Top Ten Debug Tips

Started by tommycataus, September 17, 2015, 07:16:30 AM

Previous topic - Next topic

tommycataus

Before I go any further I recommend the debug thread, if you are having trouble with a build.

I was just thinking, my margin of error has decreased considerably since I kept track of the type of problems that were preventing my build from working. I just had the first debug issue in a long time and since I am more aware of my own weaknesses, it took minutes instead of days to fix.

I reckon it's worth looking at your own methods, what you usually do and the type of problems you run into. I don't want to replace the debug thread, but it's worth reflecting on since every member here has their own style, preferred methods and approaches. I do a lot of vero builds so that's where my focus has been.

1. Not socketing transistors and ICs! Sometimes you're so focussed on getting the thing going that you forget to put the key parts in. Almost like forgetting to turn it on...

2. Getting component legs one row out. Careless, so worth going over the schematic or diagram to check the placement of everything.

3. Not cutting traces properly. User the DMM to check continuity here, easy done.

4. Missing a ground connection. There are usually multiple ground connections in a pedal and is easy to miss one if you're not careful.

5. Solder bridges, or anything else bridging traces for that matter. I've spent a lot of the last few years with a blade trying to eliminate solder bridges.

6. Contact with enclosure. When it works fine until you box it up, usually it's something grounding the circuit by touching the enclosure. Or even the board making contact with a switch or socket.

7. Wrong component values. Usually happens to me with electro caps, I do them last as they are often tallest,  and by then my eyes are a bit funny. Again checking the wiring diagram or tracing the schematic is the only way around this.

8. Wiring the power up wrong. Usually with positive ground effects. I never used to have a voltage inverting chip on my test board so sometimes I would get muddled.

9. Cold solder joints. Hasn't happened in a long time, but when it does it's a nightmare. Usually my last port of call is to reflow all joints. Nobody likes cold joints.

10. Problems with test equipment/not the circuit. I've been scratching my head for months before over multiple circuits, and the problem was the psu filter circuit I built. Also those breadboards where the trace only goes halfway along. Annoying.

Anyway, those are my top ten in no particular order. Maybe you guys have your own top ten? If you are new to this, definitely use the debug thread as your Bible. I just found it helps to work out what makes your own builds tick, since I believe most issues are down to common problems. Food for thought.
"Remember, there's a big difference between kneeling down and bending over." - FZ

bluebunny

Good list, Tom.  How about:

1. Don't drink beer.
2. Don't smoke joint.
3. Don't drink wine.
4. Don't stay up until 4am.
5. Don't drink whisky.
6.    . . .

You can see where this is going!   ;D
  • SUPPORTER
Ohm's Law - much like Coles Law, but with less cabbage...

tommycataus

Jeez you're no fun!

I must admit I don't do any of that these days. Not even a glass of red which is highly irregular, but as you say it clouds your judgement. Sometimes you give up, come back sober a couple of nights later and the answer is there: you forgot the diodes or something equally shameful
"Remember, there's a big difference between kneeling down and bending over." - FZ

~arph

Good list,

I'd like to add:

- missing jumper wires

mth5044

I'm guilty of plugging the input into the outpit and the other way around. When you look at the bottom side for so long, you get used to that orientation.

drummer4gc

-plugging guitar into the out jack, amp into the in jack because the pedal is upside down...
-making sure the volume control isn't turned to 0
-spending too long debugging a pedal will usually guarantee that I get frustrated, stop thinking clearly, and do one of the two things above.

blackieNYC

Resistor color codes. Make a jig for your meter - banana plug to a dual binding post, so you can just drop a resistor in there without using probes.  More preventative than a debug.
  • SUPPORTER
http://29hourmusicpeople.bandcamp.com/
Tapflo filter, Gator, Magnus Modulus +,Meathead, 4049er,Great Destroyer,Scrambler+, para EQ, Azabache, two-loop mix/blend, Slow Gear, Phase Royal, Escobedo PWM, Uglyface, Jawari,Corruptor,Tri-Vibe,Battery Warmers

bloxstompboxes

I have definitely done the accidental swapping of in and out jacks when rocking it before I boxed it, and probably more times than I would care to admit.

Floor-mat at the front entrance to my former place of employment. Oh... the irony.

duck_arse

debug starts before electricity is applied.

test continuity of B+ and B- from batter snap/dc socket to board.
test continuity of B+ line and components attached to it.
check continuity of signal in to signal out, and work the bypass if fitted.
" I will say no more "

LightSoundGeometry

Solder techniques and layout design = #1

there was a learnign curve with Small Bear solder last night...used it for the first time and had runs and bridges all over the place. I was used to using
the cheap radio shack solder.

Not measuring components, then cleaning them before installation. Sticking stuff in the wrong bin, not being organized.

