i make the area using the solid tool, and build it from rectangles - each with a fictional Net name, then when i do ground pour i delete that net which leaves the guard ring, probably not the correct way but works for me

@CANBED, That's all fine until you rebuild the copper area. What you are asking about is how to separate one net with a copper area into two or more physically different structures which are still electrivcally the same net. Note that the answer to this essentially the same as for how to separate the same electrical net into two or more differently named nets such as analogue ground and digital ground or specialised routing for decoupling capacitor tracks. This post below points to a related discussion and to a project that shows a number of ways to do the sort of thing you are asking about: [https://easyeda.com/forum/topic/How-to-create-a-four-wire-sensing-or-Kelvin-Connection-on-a-PCB-8034986947764c1fb1a872de020160e8](https://easyeda.com/forum/topic/How-to-create-a-four-wire-sensing-or-Kelvin-Connection-on-a-PCB-8034986947764c1fb1a872de020160e8)

Note that isolating an area of a circuit with an island surrounded by a moat is very offten completely inneffectual because unless there is a drawbridge joining the island to the rest of the copper area and along which tracks are routed (such as the clock output of an oscillator) then the increased inductance of the loop created in the ground return path by a track just crossing over the moat (i.e. not over the drawbridge) does more damage than the noise or potential crosstalk that the moat is intended to isolate the island from.

@andyfierman So you think it is better to add just GND vias around the oscillator? How would do it you?

I was describing a fairly general case. What's your particular application? Is it for a 2 layer PCB? Schematic? Device datasheets?

And are you talking about an island with a moat and drawbridge structure or are you talking about a guard ring (which is a ring of copper around a circuit usually to intercept surface leakage currents)?

@andyfierman LSE Crystal fail to start on my 4 layer PCB (26.5 x 12 x 1.2mm). I am using [S76G](http://www.acsip.com.tw/index.php?action=products-detail&fid1=19&fid2=&fid3=&id=41&lang=3) chip which includes STM32L073x with the TFX-03 crystal (Data sheet [here](http://www.river-ele.co.jp/catalog/E_TFX03&03C&03L_Sep04_2018.pdf)), Oscillator design guide for STM32 [here](https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf), see pdf page 31) from the RIVER company with two 6pF CL capacitors. The Drive level of the oscillator is 0.5 μW. Although I can uprgade the firmware, supplied by the AcSip Company, in the S76G chip via [STLINK-V3MINI](https://www.st.com/en/development-tools/stlink-v3mini.html) (SWD) and STM32CubeProgrammer I am unable to communicate with the chip over UART (115,200 bit/s) to be able to send a few commands to configure the chip for my task. I created test program in STM32CubeIDE, just to enable LSE on MCO PA8 pin to measure it by oscilloscope and found out the LSE is not running (as I assumed). When I touch crystal by a finger I get random oscillations I catched on attached oscilloscope screenshots. I noticed I can change the default settings of the STM32L073x from RCC\_LSEDRIVE\_LOW to HIGH \(2\.7\) to get GAINmargin about 4\,2 \(**2.7**/0.6447 µA/V). See pdf page 86 in STM32 [Datasheet](https://www.st.com/resource/en/datasheet/stm32l072cz.pdf) (LSEDRV). I will try it tomorrow. When I touch crystal by a finger I get random oscillations I catched on attached oscilloscope screenshot.  Maybe the PCB design (4 layers) is responsible for not working LSE. I didn't expect any problems with 32.768 kHz. U3 is TCVCXO oscillator 26 MHz running well (Vpp = 1V). 

You have said that your crystal is a TFX-03 but you have not said what load capacitance your part specifes.  And where then does your chosen TFX-03 crystal fit in the table on p31 of: [https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf](https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf)  "...with two 6pF CL capacitors." I don't understand what this means. Have you checked you calculations for the load capacitance in section 3.3 on p12/13 of: [https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf](https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf) ? "Although I can uprgade the firmware, supplied by the AcSip Company, in the S76G chip via [STLINK-V3MINI](https://www.st.com/en/development-tools/stlink-v3mini.html) (SWD) and STM32CubeProgrammer I am unable to communicate with the chip over UART (115,200 bit/s) to be able to send a few commands to configure the chip for my task." I think that is a question for ST support.

