开云体育

Dave,?

Mine behaves the same, maybe I have a dodgy modules as well.

73

Jim - G3ZQC

Can you show a scan of your BPF?
The "AGC overloading", as you mention it, could be caused by a too wide BPF. This possibly can be solved by moving the IF frequency to the side of the filter instead of somewhere in the middle
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

Thanks for this John.

One thing - the green line on the trace shows the gain from the LNA/PGA preamp in the SI4432.? By default the AGC algorithm adjusts the gain to suit the signal strength.

There is an error in the SI4432.h file - the #define for PGAGAIN should be 0x0F not 0x3F.? Without this correction the green gain trace will not work.

When it gets to the top of the ToDo list a selection will be added to the menus to allow this trace to be turned on/off.

I find when zoomed in I have to turn the AGC off and fix the gain to avoid spurious signals probably due to the ADC overloading before the AGC reacts, see attached capture from the web page that is under development.? I will be interesting to see if others experience the same or if I have a dodgy module.

For the moment the setting for the Pre-Amp gains is found in the config menu - again it will find a new home at some point, and again for the moment the menu does not include all the gain options available.? Using the attenuator so the AGC doesn't react is of course an alternative.

73
Dave

开云体育

When measuring harmonics ALWAYS repeat the measurement with a different level of attenuation to ensure the harmonics are not internally generated.
If the harmonics move down in level compared to the fundamental with increased attenuation the harmonics are internally generated. This happens when you apply a "strong" signal and is due to the? IIP3 level of the tinySA somewhere between +7dBm and +15dBm (depends on the mixer, mixer drive level, attenuator pads around the mixer and attenuation in the LPF)
If the harmonics stay at the same level compared to the fundamental when you increase attenuation the harmonics are actually present in the input signal.
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

I did a test this morning with Johns V1.4. as released in other thread.

This little gadget is pretty neat. I used my Rigole to compare a signal out of my HP8640b sig gen.?? As you can see, both indicate 50dB down for 2nd harmonic.

The signal is at -20dBm. (ignore the scale on the Rigole, theres a 20dB pad in front of it.)

Glenn
vk3pe

tinySA_V1.4 is now available! Thanks to David (M0WID), Glenn (VK3PE) and Jim (G3ZQC) we fixed the power output problem in the transmitter Si4432. The requested power was totally not working!

You can set the desired output level using "D" command from the serial monitor as David noted in a previous post, however, that setting would only remain in effect for the current program run.

Now the "D" command (which I added back to the serial output menu) can be used to set the power and the program saves it in the flash memory so it will still be set the next time you run the program.

The drive can be set to a specific level by entering the command "Dn" where "n" is the desired level:

??? 0 => +1dBm????? 4 => +11dBm
??? 1 => +2dBm????? 5 => +14dBm
??? 2 => +5dBm????? 6 => +17dBm
??? 3 => +8dBm????? 7 => +20dBm

The recommended drive level for the ADE-25MH mixer is +13dBm and for the ADE-1 mixer, it is +7dBm.

Just entering "D" (or "d") will give you the current setting.

I'll work on making some of the other serial input commands work correctly in the days ahead.

Enjoy!

--
John P.
WA2FZW

Hi Jim.? As you can see having big trouble focusing on my work today.? Maybe I need to redirect these emails or go to the daily summary format.

I seem to understand less each day!? Several years ago I thought I had at least a basic understanding of CSS but having had to look at some for this project it is clear I no longer have a clue, and there are new things to get my head round such as Grid and Flexbox, and getting stuff to work on multiple devices and browsers is still a nightmare.

Hope you have your version working.

The main reason for the bits being set is I just copied what Erik did in his code and it worked.

Having looked at it, according to the data sheet bit 4 is set after the SI4432 is reset, so not knowing what it does it seems sensible to leave it as is.
Bit 3 should not matter as we are using an external switch not direct tie, I think it could be either on or off.? Try it and let us know if it makes any difference!
We are not turning the LO TX on/off, and not bothered about the LO RX values.? The TX does actually stop briefly on each frequency change so it is possible there is some effect if the mosfets switch in this brief period, though I would not expect them to change.

73

Dave M0WID

Folks - Please ignore the discussion related to the low TX output power!

Working with David this morning we (he) found what the real problem was and corrected it. Since I don't have a working Si4432 (out of seven of them), I can't test the changes we made. So when Jim, Glenn and David get a chance to test the fix, I will publish the new code.

The updated version also allows you to set the output power with the "D" command and it now remembers what you set it to when you restart the program.

Stay tuned!!!
--
John P.
WA2FZW

开云体育

Dave,

I don't understand (nothing new there).? According to the datasheet, Reg 6D (aka REG_TXPWR) bit 4 is reserved, and bit 3 is the lna_switch, set in the "direct tie" configuration.? Why are you setting them on?

Jim - G3ZQC

Dave, a bit more than that.
See registers
? SI4432_Write_Byte ( REG_GPIO0, 0x12 );??? // GPIO automatic antenna switching
? SI4432_Write_Byte ( REG_GPIO1, 0x15 );

The antenna switch in the module is not in Tx position.

glenn
vk3pe

OK.? I see it.? Will post a fix after work,? but in the meantime at line 2164 or thereabouts the line
? ?? SI4432_Write_Byte(REG_TXPWR, 0x1C + (drive - 2 )/2);

I think should be
? ?? SI4432_Write_Byte(REG_TXPWR, 0x18 | (drive - 2 )/2);

To me it makes more sense for the drive variable to be the correct three bits and the UI/serial command handler sort out the conversion from dBm so this is perhaps better:
? ?? SI4432_Write_Byte(REG_TXPWR, 0x18 | drive);
and then a D7 would set full power, and until a mod is done change the inital value for drive to 7.