Working whilst tired and or stoned/drunk is not good lol

Granny Gremlin

Quote from: bluebunny on September 17, 2015, 08:16:23 AM
Good list, Tom.  How about:

1. Don't drink beer.
2. Don't smoke joint.
3. Don't drink wine.
4. Don't stay up until 4am.
5. Don't drink whisky.
6.    . . .

You can see where this is going!   ;D

I do not follow these rules. Though I don't get carried away either.  Hasn't affected my builds so far.  Flux smoke is easily a harsher buzz.
my (mostly) audio/DIY blog: http://grannygremlinaudio.tumblr.com/

bifbangpow

My mistakes are almost always one of the following:

1. the wire slightly frayed or broke in the joint and needs to be redone.
2. I did something backwards because I confused right side up with upside down in the diagram.
3. cold joint.

And one of the aforementioned happens with nearly every build.
Keep on keepn on.

hymenoptera

Quote from: LightSoundGeometry on September 17, 2015, 12:42:03 PM
Not measuring components, then cleaning them before installation. Sticking stuff in the wrong bin, not being organized.

This! I'm totally anal about organization of my bins and measuring each part before stuffing now that
I've encounter so many easy mistakes.

Confusing brown red and orange color bands on resistors, the cryptic codes on film, ceramic and mica caps, have all got me before. You HAVE to measure each part before you stuff it, and check continuity/polarity of switches, diodes, double-check transistor pinouts, I've wired 1/4" and XLR jacks wrong, wired up pots backwards...  :icon_evil:

I even once found a bad breadboard jumper wire, straight outta the bag, brand new, that was open! The cheapy 65-piece packs with the molded plastic ends? If i hadn't checked continuity before using it I would have been pulling hair, I just know it! Now I check everything!

Having multiple multimeters is handy in this regard. Different settings and test leads immediately in front of me with no fussing around.

I still need to build a faster axial componant test jig, something quick, like with two springs or something. Gators really start to wear out the thumbs after a while!
"Radio Shack has nothing for anyone who's serious about electronics." - Jeri Ellsworth

karbomusic

#13
Quote from: bluebunny on September 17, 2015, 08:16:23 AM
Good list, Tom.  How about:

1. Don't drink beer.
2. Don't smoke joint.
3. Don't drink wine.
4. Don't stay up until 4am.
5. Don't drink whisky.
6.    . . .



Without 2 and 4 I'd hardly have any working designs.  :icon_mrgreen: More seriously....

TAKE YOUR TIME

^This will save hours of debugging because it takes far more time to find something hiding from you than verifying what is in plain sight a few extra times. I often see "help me" threads with pictures of builds and it's obvious they were in a big hurry. I only know that because I've done it, I know exactly what being in too big a hurry due to excitement looks like.  :icon_redface:

amptramp


karbomusic

Styrofoam cups with coffee in them and soldering irons don't go on the same side of the bench. Trust me on that one.

tommycataus

Love it. Had I thought longer, then yes, I could have admitted wasting over an hour scratching my head having plugged the input jack into the out socket and vice-versa. Also one time I used a PNP in place of an NPN since it had escaped into a different storage compartment and my fingers happened to find it. Now I always socket transistors until I'm happy with them.

Measuring everything, oh goodness what I would do without a DMM for continuity of mechanical parts and sockets, wires and traces. I'd have jumped off a bridge long ago. I always measure resistors, never trust my eyes!
"Remember, there's a big difference between kneeling down and bending over." - FZ

mcknib

#17
Nothing new but using different coloured wire is great for instant visual de bugging I always use green for IN yellow for OUT red for power black for ground and for all pots and switch lugs always white for lug 1 purple for 2 and any other colour handy for 3. That way I can see instantly that IN tip green wire goes to in tip 3pdt lug and pot lug 1 goes to pad 1.

Permanent markers are great I always write IN on the inside of the enclosure so I don't wire the IN and OUT's the wrong way round when the box is the wrong way up......I've done it 1 or 2........hundred times and putting small marks on bi colour LED domes for red side and marking the emitter leg on TO18 can types and writing the hFE value on them and for marking cuts and links on vero board component side and poking a wire through to make sure you cut it in the right place etc.


and definitely +1 on checking test gear I spent forever audio probing a circuit - first probe got audio left it overnight.

Went back to it the next day and I had difficulty getting it in the amp input socket so just used the other input anyway long story short got zero audio nothing, nada spent some considerable time checking power supplies, cables, patch leads, guitar jacks, taking voltages, kicking cupboards and ripping out clumps of hair when by pure chance I pulled the probe out the amp tried cramming it in the 1st input again, looked at it and noticed the tip had broken off the probe jack and got stuck in the amp input. What are the chances of that happening? pretty good in my case I've still got bald patches to proove it!


and as luck would have it I took the amp apart only to find inacessable board mounted jack sockets where I had to try and coax the tip out with a screwdriver but every cloud has a silver lining I never got shocked foaming at the mouth and dribbling all over the electrics and my neighbours now know my complete repotoire of expletives I'm now known as Mr C'mon you F**%$ piece of F**&^% S*&^