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

[David Simpson]

In many cases, the rule of thumb, reading radio diagrams - wise, is from left to right. I.e. Aerial & RF stages on the left,IF stages in the middle, and detector & audio stages on the right. Mains Transformer & Rectifier circuitry - bottom right. Looking down onto the chassis from the rear, this is also mostly the case.
Turn the chassis upside down, and as many others have experienced, one often encounters a higgledy piggledy layout of wiring. Some radio manufacurers even used to use spare valve holder pin tags for mounting components !
Myself, and perhaps other ex military folk, were used to terminal strips, paxolin tagboards & neat looming & lacing.
Many of the circuit diagrams out there in the VR fraternity are in A4 format, and also many VR folk use the BVWS "Trader Sheets". As do I, but the first thing I do on acquiring a diagram - is to go to the local chemist, who have a scanner/printer, and get them to blow it up to A3. Far, far easier to read.
I then put these blown up diagrams into A3 acetate or polythene sleeves & store them flat.

Regards, David

[Electronpusher0]

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.

Junctions should be staggered, if two wires appear in line they should cross.

Peter

[Radio Wrangler]

Originally Posted by rambo1152

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.

You've got me bang to rights. I tend to just have blobs. It's habit. I draw half-loops where lines cross without connecting where there's space to get them in, but I blob crossing lines which do connect. Habit, really. I suppose splitting one line with two 45degree bits to get two blobs in might be clearer.

I'm not consistent and sometimes turn out crossing lines straight over without loops or blobs and then turn out crossing lines which do connect with only a blob.

Sorry!

Dvid

[Chris55000]

Hi!

This discussion brings me to another point – depiction of valve electrodes!

It was traditional for more years than I care to count, to insert a black spot at the point where the electrode symbol met the envelope, Newnes (P.W. & P.T.), Caxton (R and T), Data Publications (R.C.) Illiffe, (W.W.), Odhams (Radio & Elec Repairs, Gordon King's books) all advocated the spot, but when BS 530 came out in 1948, they deprecated the spot, instead they recommended a thicker valve envelope symbol with the electrode connection drawn straight out across it.

Obviously the spot was intended to represent the base–pin but I wonder why B.S. deprecated it?

If you look at anything drawn by Bush in the 1960s, Murphy Radio (rectangular valve envelopes) and Solartron and Cossor amongst the 'scope makers, plus anything from the B.B.C., these all used the B.S. convention, which I'm now in the process of adding a new suite of valve symbols to my sPlan for!

Chris Williams

[ms660]

There was no standard way for schematics, each manufacturer to their own, valves one way up, valves t'other way up, valves sideways this way, valves sideways that way, the way they're drawn makes no difference to the workings of the circuit.

For the service tech, the Trader sheet method for showing the switchery was the best, the worst for me was some of the Philips attempts.

Lawrence.

[David Simpson]

As well as blowing A4 circuit diagrams up to A3, I also follow David's practice of "U's" & "wee blobs". If its going to be a long repair/restoration project, or the circuit diagram has been printed confusingly, I also colour-in parts of the circuitry :- HT = Red, Anode circuitry = Blue, Grid circuitry = Green, Cathode circuitry = Yellow, & Heater circuitry = Brown. These coulours were a sort of convention in some 30's/40's/50's military equipment & some domestic radios.
Some folded "pull-out" circuit diagrams from Manuals & their scanned copies often take up two or three A4 sheets. Hence can be a bit difficult following them through, stage by stage. However, some judicious snipping with scissors & careful joining with sellotape, can produce a decent A3 size diagram covering everything. Although, as many folk have experienced, often scans of copies of copies, dont always match up, so circuitry lines sometimes end up with "dog's legs" joints. Again, colouring-in often helps avoid confusion.
I've been lucky in acquiring one or two "original" "yellowish" factory circuit diagrams which are approx A2, and they are a dream to follow, & I keep them pristine in A2 sleeves.
Military Manuals & AP's are great to use, as they often contain detailed seperate sections of circuit diagrams. As David will tell you - HP's Manuals are similar. As are Tek Manuals.

Regards, David

[TonyDuell]

Originally Posted by Boulevardier

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

When I'm drawing a circuit diagram, I combine the 2 conventions. A half-circle hop-over if it's not connected, a solid dot on the cross if it is.

