Measuring the noise floor

The HP4195A doesn't have the best noise floor. It is a general purpose instrument. It works well when the measurement environment is set up properly. You need to understand the noise characteristics of your probes. You also need to be aware of ground loops when dealing with low level signals. The differential probe helps with ground loop problems.

More things to watchout for:

1. The bandwidth setting of the instrument will change your noise floor. Narrow bandwidth setting will give you a lower noise floor. If your making comparisons between the 4195 and 3588, be sure your comparing the measurements using the same bandwidths. Better yet, the

4195 does spectral density measurements.

1. The 4195 has a low noise setting for the input amplifier. Be sure that it is enabled.

2. When measuring small amplitude signals, be sure the input attenuator is set to zero dB.

1. If you need to measure signals below -90 to -100 dBm on the 4195, use a low-noise preamp. This gives the option of having an input impedance other than 50 Ohms.

2. Make sure your active probe is working.

1. If your probe has gain, be sure to adjust your readings. The 4195 has a math function that allows you to add gain offsets.

Mark

To get to the low noise mode, press the Receive Attenuation button. You will have a selection of IF RNG Normal IF RNG Low Distortion (only in spectrum mode) IF RNG Hi Sensitivity

Low distortion and high sensitivity mode will give you a better noise floor. I'm a bit baffled by your bandwidth setting limitations. The

4195 allows you to use bandwidth settings down to 3 Hz. Might take a few days to make the measurement at 3 Hz depending on your setup.

What is your basic setup? Freq range, mode, circuit impedance, ... If your doing noise floor measurements, you want to use the spectral density mode (dBuV/Hz, dBm/Hz, uV/rtHz) which reports noise in a 1 Hz bandwidth.

Mark

Thanks. I didn't know about the low noise setting, I'll have to check the manual again. The resolution bandwith with the 3588 was 17k. The

4195A only lets me set 10k and 30k in that range. Though I would expect to see a descrepancy between the data from the 3588 and 4195A, the lack of change between the on and off conditions worries me about the condition of the 4195A. The preamp is probably what I need. I'll look into renting one of those when the 4195A come back from cal.

Thank you

Paul C

Yes I am using the 'noise' mode reading in dBm/Hz with a 1Hz bandwidth and the 4195A does let me set the bandwidth down to 3Hz. I had set the

3588 to 17KHz and tried to get the same settings on the 4195A. The 4195A only allows certain quantum jumps in RBW ; 300Hz 1KHz 3KHZ 10KHz 30KHz. It lets you key in other settings but it jumps to some level. I did try it down at 300Hz and it took a long time to make it across the screen but it didn't help the readings. I missed the sensitivity settings on the Receive Attenuation buttons. I'l have to try that. I did play with attenuation settings but couldn't get the machine to recognize the on from the off condition of the circuit. I know it's probably operator error, that makes it all the more frustrating.

Paul C.