Working without Schematics?

Whatever you design/produce - some poor sod later on has to maintain it.

Even if I don't need a drawing to work from - (s)he probably will. So I document everything and the easiest way to do that is start with the documentation..

But it isn't always for the best. One dear sweet engineer at one time thought it would be kind to use the HT line in a valve servo amplifier as the zero reference. As a young apprentice I found out that the line marked as 25v was, in fact, rather hotter..

-- Sue

I have never seen a schematic diagram used at McDonald's (hamburgers). On the other hand, I would be very surprised to see a TV repair man not using a schematic. What is your point?

Bill

-- Fermez le Bush

What is your point?

I would like to know what other people are doing in the respected industries?

I repair circuit boards where I am employed. I would love to have schematic diagrams for everything I work on. But it ain't gonna happen..

Spokesman a écrit :

I'm doing schematic for years and I never made a cubicle without schematics. On the other hand if you want to make a little box with just a motor starter, it is possible to make the box, but it will be time wasting for others to understand your work. Janos

I spoke to some people at a local electrical utility about a consulting opportunity to improve their substation design configuration control problems. It turns out that the construction crews have so little regard for the engineering department (due to some past personnel problems) that they just build the stations and let engineering do as-builts.

I do engineering problems and preocess improvement, not solve HR problems, so I walked away.

Schematics are sometimes useful to help folks understand a circuit. Beyond some point (talking digital design here) you "blocks" go beyond simple gates, counters, etc.

I was on a job where the designer was a literal genius. He created and had built a encoder for an error correcting code with little or nothing in the way of schematics but just a wirelist which he turned over to a technician to wirewrap (this was about 25 years ago.)

If it's not necessary for someone to "debug" a design there is no reason for anyone to truly understand it and no reason for a schematic. A net list or equivalent is do the job.

But a schematic that just shows where every little wire goes but doesn't have notes and such saying why does what, etc., is just a waste of time and paper.

ah, but the poor tech that is paid to fix the gizmo needs something to go on. a schematic, parts layout, and parts list are much appreciated. the "theory of operation" section of a manual is sometimes helpful to even an experienced tech but is mostly unneeded for typical repairs.

Well, a wire list will show where everything is. If the gadget worked at one time and stopped working a "tech" can proceed on the assumption that a single IC or whatever stopped working right. Having the specs of the ICs and a scope will quickly find a "dead" part.

But I think you agree that while documentation is desirable, a schematic may be overkill.

I recall running into a control panel at a power station. The circuitry had had many modifications and the actual and the documented situations weren't the same. I spent a lot of time on my backside, under the panels to draw what was actually there and more time to make good "wiring diagrams" and a functional circuit diagram of what actually existed (until the next time someone made an undocumented change).

The crowning glory(??) is automotive circuits- a schematic is definitely needed. However, Volkswagon bus/ camper schematics are not all that helpful - sort of an aid to differentiate between utterly hopeless confusion and "why the hell did they do it this way" (e.g. a gas heater control which was tapped off the field circuit of the alternator).

Don Kelly snipped-for-privacy@shawcross.ca remove the X to answer

----------------------------

an assumption that may waste hours or days. multiple failures in a system are common enough. have you ever heard the electronic servicemans adage: first fix the power supply, then fix the rest of the problems" ?

you have never actully done this have you? i mean traced faults to component level.

i can think of few instances where a schematic would be "overkill" (a nebulous term). maybe a flashlight.

here is a real world example: an expensive gizmo used a 12 volt regulated linear supply. a power surge blows the bridge rectifier shorts the pass transistor, fries some resistors, and shorts a zener diode.

your mission, Mr. Phelps, is to fix it quickly and economically. that means make it like new and not redesign it. without a schematic you may not even know its a +12 supply. you don't know the zener rating, you can't read the colors on the resistors and you don't even know the amperage and PIV of the rectifiers.

in a neraby thread somone posted a link to a motor schematic. that showed the motor was DC not AC. without that everyone would still be guessing.

i will admit that the time a mother brought me her sons science project, a lemon powered buzzer, a schematic was superfluous. i wrote on the service tag "traced problem to dead lemon" "$12.50 bench charge"

My argument for schematics has two parts:

As a tradesman, when I would find something blown up during routine calibrations - I could be sure it wasn't me because I knew exactly how I hooked into the system by using the drawings, this would keep the engineer off my back, although sometimes the engineer still would assume I blew it up!

I've seen guys blow fuses because they didn't know the wire they were lifting could become live even after a pot test.

I've seen an operator almost explode because one of his alarms came in when a fuse was popped by a tradesman that left his meter on current when measuring voltage, had no idea what else happened because he had no drawing, he replaced the fuse and took a lot of sh%$ because he never called the operator.

As Engineer In Training - I have an incredibly hard time writing maintenance procedures with out the drawing. Because of the "old days" when maybe drawings we not that important, I find drawing mistakes all the time and have to fix them before I move on to the next job. This part of my job sucks - trying to figure out what guys did 30 years ago.

I do agree that there are times when drawings will get in the way, and sometimes it helps to just troubleshoot using what you know - even then drawings can help.

It has to be true that not keeping the drawing up to date is unacceptable in any plant environment.

Are there any standards out there that require all drawings must be kept accurate and assessable at all times! Maybe an IEEE standard or ISA, or NFPA- this would really help my case!

This is a "finger pointing" problem. In your case, the engineer should find a tech he trusts or "do it himself."

In my first post in this thread to mentioned an essentially undocumented design by a true genius. We had a system wide test that could easily determine whether this "board" was working but no obvious way of finding a failed IC or whatever if the board failed a system test.

The problem was complicated (as if the orginal problem wasn't bad enough) by the fact that our true genius when he designed the board knew he had some "races" which are a condition whereby a signal may change state too early. Unfortunately the genius "solved" the race problem in one part of the board by just inserting some "gates." (Another race was solved by a few inches of twisted pair wire AKA a "delay line.") Delay lines will have the same delay forever but gates? The nightmare was that because of some subtle change in the manufacture techniques, newer parts would have different (but "in spec") delays and that "new" boards just would not work over the temperature range.

I "moved on" while this issue was still being discussed by "management." I suspect they just said, "Ship It!" And hoped for the best.

That's why sometimes it's easier to "start over."

Believe it or not, when the old Bell System started to bring in "digital" switches a BIG motivation was that the old relay and electro-mechanical based systems required that a LOT of relays be adjusted by SKILLED techs and it took someone with years of experience to "debug" a system with problems.

One of the "Lessons Learned" from the "PC revolution" is that often it's cheaper just to replace the whole box that try to figure out WTF went wrong.

Keeping drawings (even or expecially CAD based drawings) up to date is about an order of magnitude more expensive than to just maintain an accurate wire list.

Look at it this way: you have secrets in your head. Unless and until your bosses pay to "start over" they will need you.