I think you gild the lily a little when you say:
"If you are a hobbyist just starting out then a commercial scope may be out of your reach. Even if you are a professional you may not feel that you can justify the expense so that you can dabble at home at the weekend."
Some people reading this statement might believe it. Frankly, I can't imagine a professional being happy with a piece of equipment with a bandwidth of 10kHz. Sure, it can serve some useful purpose, but it's three orders of magnitude lower in bandwidth than the cheapest of scopes. As a comparison, it's like saying if you can't afford a $400 Fluke multimeter that can read up to 1000V, here's a way that for $20 you can build one capable of reading up to 1V.
On a positive note you do say that:
"Audio signals range between 20 Hz and 20 kHz . So theoretically 20k Hz is the upper range of frequencies that you will be able to display. In practice anything over 3 – 4 kHz will be displayed with some noticeable distortion but you should be able to see at least the presence of a 20kHz signal."
You also mention the AC coupled nature of the input, which is another issue.
I'm not going to argue that this is not your experience:
"I know that during my entire career in electronics and software engineering my main use of oscilloscopes has been to examine signals of a few kHz. Higher frequency use has been a rarity for me."
But I can't say that it's my experience. Note that I pretty much NEVER do anything RF related, mostly microcontrollers, relatively slow logic stuff and goofing about with switchmode regulators. Both the AC coupling and the low bandwidth would cause me problems. These range from not being able to see actual logic levels to not being able to see fast rising edges. Whilst these things may be areas that you wouldn't recommend its application, there's a lot of people out there playing with microcontrollers as their first introduction. Far fewer people are building multivibrators and small class A audio amplifiers.
One major thing you don't mention as a limitation is the input impedance of this device. This further limits the utility when measuring signals in high impedance circuits. In fact you do the opposite by saying that the 470k input impedance is "reasonably high". Additionally you recommend in some cases reducing this input impedance to around 10% of this value. At this point you have an input impedance of around 50k. Using this in audio circuits (arguably where its limitations are going to be the least obvious) may result in signals being significantly attenuated.
Perhaps this is partially a terminology issue, but I found this amusing:
"Most top electronics engineers have a scope permanently on their bench and they use it practically every day. If the software engineers don’t have one then they are wishing that they did because sooner or later they will need it."
I'd say that by "top electronics engineers" you mean almost anyone from moderately well equipped hobbyist on up who needs to see the inner workings of their electronics. And by "software engineers" you probably mean something other than "software engineers" (software engineering is an attempt by the engineering guilds to label traditional programmers as something akin to barber surgeons). Software engineers are typically another form of desk-jockey who doesn't regularly see any form of hardware naked, and often wouldn't know an oscilloscope from a telescope. What you're probably thinking about are people working on embedded systems and the like.
I also have some stylistic issues with some of your schematics. Take this one for example:
You use different styles of diode for the LED and the zener. I don't refer to the fact that one is drawn as a LED and the other as a zener, but the fact that one has a line drawn through the anode and the other doesn't. The resistors are the horrible boxes. Whilst that is simply a preference, the use of different size boxes is a problem.
I also note that you don't show junctions clearly. Whilst this isn't an issue with this circuit that doesn't have crossing lines, it's non-standard.
Both of your transistors are drawn... oddly. One should be rotated (or have its load moved below the emitter) and the other has the emitter and collector not being aligned for some reason.
While I'm at it, the ground rail and some other unnamed rail wander off, the battery is shown as a cell, it is typical to use Q for transistors, you use 4.7k (as an example) instead of the more "fault tolerant" 4k7, the switch is not labelled, and the whole circuit could have been replaced by a single three terminal regulator.
Whilst none of this is particularly important on its own, it doesn't impress me as being designed by someone who is well skilled. Now, I'm not saying that you're not, but the layout of the circuit seems sloppy.
On a technical note, it seems you're using Q2 (sorry, TR2) to remove the V
BE offset from the output voltage. I question whether this is an issue, or why you don't simply tie Q2's base to its collector and dispense with R4 -- unless this is designed to limit the available current.
Here is what I consider to be a somewhat incomplete improvement (I've left out the battery and the switch for one thing)
I think it's worth telling readers that almost any small signal NPN transistor will work. For our American cousins, specifying that a 2N2222 could be used (or even a 2N3904) would make them feel more comfortable (I think they feel that BC547's are un-American or something). There's probably no reason not to use BC548's (which are what seem to be the generic jellybean NPN transistor in my neck of the woods). Of course this might raise the issue of pinouts, since they don't all share the same one.
I've really only skimmed your book, and spent a little time on things that stuck out to me. I have no doubt that this could be useful to people, but I question whether they will really understand its limitations. Nevertheless, it's one more book than I've written.
, what's it called Steve?