This is one of the few things where I disagree with The Art of Electronics. In said book they tell you not to use the half-circle, it's very old-fashioned. I care less about that than having an unambiguous diagram.

[kalee20]

Originally Posted by rambo1152

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.

Hand-drawn: I use a little loop where wires cross.

CAD: If using a program which does lines, circles etc then I make a little loop. But if using a proper electrical CAD package, the facility regrettably just isn't there.

Always, if there is a connection, I either stagger the junction, or put two short 45° angles, so that it really is clear. Two lines crossing at right angles is just too ambiguous by the third generation of photocopying!

[kalee20]

Originally Posted by Chris55000

It was traditional for more years than I care to count, to insert a black spot at the point where the electrode symbol met the envelope, Newnes (P.W. & P.T.), Caxton (R and T), Data Publications (R.C.) Illiffe, (W.W.), Odhams (Radio & Elec Repairs, Gordon King's books) all advocated the spot, but when BS 530 came out in 1948, they deprecated the spot, instead they recommended a thicker valve envelope symbol with the electrode connection drawn straight out across it.

I've never done that, not for valves, not for transistors, not for bulbs, not for anything.

It takes time when hand-drawing, and adds nothing as I see it. Straight through the envelope, and to the destination goes the line!

[duncanlowe]

I'm not sure where I got the habit from, but when drawing crossing but not connected lines, I don't use the loop, nor just cross the lines. On one of the lines, I leave a small gap either side of where it crosses the other line. For connected lines I still use the blob.

When looking at a circuit, I'm usually trying to figure out what it's supposed to do and whether it's likely to do it properly given the environment it's in. The most interesting of these is when I don't have a circuit but am looking at a PCB and trying to understand what it's actually doing compared to what it's supposed to do. For example I had an antenna module (for receiving 433MHz) that was apparently misbehaving, but I was promised it was passive. Didn't take long to break open and figure out there were active components on the PCB meaning it had gain, and therefore was suffering from feedback from it's own cable (that had been cable tied to it). I suppose I spent much of my life as a "Reverse Engineer".

[Radio Wrangler]

While following a line on a complex diagram that uses little loops, you can say "Boing, Boing, boing..." to yourself and thereby dodge the appearance of getting old too quickly.

It also enhances your personal mystery factor with any witnesses.

David

[ms660]

KISS...No jump over hoops....Connection dots with a diameter at least 4x line width.....and for the more complex schematics the same plus connection highways with junction ID's and direction indication.

Lawrence.

[TonyDuell]

Originally Posted by kalee20

I've never done that, not for valves, not for transistors, not for bulbs, not for anything.

It takes time when hand-drawing, and adds nothing as I see it. Straight through the envelope, and to the destination goes the line!

To me, a dot on the envelope means a connection to the envelope. It would make sense for metal valves, for example. I use that convention if I need to indicate a connection to the shell of a connector, for example to show that the shell of a DB25 socket is connected to ground.

[G6Tanuki]

For simple [i.e. not synthesized/computer-controlled] radios, amplifiers etc, my mind instinctively first identifies the active devices then visualises the signal-flow through them. Typically 'signal' flows from left to right.

This doesn't really work when the schematic occupies a dozen or so pages. It also doesn't work so well when looking at something like a frequency-synthesizer, counter or similar.

I wonder where the 'convention' of putting the power-supply on the right of the schematic, with the power entering on tthe rightmost side - came from?

In some situations - such as significantly-complex regulated power-supplies - it would make more sense for the diagram to be drawn so *power* flows from left to right.

[Radio Wrangler]

I don't think the original post was asking about the draughtsmanship aspects of a circuit diagram, I think the question was about the appraisal of the design illustrated in the diagram.

In the example which kicked this off, I'd looked at the way some JFETs were being used as analogue switches for an audio signal flow, and I'd had one of those 'Oh Bu**er, that's not going to work!' moments. Closely followed by a 'What idiot thought it would?' moment

It's a matter of looking past the draughtsmanship and at the design itself, seeing it working in your mind, recognising sections of it and knowing where they came from. So it's partly animation of it functioning and partly seeing it in the context of a family tree.

Poor draughting and lousy layout can make it painful to read a diagram, but the real trget is the hardware it documents, however well or badly.

David

[kalee20]

True. Though, a well-drawn circuit diagram does make it easy to follow.

