Re: Antuino

Bruce Hall


Thanks so much for your help and thoughts.  I assume you are seeing the same thing on your end?   I want to make sure this isn't just because of a single unit or bad measurement data.

Is there a way to see where in the receive chain the non-linearity starts?   In your talk you mentioned that the AD8307 had a very linear response.  Agree that 70dB would be good.


On Wed, May 29, 2019 at 3:31 AM Ashhar Farhan <farhanbox@...> wrote:
I guess, we need a better post mix amplifier? probably increasing the current there will help? I am not sure where this non-linearity is coming from. I suppose at least 70 db of linear range will be good. this is about 50 db.

- f

On Wed, May 29, 2019 at 2:08 AM Bruce Hall <bhall66@...> wrote:

I tried a the mod you suggested and bypassed the IF stage.  In my case I removed the coupling caps on either side of the stage (C27 and C11), and soldered a new 100nF coupling cap between them (see attached).

I re-ran the return loss calibration and did a plot with the step-attenuator in place, also attached.   The graph is shifted down about 5dB.   I don't see good linearity until I add about 20dB attenuation.


On Tue, May 28, 2019 at 2:06 AM Ashhar Farhan <farhanbox@...> wrote:
Bruce, it is possible that there is too much gain. You can remove R8 (2.2K) resistor that provides feedback to the Q1 transistor to switch off that stage and use a 0.1uf capacitor between C11 and C7 to bypass that stage. you will have to rerun the return loss recalibration routine before you can accurate readings. I will try these out at my end too.

- f 

On Tue, May 28, 2019 at 2:20 AM Bruce Hall <bhall66@...> wrote:

Does anyone see gain compression at the higher dBm levels?  Or does everyone else see a linear response on their units?  I have been adding 20-30 dB of attenuation to get into the linear part of the curve, but this significantly reduces the useful range of the device.

I am wondering if there is too much amplification in the IF section.   What would be a good method to troubleshoot?


On Wed, May 22, 2019 at 8:48 PM Bruce Hall via Groups.Io <> wrote:
I took a few more measurements, adding in 3dB and 6dB steps at higher attenuation values.  It looks like the additional 3dB and 6dB steps are attenuating RF appropriately.  But the shape of the graph remains non-linear.   See attached.   

I am not sure what to try at this point; any suggestions?


On Wed, May 22, 2019 at 3:35 AM Ashhar Farhan <farhanbox@...> wrote:
This is interesting. There could be two possible explanations:
1. There is gain compression. This is possible due to the two stage IF amplifiers.
2. The step attenuator is off-calibration.

One way to know is to measure readings of smaller attenuations at different attenuations. For instance how does the readings change between -50db and -53db as opposed to -10 and -13 db? My tests show quite linear relationship between actual attenation and db readings.

- f

On Wed 22 May, 2019, 6:06 AM Bruce Hall, <bhall66@...> wrote:
I have been playing around with the Anduino.   I inserted a step-attenuator between the RF input and RF output ports.  With no in-line attenuation, the measured RF output is -16 dBm.   As I added attenuation, the measured RF power predictably decreased.   The receiver bottoms out at roughly -70 dBm, and does not record levels any lower - which makes sense.  What I don't understand is the non-linear relationship at low levels of attenuation.  For instance, adding 3dB, 6dB, or even 12 dB of attenuation did not significantly change the Antduino's displayed power measurement.  Independently, I crudely measured the RF power using my oscilloscope, and the power output changes as expected.   I am curious what results others get with their units.    Why am I getting a non-linear relationship between attenuation and RF power?

On Tue, May 21, 2019 at 7:52 PM Ashhar Farhan <farhanbox@...> wrote:
I just landed back home an hour ago. I will start putting the documentation for the antuino together tomorrow.

- f

On Wed 22 May, 2019, 4:12 AM Glenn, <glennp@...> wrote:
Rob i had issue with fonts also.   I still can't find the correct glcd LIB either, did you?.

Keep in mind I am a novice at this but do have Jacks book.

What i think it should be is to have all the Fonts in a directory called "fonts". Fonts.h is just a list of the fonts used in the Sketch.
in the same directory as the sketch.

then change #include <fonts/allFonts.h> to "fonts/allFonts.h"   so when you run the sketch in the IDE, it gets the fonts from the sub directory /fonts

I have another problem though in that i can't find a zipped version of the fonts required to download. 

I found the fonts here
But if I right click on each entry  "Save As" the downloaded files have extra characters in them making them unusable.


On Tue, May 21, 2019 at 11:53 PM, Rob Bleumer wrote:
Ho Ashhar
Same question from me and I got also problems with the line containing fonts/allFonts.
Perhaps part of the same.

Join to automatically receive all group messages.