2019-11-05 04:03 AM
Hi Enrico
Can you please suggest us the correct values for data rate, deviation and channel filter for the sp-l2 in a narrowband application?
How narrow can we go?
Also in your schematic you use a trimmer for calibration .Why? Isn’t the tcxo precise enough?
Pratically ,how do you do the calibration ? do we need special instruments?
Another question .
In your sch the balun differs from others. As we can see below
Why?
Marco
ciao
2019-11-05 08:10 AM
Hi,
I kindly asked Marco to post his questions here.
I will answer question 5 today and tomorrow the other questions.
QUESTION 5
>> Pratically ,how do you do the calibration ? do we need special instruments?
Program the S2-LP transceiver in order to transmit a pure sinusoidal wave at 868 MHz, called carrier frequency.
If the board has an RF connector, connect the board to an RF frequency counter or to a spectrum snalyzer.
If the board does not have an RF connector than connect an 868 MHz antenna to the spectrum analyzer and the RF energy will be picked-up through the air.
There are other ways but those 2 are the most common.
Please note that an RF frequency counter is more precise than a spectrum analyzer.
Turn on the S2-LP and look at the instrument.
If the instrument measures, let's say, 868000120 Hz than your carrier frequency is off by 120 Hz.
Now, using a screw driver, gently impress a rotation to the trimmer in order to move the carrier to 868000000 Hz.
I'm using a low cost 10k trimmer Bourns TC33X-2-103E, to trim the frequency of the VC-TCXO.
The correct verb is "to pull" the frequency. The pull range of common VC-TCXO's is 10 KHz.
That means that you can tune the frequency up to 10 KHz.
Bourns TC33X-2-103E trimmer allows just 1 turn.
COMMENT 1
If your system is battery powered at, let's say, 3 Volt, than the TCXO must be powered at 1.8 Volt through an LDO at 1.8 Volt.
You will also need a P-MOSFET, driven by the microcontroller, to completely turn off the LDO, the TCXO and the trimmer in order to save energy.
COMMENT 2
Why do we need to power the TCXO at 1.8 Volt?
The voltage of the battery, due to its internal resistance, will drop during RF transmission bursts.
If the TCXO were powered directly from the 3 Volt battery, its output frequency will not stay stable during the transmission bursts.
That means that the S2-LP will receive an unstable reference frequency and will transmit on a different channel, let's say at 890 MHz.
The RF scenario will now be the following:
The transmitter will transmit on a channel around the 890 MHz while the receiver is listeining at 868 MHz.
It's not a nice scenario, is it?
A brand new CR2032 battery will drop easily by 0.5 Volt, if S2-LP transmits 10 dBm.
An aged CR2032 battery will drop easily by 1 Volt, if S2-LP transmits 10 dBm.
Ciao,
Enrico Migliore
2019-11-06 04:03 AM
hi enrico
Does the problem exist also with a normal power supply? i guess not
what about the narrowband settings ?
also , can you suggest a frequency counter .do we need a very precise one ?
are the SURECOM SF-103 or the SURECOM SF-104 ok?
marco
2019-11-06 07:52 AM
Dear Marco,
>> Does the problem exist also with a normal power supply?
If the on-board power supply is derived from a wired connection and if you have a capacitor on the VDD of the board of at least 10 μF than the problem doesn't exist. The VDD will not drop and the output frequency of the TCXO will stay stable at 52 MHz.
>> what about the narrowband settings ?
In the next days. I'm pretty busy.
>> also , can you suggest a frequency counter .do we need a very precise one ?
>> are the SURECOM SF-103 or the SURECOM SF-104 ok?
I'll ask around and let you know.
I have an HP 8566B spectrum analyzer and I'm using that.
Ciao,
Enrico