This thread is for discussion about the new book and any errata. It was a rewarding and learning experience for me helping Bob by reviewing the manuscripts. But none the less a few errors got by me and were introduced past my review, as I discovered when the book was in hand.
Cheers to Bob for his dedication in creating a large collection of information for us to enjoy 😉
Errata:
1) Chapter 3, p56, Figure 3.16(c) should be "ring-of-two LED" since D1 and D2 are shown as "Blue" and "Green"
2) Chapter 15, p336, Figure 15.13 was changed from my review. Text in book was for attached fig 15.13
3) Chapter 15, p339, Low-Frequency Monoaural Merge, first paragraph, last sentence, "The outputs are then bridged with (R3) to mix the signals at low frequencies"
4) Chapter 18, p396, Figure 18.11, should read "THAT 1510/1512 IC Microphone Preamplifier" to match what's shown in the figure
5) Chapter 21, p483, Figure 21.6, the function block should be labeled as "TA7328" and not uPC1237. uPC1237 example as attached
6) Chapter 26, p616, 3rd paragraph, "Applications for compandors can be found in Philps/Signetics NE570/571/SA571 AN174 and Onsemi AND8159,AN8227"
7) Chapter 28, p642, section 28.10, to clarify a FPGA is defined as a "Field Programmable Gate Array" but it's true that they are logic elements
Cheers to Bob for his dedication in creating a large collection of information for us to enjoy 😉
Errata:
1) Chapter 3, p56, Figure 3.16(c) should be "ring-of-two LED" since D1 and D2 are shown as "Blue" and "Green"
2) Chapter 15, p336, Figure 15.13 was changed from my review. Text in book was for attached fig 15.13
3) Chapter 15, p339, Low-Frequency Monoaural Merge, first paragraph, last sentence, "The outputs are then bridged with (R3) to mix the signals at low frequencies"
4) Chapter 18, p396, Figure 18.11, should read "THAT 1510/1512 IC Microphone Preamplifier" to match what's shown in the figure
5) Chapter 21, p483, Figure 21.6, the function block should be labeled as "TA7328" and not uPC1237. uPC1237 example as attached
6) Chapter 26, p616, 3rd paragraph, "Applications for compandors can be found in Philps/Signetics NE570/571/SA571 AN174 and Onsemi AND8159,AN8227"
7) Chapter 28, p642, section 28.10, to clarify a FPGA is defined as a "Field Programmable Gate Array" but it's true that they are logic elements
Attachments
Last edited:
Nice cover!
But sadly $162AUD won't make it past the Domestic Minister of Finance. 😥
But sadly $162AUD won't make it past the Domestic Minister of Finance. 😥
Routledge has a 20-25% sale going on right now. https://www.routledge.com/Designing-Audio-Circuits-and-Systems/Cordell/p/book/9781032010892.
Just ordered my copy.
Just ordered my copy.
The above routledge link gets redirected to some other page.
Other than that, have discussed the book with a couple of pro audio designer friends. We all have our own copies to peruse. Pretty much independently ended up with about the same point of view on it. However would still like to see a plainspoken, evenhanded print review. If anyone happens to see one, please let us know.
Other than that, have discussed the book with a couple of pro audio designer friends. We all have our own copies to peruse. Pretty much independently ended up with about the same point of view on it. However would still like to see a plainspoken, evenhanded print review. If anyone happens to see one, please let us know.
Last edited:
An FPGA also contains flip-flops. I don't think that's technically a logic element. It's definitely not a logic gate.
I hope some feedback is allowed here...
I read through the chapter on MC phono stages (Ch. 16). I think it would flow better if the equation for thermal noise had been stated along with a table that showed the thermal noise for an example list of resistors. In its current form, Cordell ends up basically describing the equation and writing several of the examples. Example:
P. 344: "For example, in a 20-kHz audio bandwidth there are 141 √Hz. The total noise created by a 1-kΩ resistor in that bandwidth is then 141 √Hz times 4.1 nV/√Hz, or 578 nVRMS."
It seems like a kludge, especially given that the equation for thermal noise is on page 181 (eqn. 10.2). This structure is used quite a few times throughout Ch. 16.
The text would flow better with something like this: ".... 578 nVRMS, or the equivalent noise of a 1-kΩ resistor at room temperature."
Tom
P. 20: The schematic could be greatly simplified by using an LM393 (dual comparator) instead of the quad LM339.
P. 339: "The outputs are then bridged with a resistor (R4) ..." Should be R3.
Tom
P. 339: "The outputs are then bridged with a resistor (R4) ..." Should be R3.
Tom
P. 330: "JFET input amplifiers are best used in these locations because of their absence of input current noise."
JFET (and MOSFET) input amps have input current noise. It's much lower than that of a BJT input amp, but it's not absent. I'd reword that as, "... their much lower input current noise".
Tom
JFET (and MOSFET) input amps have input current noise. It's much lower than that of a BJT input amp, but it's not absent. I'd reword that as, "... their much lower input current noise".
Tom
If you type “Bob Cordell” in the search, it will direct you to a page showing the book on sale.
Thanks for corrections - duly applied. Couple of points, as follows (with reference to Rick's nomenclature above);-
2) Figure 15.13 - the revised diagram has D1 & D2 missing - is that intentional?
7) Comments regarding FPGA's start on page 641.
Very informative and interesting book. Thank you both.
2) Figure 15.13 - the revised diagram has D1 & D2 missing - is that intentional?
7) Comments regarding FPGA's start on page 641.
Very informative and interesting book. Thank you both.
'Flops and latches are the architypal logic elements, most circuitry is built from them in synchronous logic, yes there's combinatorial logic with gates of course, but that's likely done with look-up tables or multiplexers anyway. I'd hate to have to design a state machine using just individual gates!
Fair enough. I'm not a digital guy so I won't pretend to be an expert on that subject. 🙂
Tom
It can be shown, that through logic reduction, any logic function can be reduced to a sum of products and therefore
FPGA's and the earlier PAL's (Programmable Array Logic) usually have sum of products arrays with optional inverters to get
the polarity correct. Given that some FPGA's are programmed from a small ROM on power up it is possible that their "logic"
is done in look up tables, but that's simply done so that SRAM can be used to implement the logic. Combinatorial logic
(unclocked) has been traditionally done with gates, in a hard programmed ASIC for example and for the last 20 years or so logic
synthesis tools have been used to take logic equations and turn them "efficiently" into what the target architecture requires.
VHDL, Verilog, etc. Registers are used to build synchronous (clocked) systems, or to synchronize async inputs.
Many synthesis tools, early on, even from the best companies were very poorly done.
FPGA's and the earlier PAL's (Programmable Array Logic) usually have sum of products arrays with optional inverters to get
the polarity correct. Given that some FPGA's are programmed from a small ROM on power up it is possible that their "logic"
is done in look up tables, but that's simply done so that SRAM can be used to implement the logic. Combinatorial logic
(unclocked) has been traditionally done with gates, in a hard programmed ASIC for example and for the last 20 years or so logic
synthesis tools have been used to take logic equations and turn them "efficiently" into what the target architecture requires.
VHDL, Verilog, etc. Registers are used to build synchronous (clocked) systems, or to synchronize async inputs.
Many synthesis tools, early on, even from the best companies were very poorly done.
I got something in the mail!
I ordered it July 21, it's a lot to pay for a book, but looking at the table of contents, the number of pages, and that it's a brand new book all sold me. The website said shipping was 6 to 12 business days, but I count 17. At least I have it now. Just looking through, it will augment many other books I have. I saw something about filters and I think of Active Filter Cookbook that I've had for probably 40 years and referenced many times.
I ordered it July 21, it's a lot to pay for a book, but looking at the table of contents, the number of pages, and that it's a brand new book all sold me. The website said shipping was 6 to 12 business days, but I count 17. At least I have it now. Just looking through, it will augment many other books I have. I saw something about filters and I think of Active Filter Cookbook that I've had for probably 40 years and referenced many times.
Still haven't seen any print reviews of this book, although privately I'm aware of some mild criticism. Is it fair game to discuss the pros and cons?
Further to Tom's comment about thermal noise I've had a page on my website with a handy table that is quite useful when estimating resistor noise:
https://www.njohnson.co.uk/index.php?menu=2&submenu=2&subsubmenu=8
Neil
https://www.njohnson.co.uk/index.php?menu=2&submenu=2&subsubmenu=8
Neil
Thank you very much for buying my book. It is a significant price for a book, but it is somewhat in line with the cost of my amplifier book when it came out. I have also noted significant price increases in technical books over the last few years, as I had to buy a number of them for researching this book. I hope you enjoy the book.
Cheers,
Bob
I think it is definitely fair game to discuss both errata and pros and cons of the book here. That helps me do a better job and also gives me an opportunity to explain any possible misunderstandings. Thank you in advance for any input you have.
Cheers,
Bob
Hi Tom,
This is a very good point. JFET noise is very low, but because there is a semiconductor junction at the gate, there is inevitably some current noise. This can come into play with sources that have very high source impedance. Under those conditions, it is useful to compare the noise that the current noise causes across the source impedance with the thermal noise of the source impedance. It takes a pretty high source impedance to have as much thermal noise as the noise cause across it by JFET gate current, especially at room temperature. Note that the gate current noise is shot noise. These kinds of considerations come into play in small-diaphragm condenser microphones, for example. There is discussion of that in the microphone r in my book.
Small-signal MOSFETs typically have vanishingly low input current noise (and more 1/f noise).
Cheers,
Bob
Hi Neil,
I just got a chance to look at your web page on thermal noise. That is a really handy, succinct set of information on noise. Thanks! Knowing the effective noise bandwidth (ENBW) of a frequency response function can be important to any analysis. Some might find Table 10.1 on page 179 of my book to be useful. There the ENBW of low pass filters of order 1 through 6 is shown for both Butterworth and Bessel filter functions. It is of course the case that as the filter order increases, the ratio of ENBW to filter cutoff frequency approaches unity. It is also very handy to know that the ENBW of the A-weighting noise function is 13.5 kHz.
Cheers,
Bob