5
\$\begingroup\$

I asked this question about replacing a Darlington transistor in a power amplifier with something else. It was closed with the reason:

Questions seeking recommendations for specific products or places to purchase them are off-topic as they are rarely useful to others and quickly obsolete. Instead, describe your situation and the specific problem you're trying to solve.

Well, I sort of disagree with that characterization of the original question, but nevertheless I edited the question to make it as generic as I could while still retaining the concrete example.

"Describe your situation": I want to create a high power precision voltage power supply from a low power precision voltage source. I have a schematic for one that uses a Darlington transistor.

"The specific problem you're trying to solve": it seems Darlingtons have gone out of favor. At the very least, the one in the schematic is deemed obsolete and there is no obvious replacement for it. That suggests to me that Darlingtons as a class have been replaced by a better kind of component. So 1) what is that component and 2) how do I use it in this application.

These edits failed to reopen the post. "Original close reason(s) were not resolved".

I'm pretty much at a loss about how to move forward and resolve the close reason. I am not asking for a part number or vendor, I am asking about how to update an apparently obsolete amplifier design. (Thanks to a commenter, I've already found a source for the obsolete component in the schematic, but that still leaves me with all these other questions.)

I mean, I suppose I could remove all references to specific components and ask "How do I build a high power precision voltage power supply?" but that is awfully vague and would likely be closed for needing details, clarity, and/or focus.

The answer to my revised question would not be quickly obsolete and would be helpful to anyone looking to build a highly regulated low-noise power supply, and especially anyone looking to replace a Darlington pair in an existing circuit design given that they are now harder to find. Additionally, it would be helpful to many people to know why Darlingtons are obsolete and what they have been supplanted by.

Please help me understand how to make my question appropriate for the site.

\$\endgroup\$

1 Answer 1

2
\$\begingroup\$

I can't speak for those that voted to close, of course, but I would offer my perception of the situation:

Readers don't like presuppositions or unchecked (esp. mistaken) assumptions in questions. It's generally better to ask a question neutrally, rather than asking why some given (shaky) position is right or wrong -- most often, it's neither and both, simultaneously! It makes more effort to answer, and tends to make a mess that's both confusing for the asker and readers, while diluting what could've been a more concise point.

Whether this is a correct, right and good use of the Close function, or reflected by the reason selected, is another matter. I tend to agree, at least the wrong reason was selected here. I would not have closed it myself.

To be perfectly honest, I don't have a great substitute to offer for this strategy. I tend to do it a lot myself; the intent is usually to provide context around the question, and to illustrate what basic level of competency one has on the subject (for example, to anticipate trivial / oversimplified "did you turn it off and back on again?" sorts of comments/answers). There's also the intended good will of "see I did the 'homework', could you at least check my results?" (Which can also be, perhaps one isn't aware of, or expecting, what degree of answer may come, and being overly humble, when a more comprehensive answer would indeed be gladly given.) Of course, when ones' state of knowledge is meager on a topic, it... well yeah, it sure does serve its purpose: those assumptions tend to reveal, probably more ignorance, or more mistaken conclusions, than was likely intended. So you can see why it can cause problems.

(Also, to note: a close action isn't negative, at least in and of itself. It serves the purpose of, I would say, setting a minimum threshold for clarity and quality of a question. It can tend to be used "in anger" so to speak, though. Voters can vote to close for any reasons they see fit, so it can be a bit arbitrary at times I'm afraid.)

Anyway. Here are some alternative questions that, I think, would likely go over better (or have already been asked; of course, one should search first):

  • "How to substitute a Darlington?" A sufficiently comprehensive answer to which, I think, is likely to check their own assumptions by going shopping and checking that such parts are, in fact, available; at least presently (bonus points for checking life cycle so that the answer is likely to remain valid some years into the future). Supplier links might even be provided! This can be a sneaky way the right way to ask "where to buy--?" questions. (Mind, if its "where to buy"-ness is too obvious, it's still likely to get closed.)

    Regarding closed-ness, I think the crux of the reaction here was: looking at just the top two paragraphs, and the title, one can reasonably deduce it's either a "repair" ("oh they're just trying to substitute some existing part") or "where to buy" question without having to read the rest of the post. Is this hasty? Unfair? Yes; but one can also view it as a lesson to be concise. Not a great lesson, granted, but there are negatives and positives to pretty much everything, and it's up to ourselves which ones we focus on.

