Grey beards and boffins; what do see when you see a circuit.

[McMurdo]

a true circuit diagram much easier to follow than a diagram which tries to place components in their mechanical location - that usually leads to many crossing lines and a great loss of clarity

A similar philisophy to Harry Beck of London Underground fame!

[jamesperrett]

Originally Posted by Diabolical Artificer

The best circuit description I've ever come across was in a Tascam 38 R2R manual. the full schematics were shown, and then each individual function was clearly explained, IE "when the play button is pressed, Q1 switches on, turning Q2 off, which in turn also turns off Q5. IC3 pin 6 then goes high, enabling....." etc, hope you get the picture. Each description was accompanied by a small schematic, showing voltages etc. I think I gave it to Michael Maurice, wish I'd have kept it, it was an education in itself.

The Tascam 38 manual is available on the net. Here's one link

https://www.vintageshifi.com/reperto...ice-Manual.pdf

[unitelex]

I think you can pick up some of the character of the equipment quite quickly by looking at the schematic even if you have not see the product. How many valves does it have?, (budget vs high end set) Which series of valves were used. Component choice often gives away the manufacturer's nationality and design priorities BC... vs 2N... vs 2SC... and also the era of the design of course. Sometimes you can guess the OEM by the circuit, you can certainly make an assessment of cost vs quality. Some of the older hand drafted schematics are interesting in that you can often see the revision box history, how many changes were made and over what timeline, it can give some interesting insight into the original design process.
As others have said, I first look to recognise the stages. Some schematic layouts make this easy by separating and even labelling the stages, others are awful and condense/merge the stages in the drawing making it difficult. It helps me to see the resistor and capacitor values on the schematic as their values help to confirm the understanding of the coupling paths. So I guess R and C values are more useful to me in understanding the functions than voltages and currents, or put another way the resistor values allow one to estimate the currents. As others have said voltages and power levels are needed at the repair stage.
One feature sometimes used in schematic drawings I particularly like. I have see this on some early PYE transceivers and more recently on Panasonic schematics. The main signal path is shown in bold lines or in an alternative colour. Makes it very easy to see the flow.
I am not keen on schematics where there are many layers of hierarchy. At most I would have a top level schematic/block diagram with key signals interconnect shown and then a single layer of pages underneath. I am not keen on tracing signals up and down several layers of hierarchy, often resorting in re-drawing the section of interest on a single sheet. I would rather have a cramped schematic than too much hierarchy.
Some designs are standard fare and somewhat dull, others have unique or clever features, clever ways to reduce component count eg single stages that process both RF and AF simultaneously or switching a valve between completely different functions according to a mode/band selection etc. It is the exceptions that get my interest.

[jjl]

Originally Posted by merlinmaxwell

It is the way I see currents to through them, almost in colour (blue for caps.).

Electronic synaesthesia - that's interesting, and I can imagine that one could find it helpful. My daughter tells me that she sees certain numbers as colours.

John

[Hartley118]

Originally Posted by McMurdo

A similar philisophy to Harry Beck of London Underground fame!

It’s no coincidence that Harry Beck was an electrical draughtsman well experienced in the clear drawing of complex electrical signalling systems.

Martin

[Malcolm G6ANZ]

The worst manual I had to work with was American! It had the board layouts and circuits of the boards in one section. But the way the boards and racks were connected were in another section. So to follow a signal path you had to find the labelling where the signal left the board ,crossreference that to the interconnect diagrams. That led you to where the signal entered another board or distribution point. Then back to the board diagrams and trace to where the signal left the board ,back to the interconnect and so on.
It could take a long time to trace a signal path.

[Electronpusher0]

Originally Posted by Malcolm G6ANZ

The worst manual I had to work with was American! It had the board layouts and circuits of the boards in one section. But the way the boards and racks were connected were in another section. So to follow a signal path you had to find the labelling where the signal left the board ,crossreference that to the interconnect diagrams. That led you to where the signal entered another board or distribution point. Then back to the board diagrams and trace to where the signal left the board ,back to the interconnect and so on.
It could take a long time to trace a signal path.

Drawn for ease of manufacture, not service

Peter

[Lucien Nunes]

As alluded to in many posts above, a circuit can be read at different levels of granularity that explain its operation in ways that are optimal for different purposes. E.g. in a troubleshooting situation:

a) 0.2mA flows into the base of Q7, which turns on and sinks the 2mA sourced from Vcc through R18. It saturates, reducing the voltage at pin 2 of IC3 to about 0.2V.

Working with the currents and voltages is necessary if something in a known stage is a bit adrift and you need to find exactly which part is out of spec. But at this level of detail, you would need to absorb a lot of information and side-step many distractions to see the big picture.

b) Q7 pulls one input of IC3(a) low.

Working with logic states and signal levels is sufficient to trace the progress of analogue or digital intelligence around a circuit. The analysis takes certain things as given, such as that a logic input must swing between certain thresholds and that the designer has allowed for this. You can home in on those specific voltages and currents later when something is anomalous or suspect.

c) Q7 inhibits record.

Working with functional-level descriptions is quick and easy and allows you to take in much of interest in seconds, although it requires you to turn a blind eye to most of the actual electronics. This analysis takes for granted that TTL is often used active-low, in fact does not even care what kind of logic is used, just that it exists and in this case inhibits record when asserted. That fact is not evident at all in description a).

One important skill is tuning the focus of attention fluidly to varying levels of detail, narrowing down to discover specifics when relevant, widening out to gloss over them when not. It's like tweaking the Q of a filter to allow the right passband of information. With practice, you can go further and have mental filters with different bandwidths all active at once, at which point oddities like the Fostex signal switching example leap off the page. You might be trundling along looking for a problem at 'functional block' level, but the 'voltages currents and impedances' parsing is still going on in the background. When the JFETs come into view with just diodes connected to their gates, the background analysis flags it up as 'does-not-compute' and diverts the main search down to component level to discover what is wrong or not understood about it. For me it triggered a sanity check: 'is this a block diagram indicating a FET signal switch, is it a drawing error, or is it what they have actually built on the PCB?'
The mind boggles.

[CambridgeWorks]

Post #23
Later Pye Telecom manuals introduced the colours. Even the R, C, and Q? was it for semiconductots in lines at the bottom of the page for easier finding of actual components. Earlier ones, even back to the 1950s were still very good layouts. Overrall, in non-domestic equipment I feel their manuals were some of the best.
I have in my time used some horrendous diagrams for extremely complex equipment. Such as a short line from an ic with an arrow tip marked R351 as an example. This then involved searching maybe 4 other pull out sheets to find R351. It could be next to R215 and R233! You needed quite an educated guess to locate it. This is where my own manual started to receive my own coloured markings for guidance. It no longer was a "Registerd Copy" within the department because of this. Then there were the mistakes that were found!
Rob

[unitelex]

I am also not a fan of schematics without a ground busbar, you know the ones where they have multiple individual ground symbols for each grounded component littered randomly on the schematic.
I prefer to see transistors with the circle because this is how they were shown in the electronics magazines in the 70s 80s when I was learning,but it no longer bothers me for transistors if it is missing. For some reason I find missing oval/circle around valve symbols just looks wrong.

[TonyDuell]

I prefer circles round transistors, but curiously I don't care about them on diodes or LEDs. I also prefer the zigzag line for a resistor and the hooped line for an inductor/winding rather then the rectangle and thick line used on modern diagrams. I guess it's what I learnt from all those years ago.

I also prefer a multi-page diagrams with sensibly labelled interconnection signals to cramming everything on one large page. In fact I often prefer named interconnections to a long line on the same page, it's easy to jump from one line to an adjacent one and end up in the wrong place.