What do I see? A load of components, hopefully familiar - and a load of connections.

If the components look familiar (R's, C's, L's, diodes, valves, transistors, op-amps) then I try to see what it is - or break it down to circuit segments, which are recognisable.

Usually, separate bits of an overall circuit are amplifiers, oscillators, rectifiers, or logic. But not always. Again, component values make the difference from a circuit being an RF amplifier with a clever bit of neutralisation, or an oscillator.

Component values are important, and that comes with experience. A capacitor from a signal line to ground, for instance, may be a treble cut. But it could be just an RFI suppressor. A glance at the value, and having an idea of what the circuit is of, gives me a strong clue. So, having values on the circuit diagram (as long as it doesn't get too cluttered) is a big help.

Seeing FETs with a badly-defined DC path for their gates, as RW observes, says that the circuit relies on luck to work, and makes me look for any other no-no's.

If the components are not familiar (a rectangular box with a lot of connections and labelled ICwhatever) then the only thing is to identify the box, and look up the data sheet for the IC type.

[Radio Wrangler]

That's a thought, I REALLY hate diagrams without values on them, the sort that have you flicking backwards and forwards to a parts list.

I know why they do it... it makes documentation easier to update and easier to be sure there aren't any contradictions left lurking.

But a diagram is drawn once, and then should be updated only infrequently. But it will be read many times.

So the best deployment of labour is to keep it with updated values drawn right by the part, so that a bit of labour can be saved on each and every time the diagram is looked at.

David

[G6Tanuki]

I guess the values-printed-on-the-diagram or referenced-separately thing is a matter of preference. Personally I rather like the Military approach where there are letters/numbers along the horizontal and vertical sides of the diagram which are used as references to the values/parts-list.

Printing the full specification of each part on the diagram is not really possible - yes you can print "10nF" next to a capacitor but you still need to refer back to the parts-list to get the voltage-rating and construction/dielectric-type.

[In the early-1990s I worked - briefly - with a really-nice Intergraph CAD system; you could mouse-over any component on the schematic and a little pop-up box would then appear giving component location-on-PCB, its parameters/values, NATO stock-numbers, current preferred-supplier/price etc. - you could then right-click on the mouse to pull-up a small picture of the part as-placed-on-the-PCB].

[Hartley118]

Originally Posted by Radio Wrangler

That's a thought, I REALLY hate diagrams without values on them, the sort that have you flicking backwards and forwards to a parts list.

I know why they do it... it makes documentation easier to update and easier to be sure there aren't any contradictions left lurking.

But a diagram is drawn once, and then should be updated only infrequently. But it will be read many times.

So the best deployment of labour is to keep it with updated values drawn right by the part, so that a bit of labour can be saved on each and every time the diagram is looked at.

David

This is where I could really do with a ‘like’ button.

Martin

[trh01uk]

Originally Posted by Hartley118

Valve circuits have for me a useful clarity in that the valve symbol is instantly recognisable as the basic amplifying block in the circuit. After all, it's amplification that's at the heart of electronics. The valve may of course be sitting in a circuit that uses its gain to oscillate, or uses particular electrodes in an RF mixer circuit, but, as long as the signal flow is shown in the conventional left to right form, it's not usually too difficult to work out what's going on.
Martin

One famous example which is an exception that defies your very sensible comment is the detector in the German E52 "Koln" HF receiver, which was produced during WWII. The receiver famously insisted on pressing a single valve type - a small signal pentode type RV12P2000 - into all stages of the receiver (except for the HT rectifier). This was to ease repairs in the field apparently - no plugging the wrong valve into the wrong slot!

This detector circuit will defy most attempts at analysis until you understand how the normal pentode elements are "redefined". So the IF signal is injected into the pentode's suppressor grid, which become in effect the diode detector anode. The actual pentode anode functions as another diode for delayed AGC. The pentode's control grid receives detected audio, and is part of a "triode" AF amplifier, the "anode" of which is the pentode's screen grid!!

You have to ask what the designer of this little gem was on.....?! Anyway, I enclose the only schematic I could find on the net at the moment. And this one has been re-drawn to hide the fact that the valve is a standard pentode - the original E52 schematic offers no such "help"!

To give due reference to this schematic, it came from: http://www.noding.com/la8ak/e43.htm