-9
\$\begingroup\$

I see so many questions, including some of mine, in which a particular implementation is described and help is requested.

And, without fail, the team of MCU pushers comes around and strikes down any solution which does not use the MCU.

This occurs without even considering the original design requests.

A good example is I asked for a way to wirelessly send data. The question was even downvoted and all the information I got was to 'USE A MCU'.

Well, with only two chips I was able to solve my problem. Of course, I did not use an MCU. I stated in the question my reasons for avoiding on in this case. But, inevitably, I am given the normal responses: "USE A MCU! WASTE OF TIME OTHERWISE".

Just today there is a question regarding making a clicky sound with a relay. Of course, the gang comes around again. "USE AN MCU". Why? For god's sake, the OP is requesting a retro sound via analog means (relay or turn-signal clicker). It's not like he asks "How can I make the clicky sound". No, he asks "I want to make the clicky sound with a relay, how can I make it louder".

Why is there this great push for all solutions to all problems to be solved with an MCU? They are arguably less robust than most direct solutions, and they abstract the problem space to an area which may not benefit the person asking the question.

I understand that MCU's are powerful tools that can do great amazing things.

For example:

  • light up a matrix LED display to music;
  • control a remote car or drone;
  • control a smart fridge;
  • control a camera;
  • track my energy usage and wirelessly relay this to the energy company;
  • run my thermostat with my time and temperature preferences;
  • send sensor data in real time to central hub;
  • control the timing of my garden watering schedule taking into account rain and sunshine;
  • analyze plant controls and implement feedback;
  • and a million other practical examples

But I am getting a little tired of every question in which a particular implementation is provided being rebuffed with "USE AN MCU" (without even analyzing the originally presented idea by the OP).

Can we please judge each question on its own and not automatically assume the user is an ill-informed idiot who should use an MCU?

