Mixer spurii - RSGB H/B 4th. Ed.

[Skywave]

Originally Posted by MichaelR

If you look at the table in the Collins material there is a footnote which explains numbers as shown in their table with a big black dot in front of them are in fact sum mixing; others are difference mixing.

The RSGB material may have used a typo error with the decimal point and no footnote.

Mike

The RSGB entry includes a footnote that clearly states that a dot "indicates sum mixing & that 'others' (viz. no dot - there are no other alternatives) indicate difference mixing".

Dave, G8HQP, clearly explaines above that you choose the 'dot' results is you are selecting the sum for your mixing product, (F1 + F2), and the 'no dot' results if you are selecting the difference for your mixing product, (F1~F2).

The 'dots' in the RSGB table are clearly not decimal points.

HTH

Al.

[Skywave]

Originally Posted by MichaelR

This material is assuming that F2 is close in Frequency to F1; this subject is extremely complex and they have tried to show mix levels which would or could cause a problem.

Mike

I'm genuinely puzzled by that remark, Michael - but then I do have a dual problem of sometimes failing to see the blatantly obvious or, alternatively, recognising deep subtleties.

Since the products of a non-linear so-called mixing process can be represented by a chain of signals whose frequencies are mF1 + or ~ nF2, where m and n are integers, why the restriction "F2 is close in freq. to F1"? Although the subject - in the broadest sense is 'complex' - there doesn't seem to be anything particularly complex about these two equations.

If I've misunderstood you, please accept my apologies; perhaps you'd care to expand on your thinking here?

Al. / Skywave.

[G8HQP Dave]

The RSGB material may have used a typo error with the decimal point and no footnote.

As far as I can see the RSGB have correctly copied the Collins chart, table and footnote. They have redone the type-setting (i.e. not just used the same printing plates), but it is an accurate copy. The error we have found is there in the original, unless we have completely misunderstood it. I am not being critical of Collins, as I said it is very difficult to get this sort of thing right (especially in the days before spreadsheets). It just serves to show that you should never take on trust any technical information you cannot understand and reproduce yourself, even if it comes from a normally reliable source.

[MichaelR]

Hi Dave/Al,

You are correct I did not see the RSGB scans posted earlier.

Mike

[MichaelR]

Originally Posted by Skywave

I'm genuinely puzzled by that remark, Michael - but then I do have a dual problem of sometimes failing to see the blatantly obvious or, alternatively, recognising deep subtleties.

Since the products of a non-linear so-called mixing process can be represented by a chain of signals whose frequencies are mF1 + or ~ nF2, where m and n are integers, why the restriction "F2 is close in freq. to F1"? Although the subject - in the broadest sense is 'complex' - there doesn't seem to be anything particularly complex about these two equations.

If I've misunderstood you, please accept my apologies; perhaps you'd care to expand on your thinking here?

Al. / Skywave.

I was more making the comment on the production of these tables. If you read the Collins article they indicate the levels of mixing products db relative to fundamentals. This is with the example of a 12Au7 valve operating in a defined non linear mode. Someone has calculated the complex ( for me ) 5th order coefficients .

Mike

[Skywave]

Originally Posted by AlanBeckett

Some twelve years ago I wrote a comprehensive manual . . . Recently, I had cause to re-read it - it's hopeless. Unless you understood what it was supposed to do you wouldn't stand a dog's chance of making any sense of it. But then I understood it when I wrote it!
Alan

Originally Posted by G8HQP Dave

It just serves to show that you should never take on trust any technical information you cannot understand and reproduce yourself, even if it comes from a normally reliable source.

Agreed, gentlemen! I've got the T-shirt - and the scars - to show on both fronts!

Two things are established by the above:
(a 1) when you do write technical documentation, (even if you know that you are only the one who is ever going to read it), read it several times through and ask yourself "does this make sense?" Then repeat this a few days later;
(a 2) assuming the necessary technical understanding, could I plonk this in front of somebody and feel confident that he / she could go through it without having to come to me and ask 'what does this bit mean?' ?"
(b) the merits of a technical education, be this formal or self-taught.

That aside, I (& Dave,G8HQP) referred earlier to a Spreadsheet version of the concept behind this graph / table combination. Not wising to forestall G8HQP's efforts on this, here's mine. It's an 'engineer's approach'; not particularly elegant, but - as far as I can tell - it works.

If you find otherwise - please let me know.

Al./ Skywave.

[G8HQP Dave]

I too learnt, the hard way, that something I write now has to be understandable by me in six months time when I have forgotten the context. When I used to write software I tried to get people not just to document what they had done, but also why they had done it that way, and what other apparently superior ways they had rejected and why. This was because I found we were often sitting in technical meetings scratching our heads and wondering "Why did we do it that way when we wrote that last year?"

My Firefox opened the zip file and found nothing in it.

I was going to attach my spreadsheet but I don't seem to have a copy of Zip on this machine - I need another round tuit!

[dsergeant]

Well I guess Firefox wouldn't know what to do with a .xls file anyway....

Yes, I like your spreadsheet and a short play with it gives one a good understanding about what the issue is about. Much easier than trying to understand that chart of course. Now just need a mixer application to put it to use...

One suggestion, the selection of f1+f2 or f1-f2 could do to being made a little easier. Since we need either one or the other then putting anything in the first box should automatically clear the second - or maybe some other means of setting this parameter.

(and I can never understand people who don't have WinZip or equivalent on their computter, it is probably the most used program I have on mine and couldn't live without it.).

Dave

[G8HQP Dave]

Here is my version of the spreadsheet - in three flavours: Works, Excel and OpenOffice.

[Skywave]

Originally Posted by dsergeant

Yes, I like your {Skywave} spreadsheet and a short play with it gives one a good understanding about what the issue is about. Much easier than trying to understand that chart of course.

Now just need a mixer application to put it to use...

One suggestion, the selection of f1+f2 or f1-f2 could do to being made a little easier. Since we need either one or the other then putting anything in the first box should automatically clear the second - or maybe some other means of setting this parameter.

Dave

Thank you the feedback, Dave.

"An application to put it to use . . . ?"
Ah-Ha! You've seen through my cover to where it all began! This all started when I got to the number-crunching bit of an AmRad HF bands ssb transmitter I am trying to design & build. Don't hold your breath though - it's a decidedly long-term, back-burner project.

Re the 'input validation'; choosing one F1 + / ~ F2 option that inhibits the other choice. I did tinker around with a few ideas, but couldn't find a way within Excel of achieving that result. (I'm open to suggestions.) Anyway, my 'get out' on this is that I did say that this was an "engineer's approach" (read: to be used intelligently)

Al. / Skywave.

[G8HQP Dave]

I have just obtained a copy of the Collins SSB book (2nd ed, 1959). I know it is on the internet for free, but I'm old enough to prefer reading from paper. Lots of useful information, although the earlier sections are clearly intended to convince their customers that they should move from AM to SSB (using Collins equipment of course!).