Date: Fri, 22 Jul 2011 11:17:01 -0400 From: Patrick Allison Just a few details on the TDA that's being sent to UH: First, schematic is here: http://ara.physics.wisc.edu/cgi-bin/DocDB/ShowDocument?docid=256 (as a note: you may notice that TRIG0 becomes TRIG1+/TRIG- and TRIG1 becomes TRIG0+/TRIG0-. This was done for routing, and it's just a renaming of the pins on the connector. I'll clean up the schematic soon. Also note that the order of the DAC outputs is wonderfully funky - effectively, in terms of the TRIG+/TRIG- numbering, it's channel 3, 1, 2, 0 - again, routing) The TDA has several resistors that determine the threshold scan curve that may need to be tweaked depending on how it looks with an actual tunnel diode and representative noise: To simplify things a little, just consider that on, say, sheet 3, the "tunnel diode threshold" is basically (VBIAS - THRESH0 - 200 mV). VBIAS is currently around 1.3V, and is common for all 4 channels. The 200 mV is more like +/-50 mV - it's the LVDS input threshold. The output of the DACs is nominally 0-2.5V, but if you look on sheet 6, R3/R4/R5/R6 form a resistive divider between the DAC output and something close to 1.25V (the diffamp shifts so the outputs are symmetric about VOCM = 1.25V). I'm going to try to do a simulation showing what this does, but basically you can think that with a large (compared to the sum of RA3/RA4/RA6/50 ohms, or roughly ~900 ohms) resistor at R3, it compresses THRESH0 from a 0-2.5V range to more like 0.625-1.875. From a "threshold curve" perspective it should stretch the curve around the rise. But for first testing I'm actually now thinking that we should just make R3/R4/R5/R6 = 0 ohms just to get the full threshold curve and then if we want to tweak it later we can. The other thing that alters the threshold curve is VBIAS (obviously) which basically moves the curve left/right. Right now VBIAS is probably too low - it should probably be more like 1.5V. Once we do figure out what value of (VBIAS-THRESH0) we nominally think we will probably run at (it doesn't have to be exact, just a rough guess) then it would be good to tweak VBIAS or VOCM such that (VBIAS+THRESH0)/2 is VOCM. This just reduces the static power consumption of the diffamp and balances the two amplifier halves. So to summarize I think the right thing to do is maybe: 1: threshold scan with the current setup 2: change R3/R4/R5/R6 to 0 ohms, VBIAS to 1.5V. I'm working on software for performing a threshold scan right now. Hopefully should be done soon. Patrick