For a test you could just use this:
SI4432_Write_Byte(REG_TXPWR, 0x1F);

Thanks again for pointing this out.

Back to work now after the welcome distraction!

Dave M0WID

Dave,
I found a bug in the software which doesn't allow the Tx LO to develop full output level. Changing the variable 'drive' does change levels but the max. level is still way way down.

I passed the info onto John and no doubt he will contact you.

Attached is the tinySA in action with a 50MHz signal at -30dBm from an HP8640b sig gen.?? As you can see, the signal is showing a level of about -26dBm.?? I have provision on my PCB to add a pad into the Rx module to correct this error. (It's not yet fitted.)
Linearity seems to be very good also, stepping the sig gen down in 10dB steps, the tinySA follows nicely.??? Freq response looks pretty good also.
Current drain from 12v supply is about 174mA.

Pretty remarkable for what on the surface is a simple SA although there's a lot going on in those modules !

Glenn
vk3pe

Hi Marty.

I am using version 1.04 of the ESP32 arduino environment,? but was using 1.03 until a few days ago.? No pressing need to upgrade from that version, but 1.04 produces slightly faster code.? My arduino IDE is version 1.8.10, I keep on avoiding the update requests but don't expect a problem.? Upgrades can be done from the Arduino boards manager.

If you don't have the ESP32 environment then follow the instructions here:

I think I needed to close the IDE and re-open before it worked properly.

I have the ESP32 Dev Module as the selected board in Arduino.

You will probably also need the tool to upload data files to the SPIFFS (SPI Flat File System) on the ESP32.? If not installed as part of the above then you can follow the instructions here:

I think the Data upload tool now comes with the ESP32 development environment, not 100% certain.
?
Using the arduino library manager you will need to install:
TFT_eSPI - I am using 2.1.4
WebSockets by Markus Saddler version 2.1.4

ESPAsyncWebServer can be downloaded from?
Download as a ZIP to somewhere sensible then install using the Sketch/Include library/Add .zip library feature in the Arduino IDE.

ESPAsysncWebServer depends on?
Download as a ZIP and install as above

In the ZIP for this version I forgot to include the subfolder with the files needed to get the TFT_ESPI library to work, so they are attached to this message.? ?You need to put the User_Setup_select.h in the TFT_eSPI library folder, and add the M0WID_Setup_ILI9341_TinySA.h in the User_Setups subfolder.

I think thats it, but its so long since I installed everything I cannot be 100% certain.? If you still get errors send me a screen shot of the errors and I can take a look.

Good luck

73

Dave M0WID

On Thu, Apr 2, 2020 at 05:39 AM, Dragan Milivojevic wrote:

Please no loitering of . Use github ..

Loitering?? :-|)? ?Yes, does seem like the better resource as long as the mission doesn't change with newer Microsoft ownership. GoogleCode as with other Google services has changed for the worse.? SourceFourge though seems to be going smoothly with their mission for the longer term and I hope the same with Github.??

Hi Glenn.? Good spot.

There is a variable named drive that is initialised at 6, and this determines the LO drive.? The LO drive level is updated at the start of every sweep to whatever value the drive variable contains.

The drive varible can be adjusted through the serial port using the D command - so in Arduino bring up the serial port monitor and enter Dvalue, where value is what you need it to be. ?
To obtain full strength, which is probably what you want with the ADE25 mixer you have (from looking at photo, with i think a pad?) then D16 should do the job.

There is no facility to set the drive via the menu, so I will add this, as well as being able to set from the web page (along with IF frequency, crystal calibration values etc), and the value will be saved to flash so that value is used next time it boots up.? The ability to set via the serial port will remain.

I think the SI4432 object that John is creating sets the LO power based on a #define to specify which mixer is used, but some users may choose to have a lower level mixer and a higher attenuation pad and still require full power, so I think setting to a low value initially, then letting the user configure to what they want is best.? Other views welcome.

Keep finding those missing features and bugs!

73

Dave M0WID

Here's the RF side of things. Note the? crude Atten module bypass for now. (coax would be nicer)

An initial test showed the LO power out to the mixer to be way down.?? Something in the sketch causing this, as a test program gave healthy levels.

Glenn
vk3pe

On Wed, Apr 1, 2020 at 04:25 PM, Marty Wittrock wrote:

How is your project (and libraries) set up to make the compile clean such that I can load my ESP32??

Thinking the version for each is excellent to note somewhere, i.e. Arduino IDE version, Board version (unless maybe like linked via File--> Preferences-->Additional Boards Manager URL's) and Libraries version with maybe source(s) included also so know where found.?

Might not be bad to save links on the Internet Archive (? also for future reference where you can also just add the URL wanting to be saved to the address bar link https://web.archive.org/save/%url%?like in a batch file if iterating through a site map list of each link with SET BROWSER=%browser% and START %BROWSER% "%url%" with a wait_time of some sort or pause?and taskkill so not a bunch of tabs or browsers open at the same time.?

I personally don't like to update any software unless absolutely required other than with the Arduino IDE where I use the Boards Manager under Preferences to update each time I use... and I guess Spybot S&D and Antivirus definitions.??

Just opened my WROOM ESP32 and uChip that came in the mail today after sitting in a bag with IPA for a few.?

Thanks for all your work on this and sharing.?