Voltage spikes using USB-6009

[johnsold]

Many multichannel data acquisition devices multiplex several inputs into one A/D converter.  Both the converter and the multiplexer may be switched capacitor devices.  Such devices require input current to charge the capacitors and experience large, fast changes in the instantaneous input impedance when switching.  You may see an improvement by putting a linear differential amplifier, such as an instrumentation amplifier, between your transducer and the DAQ device.

What is the output impedance of the anemometer?

Lynn

[cevert73]

Thanks for that.  The output impedance of the CTA is 50Ohms as specified in the manual.

I have actually tried experimenting with the gain settings built into the CTA device which possibly reduced the spikes for zero flow (or rather there influence) but once there was flow on the wire the anomalous spikes were as before.

[johnsold]

With a 50 ohm output impedance you should be fine.

I suspect that the problem is something else but that the symptom shows up as a result of the overall configuration or something subtle.  That is why I asked about what you saw with the scope.

How is everything grounded?  Do you have resistors to ground on both of the differential inputs to the 6009?  How is the motor for the wind tunnel controlled?

Lynn

[cevert73]

The motor for the wind tunnel has its own separate supply.  The blower is around 3m in diameter and has a lot of inertia so I can actually cut the supply and still have significant flow, upon which I'm still seeing the spikes.  I know the tunnel has previously been used successfully (albeit with different hardware).

I only have one resistor from -ve to ground.  I have read through this link ( http://zone.ni.com/devzone/cda/tut/p/id/3344 ) regarding field wiring and noise considerations for analogue signals and believe this to be adequate as I want to run DC-coupled, please correct me if I'm wrong here.

Thanks.

[johnsold]

Look at Figure 11 and related text from that document.

Lynn

[cevert73]

You reckon I should have two resistors tied to ground, where R1=R2?  Why would that be, because the source impedance is relatively high?

I can give it a try.

[johnsold]

No.  It has nothing to do with source resistance.  It is a path for the amplifier's bias current.

Lynn

[cevert73]

Could this explain what it is that I'm seeing with the voltage spikes?

Wouldn't that result in the mean voltage being brought down to 0V, as in AC coupling?

Apologies for all the questions!  I'll try this tomorrow with two 10k resistors.

[johnsold]

It is not clear that this is related to the spikes.  However the input stages of the multiplexer and converter may do strange things when not terminated properly.

With the 50 ohm output impedance of the anemometer 10000 ohm resistors are probably OK.  Any mismatch in the resistances can introduce errors and both the common mode and differential input impedances of the system are affected by these resistors.  Unless the specifications for your device say otherwise I would probably start with 1000000 ohm resistors.

Lynn

[Coq_Rouge]

This is the input circuit of the daq 6009 board. I know the daq 6009 is low cost. But still the input stage is very clumsy designed. As you can see resistor are already in place.

What is the noise in your system with only a battery connected to daq unit. It could be nice to know that.