# I’ve obviously rubbed people up the wrong way - no idea how - am I doing something wrong culture-wise on this site?

Have used StackExchange quite a lot, first time user on the electronics site though. Posted a question where I am after a value:

Length of Signal When Simulating a Computer Fan Speed Sensor

Seemed to have annoyed Elliot here and can’t see why they’re acting hostile.

Ready to eat humble pie if I’m being antagonistic, however this is the only place where this question received this kind of response. Elsewhere we’ve been going through motherboard’s supported maximums and are trying to find a standard.

Am I doing something wrong culture-wise on this site?

• Many of the EEs are old grumpy men, just ask them directly. They don’t wish to crush you, they are only emotionally violent because it helps you learn. This is not a philosophical riddle, after all. – user2497 Jan 12 '19 at 15:38
• This is by far the most hostile SE site I’ve encountered. – OnStrike Jan 14 '19 at 3:46
• @theNamesCross You're not getting a medal for participating, that's for sure. – pipe Jan 14 '19 at 14:21
• @pipe Why are you talking about participation medals? If you meant to insult me specifically, fine, but that reinforces my statement. I have referenced and used many sites within SE for long enough to render my opinion. – OnStrike Jan 14 '19 at 14:39
• @theNamesCross The "collective" you. I see loads of questions and answers on other sites on the network with positive scores even when they're just crap and does not bring any value - together with a friendly "thanks for joining!" message. Completely counterproductive. – pipe Jan 14 '19 at 14:52
• Too bad the question has been deleted. Is it possible that you suggested that Elliott did not understand how real engineers do things, that they rely on standards? Sounds kind of condescending to me. And then you felt it was hostile when he suggested that if you are familiar with standards then you should know how to find them, thereby answering your own question? Maybe you interpreted terse as hostile? – Elliot Alderson Jan 16 '19 at 22:40
• @user2497 Only arrogant, naive young people would resort to the "grumpy old men" stereotype. :) Just kidding, my skin is not that thin. – Elliot Alderson Jan 16 '19 at 22:42
• @ElliotAlderson I didn’t use the word grumpy, I said you were all educators in critical thinking :) – user2497 Jan 17 '19 at 15:41
• @ElliotAlderson merely looking for a standard or at least a convention. Asked the question here as in future would like to avoid discussion veering into personal attacks and avoiding the answer required, neither of which interests me, benefits sites like this or is a good use of time. I appreciate that you took time to look at my question, this is your own time on a voluntary basis. Received the answer I needed in the end (convention is 10k RPM max) with the same wording as the original question elsewhere. From the answers below I can see some failings of mine when posting on this site though. – Metalshark Jan 18 '19 at 5:05
• @theNamesCross EE.SE is a tough board. – Nick Alexeev Jan 20 '19 at 20:59
• This question and all the comments are proof that EE thinks the new code of conduct doesn't apply here... – Passerby Jan 21 '19 at 23:30
• @theNamesCross If you think this is a hostile site you should try visiting the Physics site as an experimental physicist and having your (physics) questions closed for being "engineering". I had a 10,000+ rep on that and got so pissed off I unsubbed entirely – Dirk Bruere Apr 12 '19 at 10:47
• Received the answer I needed in the end (convention is 10k RPM max) - It is a shame that you deleted the question though, as you could have answered it yourself, and that answer may have helped someone else in the future, who is experiencing that same, or similar, issue. – Greenonline Apr 30 '19 at 20:11

Dunno, looks OK to me. The extra code is just in the way because I'm not going to read a wall full of untested code, and as long as it's in the question it makes me think I have to read it just to answer it.

Elliot should go look for the standards or check a few fans so he can answer the question instead of spending time complaining. It's an answerable question that is:

• a specific electronics design problem
• the writing of firmware for bare-metal or RTOS applications
• practical, answerable questions based on actual problems that you face
• reasonably scoped
• invite sharing experiences over opinions

(from on-topic and dont-ask)

• Fair comment, will remove the code. Originally didn’t have it in the question, but added it when the comments started. – Metalshark Jan 12 '19 at 17:48

The SE network overall likes specific questions, rather than open-ended or hypothetical ones.

