<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"/>
</head><body style=""><div>Hi Richie, Neil,</div>
<div> </div>
<div>IMO it is incorrect to generally state that the diode ring modulator "operates as a switching modulator".</div>
<div> </div>
<div>More accurate would be "can be operated as a switching modulator" or (depending on the application) "is intended as a switching modulator".</div>
<div> </div>
<div>I did a Spice simulation just now.</div>
<div> </div>
<div>With sine amplitudes of 1V of both input signals, the output is similar to what you'd expect from a four quadrant multiplier: at any point in time the output signal is approximately proportional to *both* input voltages (the ring modulator is definitely not as distortion free as a double-balanced mixer though).</div>
<div> </div>
<div>If you choose a higher amplitude of the "center taps driving" signal of, say, 10V, then it does start to act like a chopping modulator. Indeed, the inputs are not interchangeable: the output signal remains proportional to the other input signal.</div>
<div> </div>
<div>If anyone wants to see for theirself, below are the contents of an .asc file that you can load into LTSPICE IV. You may want to plot V(plusout)-V(minout), V(n001) and V(n004).</div>
<div> </div>
<div>By the way, I now look forward to build a circuit around a double balanced modulator, equipped with switches to overdrive neither, one, or both inputs, thus selecting the three "ring modulator" flavours "linear", "chopping", and "xor" :)</div>
<div> </div>
<div>Best regards</div>
<div>Simon</div>
<div> </div>
<div>Version 4<br>SHEET 1 880 680<br>WIRE -96 96 -288 96<br>WIRE -48 96 -96 96<br>WIRE 176 96 0 96<br>WIRE 192 96 176 96<br>WIRE 272 96 256 96<br>WIRE 416 96 272 96<br>WIRE 480 96 416 96<br>WIRE -48 112 -48 96<br>WIRE 176 112 176 96<br>WIRE 272 112 272 96<br>WIRE -48 192 -48 160<br>WIRE 0 192 0 176<br>WIRE 0 192 -48 192<br>WIRE 176 192 176 176<br>WIRE 192 192 176 192<br>WIRE 272 192 272 176<br>WIRE 272 192 256 192<br>WIRE 416 192 416 176<br>WIRE 416 192 320 192<br>WIRE -288 208 -288 176<br>WIRE -48 208 -48 192<br>WIRE 416 208 416 192<br>WIRE 0 224 0 192<br>WIRE -96 240 -96 96<br>WIRE -48 240 -96 240<br>WIRE 176 288 176 192<br>WIRE 416 288 176 288<br>WIRE 480 288 416 288<br>WIRE -48 304 -48 288<br>WIRE 272 304 272 192<br>WIRE 272 304 0 304<br>WIRE 320 336 320 192<br>FLAG 480 96 PlusOut<br>IOPIN 480 96 Out<br>FLAG 480 288 MinOut<br>IOPIN 480 288 Out<br>FLAG -288 208 0<br>FLAG 320 416 0<br>FLAG -48 304 0<br>FLAG -48 208 0<br>SYMBOL diode 256 112 R0<br>SYMATTR InstName D1<br>SYMATTR Value 1N914<br>SYMBOL diode 256 176 R90<br>WINDOW 0 0 32 VBottom 2<br>WINDOW 3 32 32 VTop 2<br>SYMATTR InstName D2<br>SYMATTR Value 1N914<br>SYMBOL diode 192 176 R180<br>WINDOW 0 24 64 Left 2<br>WINDOW 3 24 0 Left 2<br>SYMATTR InstName D3<br>SYMATTR Value 1N914<br>SYMBOL diode 192 112 R270<br>WINDOW 0 32 32 VTop 2<br>WINDOW 3 0 32 VBottom 2<br>SYMATTR InstName D4<br>SYMATTR Value 1N914<br>SYMBOL res 400 80 R0<br>SYMATTR InstName R1<br>SYMATTR Value 10k<br>SYMBOL res 400 192 R0<br>SYMATTR InstName R2<br>SYMATTR Value 10k<br>SYMBOL voltage -288 80 R0<br>WINDOW 123 0 0 Left 2<br>WINDOW 39 0 0 Left 2<br>SYMATTR InstName V1<br>SYMATTR Value SINE(0 1 100)<br>SYMBOL voltage 320 320 R0<br>WINDOW 123 0 0 Left 2<br>WINDOW 39 0 0 Left 2<br>SYMATTR InstName V2<br>SYMATTR Value SINE(0 1 1000)<br>SYMBOL e 0 208 R0<br>SYMATTR InstName E1<br>SYMATTR Value 0.5<br>SYMBOL e 0 80 R0<br>SYMATTR InstName E2<br>SYMATTR Value 0.5<br>TEXT -224 440 Left 2 !.tran 0 0.01 0<br><br><br></div>
<div> </div>
<div> </div>
<div> </div>
