Hey, does anyone know what the actual layer spacing is on BatchPCB’s 4 layer boards, and how consistent that layer spacing is?
For a lot of designs that make use of 4 layer boards, knowing the spacing is critical… I’ve had experiences with some vendors being all over the map for layer spacing, including up to 30% variance between runs, which caused huge headaches.
Unfortunately being an RF Design Engineer (well, I was until I switched jobs last month), I still have the urge to design crazy things, but can’t afford to get more expensive PCBs like I used to for work…
I’m aware of the options for commercial PCB runs, however I’m inquiring from the hobby standpoint - Most hobbyists can not afford impedance controlled boards.
Further, non-impedance controlled boards can in fact be used for RF design, as long as the layer spacing is generally known – LOTS of commercial RF electronics that operate well aren’t produced on impedance controlled boards.
In a perfect world I’d have impedance controlled multi-layer PCBs with exotic substrates for my designs, but doing this as a hobby I’m stuck with using inexpensive 4 layer FR4 PCB, which returns me to the original question…
I’m in the same boat here. Since I’m designing a 4-layer RF PCB, I need some basic guidance on the dielectric thicknesses. An approximate value for each thickness would be adequate, I’m sure there’s a target that the PCB house uses. This is an absolute minimum information requirement in my opinion, as important as minimum hole size or trace spacing.
Of course, additional info on thickness variation would be a bonus.
For some cases (mainly where there are no high-Q elements) you can control the design better by only using the top and bottom layer. You can selectively do this on the places you want to better maintain the control (for RF purposes).
Leon - Sorry, forgot to mention that I’m also using BatchPCB. I’ll send off an email to Gold Phoenix, but I really wonder why BatchPCB doesn’t publish this info.
Languer - Thanks for the suggestion, although I think that the required 50 ohm trace width is very large, and not really compatible with modern RF designs (eg: at 2.4GHz) due to the huge discontinuity when entering device pads.