XPost: sci.electronics.design   
   From: nospam@no-address.com   
      
   On 18.11.2015 19:17, Phil Hobbs wrote:   
   > I'm working on an interesting project for a biotech customer, and I   
   > could use some wisdom from the assembled multitude.   
   >   
   > They're building a scanning fluorescence microscope that scans   
   > several lines at once with a single photomultiplier, using a clever   
   > scheme to make each line come out at a different RF frequency.   
   >   
   > The frequencies are inherently evenly spaced, so third-order IM   
   > products cause image artifacts.  So they asked me to come up with a   
   > way to reduce the IMD.   
   >   
   > I've been doing some measurements, and have found the usual cubic   
   > dependence of the IM3 products up above about 50 uA of anode   
   > current. That's generally a combination of voltage sag in the dynode   
   > bias string, plus some space-charge effects.  It can be reduced by   
   > changing the dynode voltage distribution so that the last few stages   
   >  have more bias.   
   >   
   > Interestingly, though, below ~50 uA there's a broad range of anode   
   > currents (20 dB or so) where the IM3 products go up linearly with   
   > the signal, so that the SFDR is a nearly constant 60 dB, plus or   
   > minus a few.  This is without changing ranges on the spectrum   
   > analyzer or anything like that.   
   >   
   > Ahah, you say, the IM3 is in the illumination source, which runs at   
   > a constant drive level.  So I thought, but it ain't so.   
   >   
   > My illumination source is a pair of blue LEDs, each driven from a   
   > separate amp with separate power supplies.  To get the spurs big   
   > enough to see, I'm using two synthesizers locked together--one is   
   > just a frequency doubler hung off my 10 MHz rubidium reference, and   
   > the other is a PTS1000 locked to the same reference, running at   
   > 20.001 MHz.  The analyzer is an HP 8566B, locked to the same   
   > reference.  That way I can use a 10-Hz resolution bandwidth to see   
   > the spurs.   
   >   
   > Using a photodiode, I've verified that the spurs in the light   
   > sources are below -80 dBc.   
   >   
   > There's mention in the literature of a space-charge effect at the   
   > photocathode, which would vary less with gain than in the last   
   > stages. If that's it, the -60 dB would be set at the very beginning,   
   >  so it should depend mostly on the actual light intensity at the   
   > photocathode.   
   >   
   > However, the linear behaviour occurs regardless of whether I'm using   
   >  a fixed light level and adjusting the anode current with the bias   
   > voltage pot or leaving the voltage alone and using a variable optical   
   > attenuator.   
   >   
   > It also doesn't depend on the attenuation of the signal in the RX   
   > chain--adding a 10 dB pad doesn't change the SFDR in that anode   
   > current range.   
   >   
   > This is, in other words, weird as hell.   
   >   
   > Any ideas?   
   >   
   > Thanks   
   >   
   > Phil Hobbs   
      
   Hi Phil   
      
   Have you checked for voltage sag at the last (highest current) dynodes?   
   Tried to stiffen them up? Maybe even run them from separate stabilized   
   power supplies connected in series instead of the usual resistor chain?   
      
   If the frequency is not too high, could you "compensate" by driving the   
   last one or two dynodes actively, proportionally to the anode current   
   (sort of like the "ultralinear" transformer tap in tube amplifiers is   
   used to drive G2, but in this case controlled and driven by fast solid   
   state amplifiers instead)? It's an old linearization trick, but if you   
   can get fast enough drivers, maybe worth a try...   
      
   Regards   
   Dimitrij   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)