- This topic has 40 replies, 5 voices, and was last updated 13 hours ago by
.
-
When you say to avoid the urge to debug by part swapping, how would I test components properly then? In this case, I was going to try to trace the failure point from the power supply pins, where I can now detect continuity between two pin pairs, unplugging components as I test them. Is there a better option?
Apologies in advance it you already know this… What I consider “debugging by part swapping” would be something like blindly replacing any component because it might be the one causing the problem. For example, if you aren’t getting a good 5v signal, you could replace D16 to D19 with new parts. It is unlikely that this would fix the problem. Instead, I think that it is better to understand how the power supply works, then devise and run tests to verify it it is working or not. The 5v supply is very similar to that described here. Using your meter, check to see if the voltages and waveforms are what they should be. If not, the problem could be in the circuit or external to it. Isolating the power circuit would be the next step. I would start by removing connector P2 and verifying the output of the rectifier and transformer. If that checks out, remove P2-2 (pin2 2 of connector 2) and reinstall the connector. This isolates most of the 5v power supply from the external circuit (it also provides a convenient place to measure 5v current draw later). Retest the 5v power signals. Still no joy? The problem must be in either the power supply circuit or reset circuit. Lifting a leg of R75 further isolates the power supply from anything else. If the power supply still doesn’t work, there are only a couple of parts that could be the problem and it should be pretty obvious which one(s) based on the meter readings. Now is the time to start removing, testing, and possibly replacing individual components (based on data instead of hunches). I purposefully went into painful detail here to illustrate an approach that is based on 1) understanding how the circuit works, 2) measuring to verify proper operation and 3) isolating the circuit for ease of debugging. Also, note that there was a minimal amount of soldering involved. Much of what I described above is implicit in the advice that you got from others. Hopefully my explanation provides insights into why things are being suggested.
Based on your posts it sounds like you are getting more serious about this hobby. Check out the Digilent Analog Discovery. It might be useful to you. In general, anything that it does can be done better with dedicated bench equipment. However, everything that it does, it does well and is often all that is needed.
Glitch
- You must be logged in to reply to this topic.