@Caster What I do is declare an area around the crystal guard a +3.3V fill.  Then the copper fill won't attach itself to the crystal guard which is an essential part of any crystal oscillator.

@andyfierman I mentioned it above: "_with two 6pF CL capacitors_" it corresponds with the value in the table above TFX-03 e.g. 5pF. CL means Load Capacitance for the crystal. See S76G Applicatin Note: 

@MikeDB Please can you post here picture of your PCB LSE 32.768 kHz crystal oscillator design for inspiration?

@Caster, If you are using the TX03 type F3A-32768-050 then calculations for the load capacitance in section 3.3 on p12/13 of: [https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf](https://www.st.com/resource/en/application_note/cd00221665-oscillator-design-guide-for-stm8afals-stm32-mcus-and-mpus-stmicroelectronics.pdf): | | | | | | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | 3.3 Load capacitanceThe load capacitance is the terminal capacitance of the circuit connected to the crystal oscillator.This value is determined by the external capacitors CL1 and CL2 and the stray capacitance of the printed circuit boardand connections (Cs). The CL value is specified by the crystal manufacturer. For the frequency to be accurate, theoscillator circuit has to show the same load capacitance to the crystal as the one the crystal was adjusted for.Frequency stability mainly requires that the load capacitance be constant.The external capacitors CL1 and CL2 are used to tune the desired value of CL to reach the value specified by thecrystal manufacturer.The following equation gives the expression of CL:CL = CL1*CL2/(CL1+CL2) + Cs | | | | | | | | | | | | | | | | | | | | Let’s assume that CL1=CL2=CLx | | | | | | | | | | | | | | | | | | | | CL=CLx^2/(2*CLx)+Cs | | | | | | | | | | | | | | | | | | | | CL=CLx/2+Cs | | | | | | | | | | | | | | | | | | | | CLx=2*(CL-Cs) | | | | | | | | | | | | | | | | | | | | | | For a crystal with a quoted CL of | 5 | **5** | 5 | 5 | pF | | | | | and a stray capacitance of | 1 | **2** | 3 | 4 | pF | | | | | CLx = | 8 | **6** | 4 | 2 | pF | | | | | | | | | | | | give the required value for the C22 and C23 in:   Note however that a value of 6pF is only valid for a total stray capacitance of 2pF. Since your layout may differ from that of the board from which 6pF values in the above image is derived there is no guarantee that 6pF will be correct for your PCB.

Sorry, the forum has trashed the formatting of my copy and paste. Here it is an image: 

@andyfierman It was my first design of the 4 layers PCB. Unfortunately the AcSip didn't state anywhere in the docummention sample PCB design for the LSE 32.768 kHz oscillator. I was not aware about any problems which may arise from the bad PCB design. I just followed the instructions to use the right oscillator type and capacitors. Tomorrow I should get MEMS oscillator SiT1533AI-H4-AA3-32.768E and I will try to replace the TFX-03 crystall with it and let you know the results in a few days.

"Tomorrow I should get MEMS oscillator SiT1533AI-H4-AA3-32.768E" Yes, I  was going to suggest that as an option. Several things I have had with 32kHz xtals in them have had problems of non-starting or, in the case of the Humax PVR9200, the clock stopping after a couple of years as the PCB gathered dirt from the airflow over it from the cooling fan! On a recent project I decided to take the more expensive but easier option and used the Epson SG-3030LC oscillator module (RS p/n 667-6284).

Unfortunately removing crystal oscillator and soldering MEMS oscillator on assembled PCB made short circuit on PCB and I burn out two resistors. I have to make a new PCB.