<div><br>> Op 17 januari 2016 om 23:14 schreef Richie Burnett <rburnett@richieburnett.co.uk>:<br>> <br>> <br>> Thanks Neil. They are all pretty clear on the matter! A diode ring <br>> modulator (double-balanced mixer) operates as a switching modulator, not a <br>> product modulator like I had always thought. So it doesn't really matter <br>> what waveshape you feed into one of the inputs of a real diode ring <br>> modulator, as long as it has sufficient amplitude to commutate the diodes!<br>> <br>> -Richie,<br>> <br>> -----Original Message----- <br>> From: Neil Johnson<br>> Sent: Sunday, January 17, 2016 8:04 PM<br>> To: Richie Burnett<br>> Cc: Simon Brouwer ; synthdiy diy<br>> Subject: Re: [sdiy] passive ring modulator transformers<br>> <br>> Hi Richie,<br>> <br>> Here is some more light reading for you:<br>> <br>> The original Cowan patent:<br>> http://patft.uspto.gov/netacgi/nph-Parser?patentnumber=2025158<br>> <br>> The wikipedia page:<br>> https://en.wikipedia.org/wiki/Ring_modulator<br>> <br>> My own take on things:<br>> http://www.milton.arachsys.com/nj71/index.php?menu=2&submenu=2&subsubmenu=12<br>> <br>> Cheers,<br>> Neil<br>> <br>> <br>> On 17 January 2016 at 19:50, Richie Burnett<br>> <rburnett@richieburnett.co.uk> wrote:<br>> > Fair enough :-) who am I to argue with minicircuits!!! I'm still not sure <br>> > whether that app note is actually saying that you *can* use a diode ring <br>> > mixer as a switching modulator if you have square waves conveniently <br>> > available instead of sines, or if they are saying that it *always* <br>> > operates in a switching mode regardless of the drive waveform. I guess <br>> > whether this mattersor not depends on what you're using it for.<br>> ><br>> > I've always thought of ring modulators as proportional devices. Albeit <br>> > ones that aren't particularly linear and distort a lot. And for the right <br>> > range of amplitudes in, you can get a decent approximation to a linear <br>> > multiplication function, as shown in the video you linked.<br>> ><br>> > It's interesting that they say pulse inputs actually reduce distortion <br>> > though! I know switching modulators are often used by the software defined <br>> > radio crowd so they obviously work fine that way.<br>> ><br>> > Cheers,<br>> ><br>> > -Richie,<br>> ><br>> > Sent from my Xperia SP on O2<br>> ><br>> > ---- Neil Johnson wrote ----<br>> ><br>> >>Hi Ritchie,<br>> >><br>> >>> My point was that his two cases of a gain of +1 and -1 are just the <br>> >>> saturated gains.<br>> >><br>> >>This Minicircuits document: https://www.minicircuits.com/app/AN00-011.pdf<br>> >>would seem to disagree with you:<br>> >><br>> >>Q. I am a digital designer, dealing with pulses rather than sine<br>> >>waves. Is it necessary to<br>> >>furnish only sine waves to a double-balanced mixer?<br>> >>A.<br>> >>No. A double-balanced mixer operates as a switching device; pulses are<br>> >>fine, and may<br>> >>even reduce distortion.<br>> >><br>> >>> The diode mixer or ring modulator is a proportional device, and in fact <br>> >>> Minicircuits even recommend a sinusoidal local oscillator signal at <br>> >>> +7dBm level for optimum operation. With harder LO drive the behaviour is <br>> >>> less linear and tends more towards the saturated switching behaviour, <br>> >>> with the addition of a whole load more distortion products in the <br>> >>> output.<br>> >><br>> >>The +7dBm (level 7 device) spec is to do with conversion gain or loss,<br>> >>not about driving the diodes. Other mixers are designed to be run at<br>> >>higher levels, upto +27dBm for some devices (they tend to be rather<br>> >>more esoteric!).<br>> >><br>> >>> Yes, that dafx paper is the one I was thinking about. Thanks for the <br>> >>> links! I only skimmed over it but it looked good.<br>> >><br>> >>Hmmm...<br>> >><br>> >>Cheers,<br>> >>Neil<br>> >>--<br>> >>http://www.njohnson.co.uk<br>> <br>> <br>> <br>> -- <br>> --<br>> http://www.njohnson.co.uk<br>> <br>> <br>> -----<br>> No virus found in this message.<br>> Checked by AVG - www.avg.com<br>> Version: 2016.0.7294 / Virus Database: 4489/11425 - Release Date: 01/17/16<br>> </div></body></html>