I'm working with the PLL VCO HMC830.
I want to use a look-up table (LUT) for manual calibration, as decribe in "Frequency Hopping with Hittite PLLVCOs Application Note" (attached) (https : // www . hittite.com/content/documents/application_notes/Fast_Frequency_Hopping_Application_Note.pdf).
It should have equivalence between the desired frequency and capacitors switches configuration (Reg 10h Register Tune VCO).
I wonder if somebody already have those data and could provide it.
Thank you very much in advance.
The LUT is device specific. From page 4 in the app-note:
If you generate the LUT with one device and use it on another the switch capacitor selection might not be optimum over temperature and time. I believe you want to use the "Direct LUT" method as described in the app-note. This method associates each desired frequency with a switch capacitor selection.
You are right. I missed that part.
So, I have to include in microcontroller code a routine to generate the LUT and store in EEPROM.
Thank you for your help.
I have a doubt regarding the table with the values of the switch capacitors.
I set the HMC830 without auto-calibration.
Then I printed the registry values "Reg 10h VCO Tune Register" to various frequencies with a step of 10KHz in order to see that frequency changes occur Reg10 value. (see attached below)
As shown in the table, there are some "jumps" in the Reg10 values.
I have seen that there are 6 jumps and are always 32 to 5 on the previous value (see column "changes").
My questions are:
Will they always be 6 jumps?
Is the jump occurs in the same registry value?
Thank you very much for your help.
Autocal will pick the best VCO and switch capacitor setting for a given frequency. The actual location of the "jumps" vary depending on the VCO.
You can decode Reg 0x10 as follows:
Reg 0x10[7:6] VCO selection, 11b= lowest frequency)
Reg 0x10 don't care, may by high or low
Reg 0x10[4:0] switch capacitor setting (11111b=lowest frequency)
As you see the cause of the large jumps is the don't care bit 5. If you create a new 7 bit number by concatenating [7:6] with [4:0] you will get a steady decrease in this number as you perform the same measurement. But for any given frequency the number will vary from part to part.
Thank you very much for your help. It has been very useful.
Sorry, I get the values of the table with self-calibration function activated (do not explain well in my previous comment).