23201a - trying to do a loopback test

Why do you continue to avoid answering my questions? Did you conduct a loopback test in both possible connection configurations or not? It’s a really simple question! And if the answer is yes, why have you started asserting that you never had two different voltages shorted together? Are you intentionally contradicting yourself or do you not understand what I’m saying? If it’s the former, I can’t really help you. If it’s the latter, aren’t you interested in trying to understand why I’m saying you had two voltages shorted together? Brushing it off as “irrelevant”, “hypothetical”, and “intellectual flatulence” makes it look like you don’t actually want to know what you’re doing.

It’s not a matter of our having “an out”, it’s a matter of your outright telling us that you connected two outputs together that were at least briefly at different voltage levels, and if your friend is an experienced electrical engineer (I’m not sure what a “full engineer” is), he should tell you that (a) this is indeed what you did and (b) it can break things. Your history with other parts is irrelevant; this part is not those parts, and this part has clearly labeled pins that are intended to be connected a very specific way. Those parts might have been more complete consumer products, not a simple component, and could have had added protections built in. Also, shorting out I/O lines might not be a 100% reliable way destroy a part, depending on the part, the voltages, the duration of the short, etc. Just because it didn’t happen once doesn’t mean it won’t happen the next time or with the next part! You might be able to get away with dropping a plate on the floor without it breaking, but that doesn’t mean you should expect all plates to survive an arbitrary fall onto an arbitrary floor, and you shouldn’t use your past history of successfully dropping plates as a reason to carelessly toss your new plates around.

You have now mentioned breaking four devices you have received recently (three from us, one from another vendor), which definitely is a pattern, so I feel like your goal should be trying to understand what you might be doing wrong so that you can avoid these kinds of problems going forward. We’re trying to help you with that and instead you’re rejecting what we’re saying as our looking for “an out”. You make it seem like if you were to get a replacement unit, you would just do the same trial and error connections because you don’t actually believe us when we say that can break things and instead think this is a story we’re making up to get out of sending you another $11.50 part. The saddest part is that your statements make it clear that you don’t even understand what you did, and yet you still irrationally cling to the incorrect and wholly unjustified position that it was “harmless”.

We test every product before we ship it, so it is very unlikely, though not impossible, that you received one that was DOA. When you add in your general approach of “connect whatever to whatever”, your lack of respect for the fragile nature of the ICs you’re working with, and your recent track record with similar electronics components, do you not see how the “very unlikely DOA” becomes “almost certainly wasn’t DOA”?

If you’re going to be working with these kinds of electronics, you need to have some appreciation of just how close you might be to the physical limits of the components you’re using. I think you could really benefit from reading this article: pololu.com/blog/5/abstractions

- Ben

No, the part from the other vendor was a GPS module that was not reliable and on their forum they posted a suggestion to contact them if you were dissatisfied, which I did. They credited me for it. I have 6 working micro-controller systems comprised of about 50 boards and modules from about half a dozen vendors. I have only had 3 parts experience a total failure, so my overall success rate is over 90% and in fact is 100% with every vendor except one. Yes, there is a pattern.

I do not connect “anything to anything”. I will try serial data Tx/Rx lines the way that seems to make the most sense and then try swapping them if that doesn’t work. Thats is a very specific well known exception to best practice that is even mentioned on the product page of a competing product. Extrapolating out that specific exception to say I just connect “anything to anything” is inaccurate at best.

You’re not being specific, but based on your past comments, it sounds like you’re talking about the SparkFun adapter, in which case you’re talking about a user comment that the poster himself prefaces by saying he has “a condition called UART dyslexia” that he further describes as “a neurological disorder”. Even if he is half joking, that doesn’t make for much of a refutation of the expert advice you’re getting here. The poster there says “I have had some success with this approach”; we are giving you a specific explanation of why it’s a bad approach.

- Jan