Easy way to test gain steps

[Brad Jeffries]

Hello, 

Just got a HL2 from the latest MakerFab build.  Unit is gross functional, realized I was overloading the RX with my OCF antenna.   I have an IC7300 with the PTX pan adapter output that is after the IC7300 filters.   I connected the HL2 to this and got much better results.   

During all this the behavior of the noise floor looked strange as I incremented or decremented the gain/attenuation in SDR Console.   Is there a table of expected noise floor vs various gain settings with the input into a dummy load?  

Others using band select filters ?   I have the N2ADR filter board and believe I have it setup correctly.  

73

Brad 

[Brad Jeffries]

I got my radio berry out and working.  If I terminate the RX with a dummy load and sweep the gain with the AGC off the noise floor moves up linearly.  If I do the same on the HL2 it moves up linearly and then jumps.  This repeats periodically its like one of the ranges in the input LNA/ attenuator is defective.   Can someone do the same on their HL2 and verify the noise floor moves linearly? 

I took everything apart and hand soldered the ground pad under the ADI chip. 

73 

Brad 

[Brad Jeffries]

https://youtu.be/5nADvY91pzg

Got a comparison of the HL2 vs the Radioberry.  Notice the distortion spurs in the HL2 and the few settings that have huge jumps in gain (either up or down).  The Radioberry is smooth and not harmonics.   I believe my front end LNA/VGA is junk. 

[Steve Haynal]

Hi Brad,

The Radioberry and HL2 share the same gateware code base. The HL2 uses a fast parallel interface to quickly set the LNA in one or two clock ticks. The Radioberry uses the slower spi interface to set the LNA. Can you please visually inspect the AD9866 and FPGA pins and traces associated with this fast path for potential shorts? 

You can find the HL2 schematic here:

http://www.hermeslite.com/

The pins to check are pins 1-12 on the AD9866. All traces for these pins are on top of the PCB. You can follow them from the AD9866 to the FPGA. If you don't see any shorts which you are comfortable fixing, then please arrange to send the boards to me for repair.

73,

Steve

kf7o

[Brad Jeffries]

Hi Steve,

Interesting about the fast port vs the SPI.  Traces look good and I ohmed them all out to check for shorts to gnd.  I believe the input of my AD9866 is defective.  The linearity is bad even if the gain is down.   I sent an email to makerfabs I will see what they say.  I am leaning toward just swapping the AD9866.

73

Brad

[Steve Haynal]

Hi Brad,

It could be the LNA on your AD9866 is defective. Makerfabs is sending me a batch of bad boards. Usually I can repair them to swap with defective boards in the field. Also, I am gathering a collection of boards which need ICs swapped out to take to a local assembly house. It may be easier to send it to me if you are in the US than back to Makerfabs.

73,

Steve

kf7o