| |
\$\endgroup\$
  • \$\begingroup\$ I'm voting to close this question as off-topic because it isn't a question at all. \$\endgroup\$ – Dave Tweed Oct 8 '16 at 20:57
  • \$\begingroup\$ @Dave Tweed: As I expected. I tagged as "discussion" so that it could be discussed. The site has a problem, which is the incessant pushing of MCU's for simple design problems. You can discuss it or you can "make me respect your authority" and close the question. Really pathetic honestly. I put this on meta because it is a meta-problem on the EE stack exchange. \$\endgroup\$ – jbord39 Oct 8 '16 at 21:06
  • \$\begingroup\$ @Dave Tweed: Put on hold because "does not appear to be about Electrical Engineering Stack Exchange " Yet the question is 100% related to Electrical Engineering Stack exchange; how can you even deny it? This is another problem. High rep users just arbitrarily closing questions based on them not liking it; taking on some random arbitrary reason to close it. So pathetic dude, I guess you retired and now no one listens to you so you build up high rep here and close questions? \$\endgroup\$ – jbord39 Oct 8 '16 at 21:08
  • \$\begingroup\$ @Dave Tweed: I insulted your answer and you lash out by closing question. If you were truly objective you would base your decision soley on the question and not on myself personally. Another high-rep user abusing his privileges. Can we make downvotes -10 rep so people can't just grind out rep like the wild boars in world of warcraft??? \$\endgroup\$ – jbord39 Oct 8 '16 at 21:09
  • \$\begingroup\$ @DaveTweed: From the help center: "Electrical Engineering Stack Exchange users to communicate with each other about Electrical Engineering Stack Exchange (asking questions about how the websites work, or about policies and community decisions)" This is a question which you may not like. But how can you say that it is not a "communicating with each other about Electrical Engineering Stack Exchange community decisions". You are a fascist if you say otherwise. \$\endgroup\$ – jbord39 Oct 8 '16 at 21:12
  • \$\begingroup\$ Wow, you are really angry about something, but I'm beginning to think that the stated topic of your question really has nothing to do with it, other than being the proximate trigger for your tirade. Would you like to take this to chat? I've got a couple of other things going on right now, but I'll try to be as responsive as I can. \$\endgroup\$ – Dave Tweed Oct 8 '16 at 21:25
  • \$\begingroup\$ I would love to. And no, it is the stated topic which is bothering me. I was hoping responses would be more like "well you see many times in the past people came and we gave them direct solutions and they never worked so now we always push mcu solutions" but instead I get ridiculed for honestly a completely legitimate inquiry about a real problem on the site. MCU's are not the end all be all. You can do stuff with transistors and resistors, even in the 21'st century. And my questions get getting downvoted because I say this. \$\endgroup\$ – jbord39 Oct 8 '16 at 21:28
  • 1
    \$\begingroup\$ Good questions yield good answers. \$\endgroup\$ – Scott Seidman Oct 8 '16 at 21:29
  • \$\begingroup\$ And in the case of my wireless transmission an MCU is stupid. It really is. I still have to send the data through some encoding and antennae. Since the data is 1-bit, changes once every couple seconds at most, is very noncritical, what is the point of the MCU? On top of that the analog aspect of it almost increases the security compared to a vanilla MCU implementation!! There is no analysis. It is like using a chainsaw to cut a flower. It was a blanket response which given the design still has no basis. Defend it if you want, but a direct solution was better. \$\endgroup\$ – jbord39 Oct 8 '16 at 21:30
  • \$\begingroup\$ Did you notice that NOBODY gave you an MCU based answer to the wireless thing, and nobody down voted the question? It's hard to understand what you're whining about. \$\endgroup\$ – Scott Seidman Oct 8 '16 at 21:57
  • \$\begingroup\$ I never linked a particular question, and I am not talking about my questions specifically, although I certainly saw it there. The question linked was immediately voted to close as being "too broad" (even though I specifically stated my constraints), because I did wanted to avoid MCU. Also see the comment section where other high-rep users literally mocked me for even considering avoiding MCU (which I successfully did for less than $4 per additional transmitter, each having individual security/address codes). Another question which I deleted was given MCU answers. \$\endgroup\$ – jbord39 Oct 8 '16 at 22:09
  • \$\begingroup\$ @ScottSeidman: I have no problem getting advice from experts, which is what I create questions on this site for. But being mocked for even considering an MCU-less implementation for ANY design is a bit absurd. A proper answer would address the specific constraints and compare the MCU to without. Maybe MCU is the only real answer. But sometimes it is not. Rather than force it down someones throat it is better to explain WHY their initial design is bad and why a MCU is a better option. \$\endgroup\$ – jbord39 Oct 8 '16 at 22:12
  • \$\begingroup\$ There might be a few unfortunately mocking replies mixed in, but the primary message you've been given is of the fact and reasons why your claims are technically wrong. The reality is that unless you need to make hundreds of thousands of something, or unless an MCU would be too slow or otherwise unworkable, non-MCU options requiring something more than a comparably well-known, inexpensive, widely available dedicated function chip do not look favorable unless you want to pursue one for purely educational or aesthetic preference reasons. \$\endgroup\$ – Chris Stratton Oct 9 '16 at 2:31
  • 3
    \$\begingroup\$ Re-ask a question with a truly legitimate technical requirement such as blindly fast speed that truly precludes an MCU, and you will discover that many of those you call "MCU pushers" actually do have other types of solutions in our quivers, but only reach for them when doing so is warranted. Conversely, I doubt you'll ever find a working engineer who hasn't seen weeks of productivity lost to an unwise choice of technology made by someone else before they were consulted... so there is good reason why many will jump to head off unwise decisions from the start. \$\endgroup\$ – Chris Stratton Oct 9 '16 at 2:34
5
\$\begingroup\$

In the case of your own questions, responders have suggested MCUs because it is clear from your question that you have not thought the problem through at all. It's obvious to an experienced person that you are going to run into issues that will end up requiring the complexity of an MCU anyway, so you might as well start with one.

For example, in your "Least overkill solution to transmit single bit wirelessly" question, you state that you want to "transmit just a single bit wirelessly". Yet, in the same paragraph, you also say "I want to [have] a couple of these located in different places". Clearly, you have not even considered the question of how the receiver distinguishes between the different transmitters. Maybe you don't care, but you didn't state that, either.

