wfotto

Just to clarify, the Hi Level output is floating. There's a DC-DC board (28V to 12V converter) and transformer isolated from the Amb board. The "V" (virtual) ground is floating, not tied to any other ground except at the DAQ. Ergo, we need differential DAQ inputs.

Maybe a drawing would help. We started with no pull down resistors. We were getting some erratic data with our High Level transducers, so changed to the 2416 recommendation of 100K pull downs. Unfortunately, this configuration, with our Low Level transducer, put the two 100Ks in parallel with the bridge 5K strain gages. The transducer output at full pressure was about 600uV lower than without the pull downs (nominal 0-30mV). I'm not sure if this was due to just the parallel combination, or if the DAQ DGND isn't totally isolated from the P GND? When we tested with the Low Levels with the 5M pull downs, they didn't affect the Low Level output at full pressure, basically the same as no pull downs. The concern is whether the 5Ms will adversely affect the High Level transducers? Note, the High Level transducers, on the Amp Board side, is a load cell with an instrumentation amplifier. DAQ Hi_Lo Level.pdf

JRys, For the load cell (full bridge): our system can record up to 10 load cells at a time. The load cell is excited with 10VDC. Isn't the USB-2416 GND isolated, therefore no current return path to the power supply? Yes, the 5M works for a load cell configuration. But is that OK with an active device, one with an op amp output? Will that high impedance cause issues, like bias currents in the USB-2416 instrumentation amp? Thanks, Bill

For the USB-2416, and I assume most MCC DAQ devices, 100K pull down resistors are recommended to avoid floating inputs. This works when the input is from an active device, such as a conditioned instrumentation amplifier output. This does not work well with pure resistance devices such as mV output bridge circuits. Even though the 100K's are tied to the USB-2416 floating ground terminal, they are still tied across the bridge outputs. This adversely affects the output. As a workaround, we found 5M pull downs essentially remove the bridge interaction. My question is, how do 5M pull downs affect the principle of floating inputs for the active device? Is there a maximum resistance, after which it doesn't work?

Can I get C of C's through this forum. If not, now do I get one for a product I bought? Bill, Taber.