As regards grounds, I like a ground busbar for a particular stage or block of mostly discrete componets. But a digital system, with all sorts of enables, etc, grounded (or tied high) will be a right mess with power and ground buses rather than symbols.

[Malcolm G6ANZ]

Ground lines were a problem. Mainly because there were usually two or sometimes three. There was analogue ground separate from digital ground which was different to signal ground

[Lucien Nunes]

Working with a diverse range of equipment types offers exposure to many different drawing conventions. Rails top and bottom for radio / TV / Audio, ladder diagrams for machine control panels, detached contact for relay auto telephony, hierarchical mixed block and signal for extensive analgoue systems, etc.

Within each format, I generally prefer the layout and symbology I grew up with. Zigzags, circles round transistors, HT at the top etc. I try not to actively hate the alternatives but do find some styles, such as USA analogue schematics with signal flow strictly horizontal (with sideways tube symbols and the B+ snaking around to keep out of its way) rather tiresome. However, as with learning multiple languages, I believe that the mental exercise of absorbing new features and conventions strengthens and develops facility with the others.

One interesting aspect of circuit diagrams that depict purely electronic devices built from discrete components is that the symbols capture and convey all that is needed to gain a general understanding of the system when correctly read. When non-programmable ICs are used it is often necessary to understand details of what goes on within the black box, but again this can be looked up and the overall behaviour of the system gleaned. In contrast, a system that is heavily interactive with non-electronic parts, such as a complex electromechanical machine, requires understanding of the electronics within the non-electronic context. If you had never seen or heard of a jukebox before, you would struggle to understand what it was all about just by reading a circuit diagram entitled 'Wurlitzer Phonograph Model 2000.' Likewise if you knew nothing of electromechanical lift controllers, it would not be obvious that a relay marked 'DJR' should never operate, even though it is energised all the time the lift is moving. It is this interaction between a 2D circuit diagram in the conceptiual plane and the real 3D world of the equipment that it depicts that I find most fascinating.

[Radio Wrangler]

Ziggy resistors and loopy inductors, please

Save rectangular boxes for generalised impedances.

David

[TonyDuell]

I forgot to mention my pet hate : multi-coloured circuit diagrams -- components one colour (or several colours according to type), connections a different colour, text another colour, etc.

I find them very tiring to read and understand.

[Hartley118]

I've always struggled in trying to follow the many and varied schematic representations of multipole multiway switches. It's often so difficult to get the head round the actual connections made in each position of the switch and relate them to the actual wiring.

For me, the extreme version of this challenge is a telephone exchange schematic which requires a special frame of mind to relate to what's actually happening in the hardware.

Martin

[kalee20]

Originally Posted by Radio Wrangler

Ziggy resistors and loopy inductors, please

Save rectangular boxes for generalised impedances.

Seconded!

Luckily, in the Day Job, I have the luxury to be able to do just that!

[Craig Sawyers]

Back sometime in the 80's I was designing a hospital bed interface to the automatic switchboard for an American company. I quickly found out that the interface standards were set out in the Bellcore documents. So I bought the relevant ones, and was greeted by a whole lexicon of acronyms.

So I had to buy another document - "The 500 most common acronyms in the Bellcore standards".

I quickly found the one that was used most frequently - POTS. Plain Old Telephone Standard.

The whole system was designed to confound the designer.

Craig

[Graham G3ZVT]

Originally Posted by Radio Wrangler

Ziggy resistors and loopy inductors, please

Save rectangular boxes for generalised impedances.

David

Amen to that David.

Where do you stand with the little U when two wires crossed?

The presence or absence of a dot which may not be too clear on the badly scanned circuits we so often have to use.

[Boulevardier]

Originally Posted by rambo1152

Amen to that David.

Where do you stand with the little U when two wires crossed?

The presence or absence of a dot which may not be too clear on the badly scanned circuits we so often have to use.

I passionately agree! The crossovers and junctions can be just about impossible to tell apart. My life has just gone right downhill since the little loops were senselessly abandoned.

Mike