It's far too easy for a beginner to drill down to what he thinks the critical issue in his project is, and to then ignore suggestions from others who have done similar projects and have a better idea of what the "big picture" is going to end up looking like. Try to be willing to listen to them.


However, I agree completely with regard to the relay noise question. I never even looked at the question before now, because I thought that the obvious solution was to add appropriate mechanical resonance to the assembly in some way. It didn't even seem like an electronics design problem to me at all, and I was considering moving it to Engineering.SE.

| |
\$\endgroup\$
  • \$\begingroup\$ Yes, I don't care which one activates the switch. It is a dead-ass simple wireless motion sensor which needs to only light up when any of the triggers are switched. There was no need to distinguish which transmitter sends the signal. And I did think of that. What is annoying is that every question someone asks is assumed that the OP is a complete idiot unless he has 10k plus rep. Anyway I found this completely simple and amazing solution HT12E/D which is 10x as robust as a MCU implementation and requires a fraction of the power. AND it does distinguish easily between transmitters. \$\endgroup\$ – jbord39 Oct 8 '16 at 20:32
  • \$\begingroup\$ And it has stood up the rain quite well. I did not feel the need to state every design constraint, and instead summed it up with "avoiding MCU if possible". And many users poo-poo'd the idea that I would even consider avoiding a MCU. Well, the HT12E/D has survived 3 texas downpours and a couple months of texas summer heat without any problems. Unfortunately no one on this site even mentioned these seemingly ubiquitous IC's. I did not come here to receive a lecture. My perceived "lack of big picture" is really just an overall snap judgement, most likely based on my low rep. \$\endgroup\$ – jbord39 Oct 8 '16 at 20:44
  • 5
    \$\begingroup\$ If you didn't come here for a response, then you're just whining, and there's no reason not to simply delete this "question". \$\endgroup\$ – Dave Tweed Oct 8 '16 at 20:56
  • \$\begingroup\$ I tagged this as "discussion". It is to bring to light a problem with the site. You can disagree that it is a problem if you want. but it is. \$\endgroup\$ – jbord39 Oct 8 '16 at 21:05
  • \$\begingroup\$ And I did come here for a response. But that was not a "lecture". I guess that is all you are capable of giving, and when it is logically rebuffed you resort to using your pathetic (yes I keep using that word because it is PATHETIC) power granted by grinding rep instead of actual engineering fundamentals or analysis. It is sickening and pathetic. You are incapable of discussion with others unless they either have high rep or ostensibly bow to the power your 80k rep yields. I was hoping for it to judge each question on its own. But you voting to close when I criticized your response ... \$\endgroup\$ – jbord39 Oct 8 '16 at 21:16
  • \$\begingroup\$ I don't see how a logical person can come to any conclusion other than that you abused your power to arbitrarily close my question because I did not like your answer. If you were a true judge you would have closed the question upon reading it. But instead you waited to hear my response to your answer (which presumably if it were of approval and subjugation the answer would not have been voted to close) before deciding for yourself. When it turned out that I disagreed with your diatribe, you decided it was no longer a valid question. It is very pathetic.I would be very interested in the defense \$\endgroup\$ – jbord39 Oct 8 '16 at 21:40
1
\$\begingroup\$

Why is there this great push for all solutions to all problems to be solved with an MCU?

  1. Design problems rarely come alone.
  2. MCUs can solve more than one design problem at a time, and do that flexibly.

p.s. For crying out loud, we're well into the XXI century.

| |
\$\endgroup\$
  • \$\begingroup\$ Yes, I know the century we are in and design CPU's for a living ... but there is no reason to over-engineer a solution to simple problems. "make clicky sound with relay" ... "no, lets use an MCU to store/replicate/and reproduce that sound instead ... much simpler than hooking two wires up to a solenoid" \$\endgroup\$ – jbord39 Oct 8 '16 at 18:09
  • \$\begingroup\$ Plus MCUs and supporting parts ($0.50) are often a lot cheaper than a relay ($5). This is just an example of course but is increasingly the case that the low cost method is to imply offload all the hard stuff to software. I'm not happy about that by the way, but it's just life. \$\endgroup\$ – user98663 Oct 14 '16 at 7:00

Not the answer you're looking for? Browse other questions tagged .