0

I am trying to test a LvPECL signal into a LVDS receiver (NI 6583). Unfortunately, I currently only have the single-ended segment/connector. Looking at the spec, the single ended receiver should still be able to take one end of the differential signal.

I've scoped it out, and the voltage ranges from 3.8V and 2.16V (according to the datasheet, the receiver can receive Voltages from -0.5 to 4.6V). I've also scoped out the difference and the voltage swing is from +/- 1.6V (so this probably wouldnt work).

I've tried connecting one end of the LvPECL signal to the receiver as well as the + end of one differential end to the + pin of the receiver and the other + end of the other differential end to the GND of the receiver.

I'm not sure how useful this will be but heres a general schematic of my setups Setup 1 Setup 2

Any idea why my receiver isnt picking up the signal?

bchang32
  • 93
  • 6
  • Try connecting the differential ground to the single-ended input ground, and one of the differential output wires to the single ended input wire. – Math Keeps Me Busy Sep 23 '21 at 11:50
  • I'm having trouble understanding what you have. For starters LVPECL, if properly implemented, is only guaranteed to swing 800 mV. So where is the voltage range of 3.8 V to 2.16 V you mentioned coming from? – SteveSh Sep 23 '21 at 13:05
  • On you receiving end, do you have a differential receiver of some sorts, or just a single ended receiver? And what is the part number? – SteveSh Sep 23 '21 at 13:07
  • Finally, schematics, even hand drawn ones, are always helpful. – SteveSh Sep 23 '21 at 13:08
  • @MathKeepsMeBusy I tried that and it was unsuccessful (The LvPECL signal is BNC and I have a BNC to terminal block connector that I'm using to connect). – bchang32 Sep 23 '21 at 13:20
  • @SteveSh I am using the NI 6583 FAM (So the TI SN74AVC2T245 single ended receiver). My LvPECL signal is from a converter from Pulse Research Lab (so its 2 SMA outputs). The voltage is what I measured across an oscilloscope from one end. When I scoped the Math of the difference, the swing was about 800 mV. I'll add a schematic but i think it would be useless since its COTS items and just a box into another box? – bchang32 Sep 23 '21 at 13:23
  • Your '245 receivers needs an input voltage greater than 2.0 V, or less than 0.8 V to recognize it as a valid logic level. The input threshold for that part is around 1.6 V. So you need to swing approximately +/- 800 mV about that point to guarantee that the receiver can distinguish low from a high on the input. – SteveSh Sep 23 '21 at 16:46
  • @SteveSh so when I connected the positive end of Q to + and positive end of Q' to Gnd (schematic 2), the differential should work then? The voltage goes from 1.6V to -1.6V (although the negative voltage is a bit out the range). I've tried that too to no success – bchang32 Sep 23 '21 at 19:16
  • No. If I'm understanding your setup correctly, the voltage has to go above 2 V at the input to the receiver, the '245. The -1.6V at the input is going to stress the ESD protection diodes. – SteveSh Sep 23 '21 at 19:19

0 Answers0