  • "How to design a precision fixed power supply?" Sometimes, a more general question is better. But be careful: there are dozens of specifications needed before a small enough set of circuits can be selected for answers. In this case, grounding it by providing your application would suffice (ordinary lab use, DC only, undefined parameters can most likely be ignored e.g. AC noise, compliance range, efficiency, cost, etc.).

    Well, I don't know about that last one, cost. Given you're scrounging appnotes for schematics here, one might guess simplicity is desired, and with it, low cost. I don't know how that compares with the labor/time spent, say, reading this answer alone... but, keep that in mind, too. Lab equipment is expensive because, well, it does its job well for one, but also, you only need one (well, per instance that you do..), and the time spent worrying about, let alone building, testing and qualifying an alternative, adds up really fast.

  • Avoid X-Y problems: don't ask about proposed solution Y to problem X. Instead, just ask about problem X. Often, proposed solutions are driven by mistaken assumptions, and just get in the way (see above!).

    For example here, it's not obvious if you would be at all interested in, or served by, solutions to your "precision 10V 4A supply" application directly -- solutions which would likely completely and utterly ignore what seems to be the bulk of your question (substitution). Either subject would seem to be a viable answer to the question -- and potential grounds for a "Needs more focus" close.

  • Check your assumptions. Is a fixed voltage really what you need here? If you're calibrating ammeters, surely you need a fixed current instead? True, a resistor converts that voltage to a current, but how are you ensuring the resistance remains fixed (it has a tempco!), and that the voltage measured is across the resistance as intended (Kelvin connections are not shown in your schematic)? Have you accounted for all sources of error in the measurement? And how you will deal with them?

  • Along similar lines, be careful whether goals are aligned between your question and your actual need.[1] To wit: presenting it as a power supply question has me asking: what about startup or transient-load conditions? Short circuit handling? (The circuit shown will blow up if overloaded, or overheated.) Drawing from application notes can be a dubious prospect; application notes are generally, at best, jumping-off points for further development or refinement, and should rarely if ever be used verbatim. (For this one, I would at least want to see better development of compensation, stability, a worked example of operating area, and current limiting added, preferably temperature limit too. But, you see -- these are all features that might simply not be at all important to your application.)

  • A more interesting focus is likely to pay off. Comparing and contrasting BJTs (plain or Darlington) with P-channel MOSFETs, in linear service, is an interesting topic; I might even ponder an answer to such a question myself. Most answer-ers have their depth of knowledge in at least narrow fields of expertise; this is a sufficiently deep topic that is likely to draw in multiple answers (perhaps even some controversy engagement between them, too!). The downside for you, of course, is: a question on a deep enough matter, may be too abstract for you to turn into your worked problem. (Keeping it focused, and grounded, with an application (say, V/I input/output range, bandwidth, precision), helps with this.)

Also, try to avoid -- ehh, I don't want to say gimmicky language, but, let me put it this way: I know what you were going for with the "your challenge, if you choose to accept it" sort of stuff, but, that particular line is a bit cliche, and, just keeping it clear, professional and matter-of-fact is most likely for the best. Certain catchy questions may indeed drive engagement, but overdoing it can also drive eye-rolling, and can add fluff to what could be a more concise question.

And, yeah, as you can see, I'm hardly one to talk about conciseness. It's a lot easier to see conciseness from a distance, but a lot harder to cut to the heart of a point when there are a dozen side-points whose importance isn't immediately obvious while one is taking a stroll through all of them. On the other hand, there are those who relish the long-form answer, and I guess you could say I tend to cater to that audience. Whether you are among it, alas, I'm not sure, and if not I apologize. Put another way: all of these points seem interesting and helpful to me, and maybe not all of them are to you, but I don't know which ones to cull so I'd rather leave them all in just in case; it's casting a wide net and hoping something sticks.

[1] Again, this can be hard to know without knowledge of what all is going on with these things. And, alas, this isn't really a good place to ask about them; SE is not a one-stop shop for all your information needs. Much knowledge is gained by everything from textbooks and class time, to practical books (some on power supply and amplifier design would be relevant here), to just plain old word-of-mouth and on-the-job experience. Obviously, quick question-answer posts cannot possibly substitute years of experience. So, I understand if there's some frustration here.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you, Tim. I considered all your advice and formulated a new question. It mostly generated snarky, unhelpful responses, but at least it didn't get closed. 😊 I guess this just isn't the right site for such questions, and I'll head over to EEVblog instead. \$\endgroup\$
    – Sara
    Jan 4 at 2:24

You must log in to answer this question.

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