Again, I think you mistake solder PASTE with solder MASK. Even though they may seem similar, they are, in fact, completely different (although it could be up to 90% inverted image.
Solder PASTE is used to solder the SMT components to the board using - solder paste.
The solder MASK is used to state where the 'copper' surface need to open, that is, not covered by the (green or your case blue) solder mask.
As I said in the first line, they may be 90% identical as a SMD pad both needs an opening (solder MASK) and PASTE to solder the component.
Solder MASK photo:
https://external-content.duckduckgo....jpg&f=1&nofb=1Solder PASTE photo:
https://external-content.duckduckgo....jpg&f=1&nofb=1Soldering process video:
https://youtu.be/xqv9HF_2GlYAnother item is the surface finish. Two of the most used finishes are HASL (Hot Air Solder Leveling) and ENIG (Electroless Nickel Immersion Gold). HASL (or lead free HASL) is the cheapest. Looking at your photo, it looks like HASL. And yes, it may look as if the board has been solder - which is EXACTLY the case. A surface finish is required as bare copper is terrible with oxygen and the finishes are not. If you have not specified a specific finish, the fabricator will select one for you.
A video about HASL:
https://youtu.be/IjsqukppWfwProduction step for PCB production:
https://www.eurocircuits.com/Making-...-step-by-step/The application note you mention only talks about keeping the solder mask from the guard ring "These guard rings have no solder mask..." (page 12). It state nothing about a surface finish or soldering.
So it should be OK.
Additional comments:
- New layout seems a lot smaller and in my expectation better.
- Do keep in mind that there is a minimum trace with.
- You can specify a solder mask region or something else on the solder mask layer like @robertferanec mentions. However, you can also set the solder mask for the traces and arc like in the picture below. With the method Robert does, it is either a more manual activity, or a large area no longer has solder mask as you uncover the entire area.
The result is here:
And you can play with the solder mask setting too, both positive and negative and like Robert said "You can simply verify that in 3D model." (well "view" not "model", but good enough)
- You mentioned "and one more thing my earlier design has failed. But when I place this guard ring. . .My design has worked." The guard ring should not be the difference between working and failing, but more reliable and accurate (less noise or drift).
- From what I know and it is also stated as bit the application note "The guard ring is biased at the same voltage as the sensitive node; it needs to be driven by a low impedance source." (page 12). In all the examples show the ring it attached to either inverting or non-inverting node AT THE OPAMP.
- Looking at the layout, and expecting schematic I think that a different layout will make a much greater improvement than the guard ring would. The output trace is long and goes around to both ends if the input circuit.