¿ªÔÆÌåÓý

Hello,

On Mon, May 8, 2023 at 06:02 AM, Gerard wrote:

Only one wire line (green), normally 2?
I think it's LM2 still present ?

I think the self LM2 is not shunted on our modules.?

Necessary or not ultimately?

LM2 should be shunted, otherwise you still have some kind of low pass filter at the input
?
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

Hello,

Here where is LM2 on the bottom
. To shunt or not
?
See? Post #846

cdt

Re,?
other thing
Do you think we can protect the input like on this vid¨¦o?

it's a Simple schematic


cdt

Hello,

A little question:
When I look at this documentation and the modification to be done on the SI4432 modules, I wonder if a short circuit is missing?
It's wroten "tinysa can work above 150 mhz...so



Only one wire line (green), normally 2?
I think it's LM2 still present ?

Hello,

¿ªÔÆÌåÓý

With JLPCB 4 layers were and are the same price as two.? 5 off 100mm x 100mm come to ?9.45 including postage (not sure if this includes tax).
I think it might be worthwhile having JLPCB source and fit the attenuators, resistors and caps, SAW filters and Si4432 but as I don't plan on making another one just yet I have not fully investigated this.? I like soldering anyway (except maybe those attenuators!).

Dave

dAVE?

Interesting find!? I had not seen the devkitC before, and nice that it is available with/without external antenna connection.

Unfortunately it highlights one of the problems with the ESP32 in that there are many (too many) different module types with different pinouts.? The pinout of this module is not the same as the board I have used, I am not sure if you adapted your PCB design to suit a different ESP32 module or not but you do need to check.

If you have the metal case already (or maybe a tin can from your soup, biscuits or whatever!) give the existing module a try in it - it is surprising how these high frequencies can find their way in and out of supposedly shielded environments.? I have had mobile phone calls with a person working inside a stainless steel tank on a site where there was very poor mobile phone coverage - seemed to work better inside the tank than outside!? Of course keeping the WiFi RF away from the SI4432 is also a good idea.? The code has to do a few tricks at certain frequencies in the scan as the 9th harmonic of the LO can block the WiFi.? Took me ages (many weeks) to work out what was happening.

Re the web page and high/low ranges - the drop down at the top right of the web page selects the mode.? The last settings for each mode are remembered (after a time delay to avoid too many writes to the Flash) and loaded back when the mode is changed.? The same controls are used on the web page, the ESP32 code knows the mode and adapts accordingly.

Dave

hello,

After a little research, I see that if I want to use an antenna outside the box, I need an ESP with a small plug.?
something like that (4€) and an antenna (0.5€)
it's will work?

Dave,

Dave,

I had to remind myself how this works before responding!

The mode allows you to see the bandpass response of the SAW filters.

Connect the reference signal to the SA input J14, and set a frequency of 30Mhz (for example) for the reference signal.? You can set the reference from the menus in SAW mode.? You can use an external generator as well if you wish.? From the Menu select Sweep Sig and enter in the frequency of the signal being input to the SA (eg 30000000, or 30 and M).? You can adjust the sweep start and stop points as you wish each side of the SAW filter midpoint.

The markers can be set to show the peak and -3dB points each side, and then the difference between the 3dB points, as shown in the photo.

I find it best to not use auto gain if the signal is quite strong.? When in Auto the SI4432 adjusts its own internal preamps and the delays associated with this can cause some spurious effects.??

In Sweep Lo mode, Display/Traces menu you can turn the gain trace on or off.? When shown you get a green trace as well as the yellow one.? If in Auto the trace is not flat (usually you will see a dip in gain around the peaks) then it is best to fix the gain through the sweep at a level corresponding to the low of the dip.
Gain of 37 works well for my example for a -30dBm signal with no attenuation.
I did consider doing something smart here but never got round to it, preferring to set things myself.
Higher gain is good in that it enables you to see smaller signals, more detail, away from the peaks, but if too high then you get spurious signals at the peaks.? After a while you get to know how to set the gain best to suit what you are trying to measure.
If you have an encoder connected you can change the gain value easily.

Dave,

Hello,

Results of tests Version _11? and version _15
see pictures
Version _11? ?---> Only Sweep low
Version _15? ----> Sweep low and High !!!!

Conclusion: I think my sweep high must works because it's was use with sweep low of version _11 (I speak of the copoments on PCB)
My real sweep low with version _15, will be work, because i changed a Pe4302. To be tested...
It's most clear now. Lol

cdt

Dave,

Just at this time i was looking version 11 and 15
I find version _11 is not for 4*SI4432
There is no variable RF_Switch in the program _11? --->>present in version _15
I have look documentation on tiny, you have in the mode menu LOW frequency, and High frequency, + other choice
In version _11, only low_sweep, not High sweep or other
I will look version 15 this afternoon, and made new tests.? May be some thing to do to activate the "Hight sweep" or other??

cdt

When you load in the code for _15 do you also upload the data files associated with that version to the ESP32?? There were some changes to the web pages and the way the data is transferred via websockets, but it was some time ago and I forget the exact details.? The data files you upload to the ESP32 include the web page source, the libraries for the chart drawing, some fonts.

Regardless of the web page you should get a sweep on the LCD.

If you inject a 50Mhz signal into J14 then you should get 50MHz through the attenuator, through the LPF and on to the mixer pin 2.
During the sweep the LO SI4432 will output RF at high frequency.? The frequency will vary through the sweep, starting at the IF frequency (434MHz) plus the frequency corresponding to the start of the sweep (0Mhz) and then ramping up through the sweep, so if the end of the sweep is 100MHz the LO will ramp up from 434Mhz through to 534MHz.? When the LO frequency is the IF frequency plus 50MHz (484Mhz) then the frequencies out from the mixer given the 50MHz input signal will mix and produce 484+50MHz(534MHz) and 484-50MHz(434MHz).? The B3555 SAW filters remove the 534MHz component and the 434Mhz signal is then measured by the RX Si4432.

In other words during the sweep Pin 5 of the RF switch will be connected to pin 3 so the RF being generated by the LO SI4432 is fed to the mixer.

When used as a signal generator both LO and RX Si4432 are producing RF.? The RX produces RF at the IF frequency (434MHz) that is mixed with the RF from the LO Si4432.? If the LOSi4432 is set to say 454MHz then there will be 454+434 and 454-434Mhz output from pin 2 of the mixer going into the low pass filter.? The filter will remove the 888Mhz signal leaving the 20Mhz to go through the attenuator to J14.

If you set your device to signal generator mode and set a frequency of 20Mhz and turn the output on then you should get 20Mhz out on J14.? If you get this then all hardware on the SA side is working.

The TG works in exactly the same way.? The swap of wires for the TG was a mistake on my board.? You need to set this to match your PCB design.? This will not affect the operation of the SA side at all so lets worry about that later.

Maybe we should try and set up a teams or zoom meeting

Dave

Sorry for point 1, it's the view WEB
cdt