So generally, a question will be much more well-received if you can post your research efforts. Such as links to the datasheets on fans or hall sensors that you have looked at, a specific partnumber of the motherboard etc. Even if you are interested by the general case, it helps if you can give a specific part for the sake of discussion.

And if you can go as far as providing a simple example schematic, the grumpy old men will just melt, like a backwards-mounted TVS diode.

Posting code is great too, but only if relevant to the question. Generally, code should be in the form of a Minimal, Complete, and Verifiable example.

(PC questions in general are more suited for https://superuser.com/, but as long as you ask about how to do something specifically related to EE, like creating a replacement for a Hall sensor, it should be fine here.)

As for why someone got grumpy, it is probably because your ideas of how to write a timer function with hardware timer interrupts seems quite naive :) You wouldn't use "delays", doing so inside an ISR is a big no-no. You would rather set up an output compare timer counter, or alternatively use a PWM output. Since this is beginner level stuff when it comes to embedded programming, some people who've been writing timer drivers for 20 years might get a bit impatient :)

Notably, just because someone gets grumpy, it doesn't mean that there's anything wrong with the question.

• Thanks @Lundin. Providing the links to standards which don’t contain the information required (like the Intel one) would have removed some of the chatter and is a great idea. FYI the PWM signal can easily be shared (normally just daisy chain it) and the interrupts are queued on Arduino, so here it saves a mile of code when doing some like: all 4 interrupts must fire, then pulse the output once. If one (or more) of the interrupts fires whilst waiting a minimum time is waited for the pulse then it is collected ready for the next iteration without having to mess with semaphores, etc – Metalshark Jan 15 '19 at 5:09
• Thanks for contributing, I think your answer was fair and I appreciate the effort you put into it. Peace. – Elliot Alderson Jan 16 '19 at 22:44

I don't want to wade through the edit trail, which is extensive, but it really looks to me like you're slowly adding the info needed to provide an answer, and it's frustrating a user. The user seems a bit grumpy, but certainly not over the top.

My bent is to simply ignore such questions, or close vote as "unclear what you're asking"

The first frustration, for me, is the long edit trail. It makes it difficult to understand the history of the question. There is no formal policy on this, but if you've gone through three edits and still need to change something, you might consider deleting and starting fresh (or ask a mod to help with deleting). The long edit trail is a pretty clear sign that there is a problem.

As near as I can figure out, your main question is "How fast a signal can a motherboard handle?" without even providing information on which motherboard. The user's issue with this is pretty valid. That makes the question underspecced. For all I know, you're talking about a PDP/8. Do you mean "all computers a person is likely to be able to buy today?" Do phones count? Chromebooks? How fast is the motherboard clock?? I suspect that will make a difference.

Then, it looks like orbiting around the main question is "how can I program this"? When the whole question has an anchor that's floating around, that can get pretty frustrating to answer.

You also really never get around to saying what "combining four fans" means. Are you trying to drive them all with one PWM signal? Are you trying to just AND their Hall Effect outputs together? Are you building a fan control board? Without details like this, this is an XY problem.

The best approach for this particular question, for a start at least, would be a question on SU to the effect of "is there a standard for a motherboard fan interface, and where can I find it?" I think that question would float on this stack, too.

FWIW, a quick google for 4 wire fan control standard yielded https://folk.uio.no/kyrrens/diverse/viftekontroller/developer-specs-REV1_2_Public.pdf as a first hit.

• Thanks Scott. I can see that if it’s still not clear that I am combining the RPM sensor output and looking at the minimum sampling rate that an RPM sensor must support (sadly missing from that Intel spec which goes to great lengths at explaining the rate for the PWM signal) then I really should delete the question and start again. Personally was frustrated that other places got the question straight away, but have to accept this as my failing on this site due to the audience and the accepted way to communicate. – Metalshark Jan 15 '19 at 5:04
• Thanks for taking the time to go through that question and the string of comments. – Elliot Alderson Jan 16 '19 at 23:28