four

I'm going to make the changes in a minute and see if having everything in the right place helps

It just goes to show that however meticulous you might think you're being, sometimes you can struggle to put one foot in front of another

[Paul_C]

Woohoo!

Now I've got things in the right place it works . . . almost.

The fuzz side is working fine, but if I turn up the oscillation much beyond zero all I'm getting is a fuzzy note, unaffected by what I'm playing.

This is most definitely an improvement, but still a bit to do - onwards ! (waves soldering iron as makeshift sword)