Electronics > Projects, Designs, and Technical Stuff
Any interests in a high impedance amplifier for spectrum analyzer?
(1/1)
ycui7:
I am working this project and would like to get some feedback on interests about it.

I am make a high impedance amplifier that can enable you to use a 10X oscilloscope probe on spectrum analyzer.

The bandwidth is about 400 MHz, considering the target use case is power supply noise diagnostic. There is no plan to go higher bandwidth.
A 400 MHz+ 10X 10 MOhm probe will be required.
The amplifier comes with two boards. One for the amplifier itself, the other one is the supporting powering and control circuitry.

The support board has USB as power input, it will pull 100 mA from the power source, so you can take power directly from you spectrum analyzer, if it has an USB port.
It also come with a 300 mAh Lithium ion battery and lower noise isolated DC DC converter to get +/-5V the amplifier required.
There is a SP3T switch that you can set to put the amplifier into Power OFF, Power ON, and Battery Only mode. The Battery Only mode will power off all digital circuitry on the amplifier assembly to give you the best noise floor.
The amplifier support simultaneous battery charging and operation, so that you can continue to work, if you forget to charge it.
The +/-5V power is generated by an isolated DC DC converter. The USB power has no direct ground connection to the amplifier.

A RF signal relay is provided such that both AC and DC coupling can be supported.
A trimmer is also present to correct the DC offset, if you want to use it non-spectrum analyzer cases.

The first prototype shows noise floor at <500uVpp at 100 MHz bandwidth (the 100 MHz bandwidth limit come from Tektronix 7A13, which is the most quiet test equipment I have at the moment, a full 400 MHz bandwidth noise floor will be provided eventually)

The dimension of the amplifier is about 2.5" x 2" x 1.2", excluding connectors. Case has not been made yet, but of course there will be a metal case to cover the whole amplifier for shielding.
David Hess:

--- Quote from: ycui7 on May 07, 2020, 04:09:34 pm ---The first prototype shows noise floor at <500uVpp at 100 MHz bandwidth (the 100 MHz bandwidth limit come from Tektronix 7A13, which is the most quiet test equipment I have at the moment, a full 400 MHz bandwidth noise floor will be provided eventually)
--- End quote ---

I would be really cagey about that noise measurement.  Because of its cascode and differential input configuration, the Tektronix 7A13 is one of the nosiest high impedance amplifiers.  I measured about 120 microvolts RMS on several of mine which is within their given specifications but implies about 600 microvolts peak-to-peak noise.  The 7A26 with twice the bandwidth has 1/4 the noise although that may not be apparent because of its more limited sensitivity.

I suspect the high noise is why the 7A13 has a 5 MHz instead of 20 MHz bandwidth limit.
ycui7:
Hi David,

How did you get your 7A26's noise measurement? I have multiple 7A26s, they all showed thicker trace than 7A13 at the highest sensitivity. For me, that mean my 7A26s has higher noise.

Could you elaborate how to get better noise floor from 7A26?

Thanks.
David Hess:

--- Quote from: ycui7 on May 08, 2020, 06:30:30 am ---How did you get your 7A26's noise measurement? I have multiple 7A26s, they all showed thicker trace than 7A13 at the highest sensitivity. For me, that mean my 7A26s has higher noise.
--- End quote ---

The 7A13's highest sensitivity is 1mV/div while it is only 5mV/div for the 7A26.  I made the measurements using the vertical output of the mainframe into another oscilloscope.


--- Quote ---Could you elaborate how to get better noise floor from 7A26?
--- End quote ---

Just use it?

How exactly did you make the noise measurement?  Did you use the tangential method shown below?  I assume not since you reported peak-to-peak noise.


mawyatt:
Nice looking circuit, PCB and concept.

Long ago when we needed to probe circuits with the SA and use a Hi Z scope probe, we would use the probe on the scope as usual and take the scope vertical output (back of most analog scopes) and feed it into the SA, this provided a scope and SA display.

Best,

Mike
Navigation
Message Index
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod