No it is toner transfer. This process is well documented.

Ok, I'm waiting for your document.

 

Here is my latest - horrible photo I know but it's the best I could get my phone to do. It includes a SOIC8, SOT23, and 1206 capacitors and resistors.

 

Toner transfer I've done some good ones but where I live don't think they let me do that you can fart next door and smell it.
@AlbertHall looks great to me

 

RE:""I want to make my boards at home.""
The CNC for PCB - very inaccurate, very expensive, very slow (hours), very bad electrically (parasitic capacitance), bad looking.

I have to disagree with speed and accuracy. Even low cost CNC machines are not that slow and the accuracy, while not super high, is more than sufficient. I agree the cost is high if only for PCBs. If you figure $500 all in costs for a small mill, then you could get 25 or more different PCB designs made by one of the PCB outfits. Of course you also have to factor in material cost, bits, shop space, your time and so on. And, I don't know how to factor in the lack of silkscreen, plated through holes/vias and solder mask but that's the same for any home-made PCB. I used to make my own boards but once the price came down and delivery times got to be less than a week, I never looked back. Can't imagine going through the hassle of making a PCB.

 

Sometimes when comparing expense for something I can have in my hand "now" (or near now) vs. in a few days, I categorize the expense as the "right now shipping" cost. Having a milling machine on your desk is more expensive, but if you're in a real hurry it allows you to have a prototype in your hand very quickly.

 

Ok, I'm waiting for your document.

Type "toner Transfer" into the search box. No less than 1000 references to toner transfer. Have you really not seen any of those threads?

https://forum.allaboutcircuits.com/threads/project-how-to-make-pcbs.47681/

Note: Taunts removed by moderator.

 

Here is my latest - horrible photo I know but it's the best I could get my phone to do. It includes a SOIC8, SOT23, and 1206 capacitors and resistors.
View attachment 144412

Why are the traces so blurry?

 

Why are the traces so blurry?

The traces aren't blurry to my eyes so it's either the phone camera or the photographer's fault.

 

RE:""Mind you the RF frequency concerned is the extremely high 60kHz.""
Yeppssssk, I understand 100-2400 MHz when I am writing RF. And there I dont know better material in the markets like Rogers Inc. teflon-based PCB boards what are duroid-TC350™ at the price of $115 for 18x12 inch plate with loss factor tgφ=0.0015. Next is duroid-6035HTC (high temperature capable material) with tgφ=0.0013 and cost $240. Last is duroid-5880® material with tgφ=0.0004 and cost $325.

RE:""but if you're in a real hurry it allows you to have a prototype in your hand very quickly""
To encarve one pcb in 10x10 cm size with some 200 components on it, the time consume will be about such:
1)CNC method: file conversion for 3D-Max (Slicer etc) - 15 minutes, carving 45 minutes, change of broken drillbits (25 USD the piece) 5 minutes. Horizontal accuracy 50 micron, vertical 200 micron. Price for weared drillbits about 40 pieces x 25 $ per 1m2 sheet of pcb = 2000 $. Machinery price 300-600 $ (except computer). Electricity cost about 1...2 kWh - 0,5 $.
2)Positivus-20 method: file printout - 1 minute, aerosoling the emulsion 2 seconds, air-drying 5 minutes, owen drying 5 minutes, exposition under 400nm LED source 11 seconds, developing in NaOH solution less than 1 minute, etching in FeCl3 (warm) less than 10 minutes. Accuracy horizontal 1 micron, vertical - 0...0.5 depending on bath device type. Price of Positivus-20 aerosol 11 USD per 1 m2 and 1 kg/m2 of FeCl3 about 4 $, transparent film 1 m2 about 5...7$ depending on place where to buy, electricity for owen ~0,12 kWh or 0,02 $, colour-ink price 20 $ (using black in color mode the density is much better, never use the blacknwhite what gives poor contrast). Machinery costs except computer and printer) - bath 20 $, UV LED pack 30 $, owen 20$.
3)Paper sheet & iron method: printout to office paper 1 minute, warming-up the iron (must be PID controlled, inner regulator is far too inaccurate) 2 minutes, ironing 1 minute, developing in warm (!) water - less than 5 minutes, etching in FeCl3 (warm) less than 10 minutes. Horizontal accuracy (globally) 0,1...0,2 mm, (locally) 0,5 mm - (the fluff problem). Expenses - electricity 0,01$, paper sheet 0,01 $, ferric chloride 4 $, black ink powder 5 $. Using the chalk paper allows to increase the accuracy about twice, paper may be re-used from typography (offset ink is not disturbing).

 

RE:""Here is my latest - horrible photo I know""
For example Your most right side`s vertical path. Has surplus capacitance over those happening through the glass fibre, now happens as through as onto both sides. If length 10 cm thickness of copper 35 micron, gap to the gnd area 0,1 mm then C=0,1*35E-6*8,85E-12/0,1E-3*2 pieces=0,7 pF. Its just the resonant tank with resonance frequency about 10 cm*10 nH/cm=100 nH what results in f(rez)=600 MHz - its detrimental need to tick with a finger.

 

RE:""Here is my latest - horrible photo I know""
For example Your most right side`s vertical path. Has surplus capacitance over those happening through the glass fibre, now happens as through as onto both sides. If length 10 cm thickness of copper 35 micron, gap to the gnd area 0,1 mm then C=0,1*35E-6*8,85E-12/0,1E-3*2 pieces=0,7 pF. Its just the resonant tank with resonance frequency about 10 cm*10 nH/cm=100 nH what results in f(rez)=600 MHz - its detrimental need to tick with a finger.

Nobody is milling a gap of 0.1mm at home. Try your calculation again with a more typical gap of 0.4 to 0.6 mm. You'll see the capacitance is quite low. Even with your estimate, the calculated capacitance or even 10x your capacitance would not impact my designs.

 

RE:""Here is my latest - horrible photo I know""
For example Your most right side`s vertical path. Has surplus capacitance over those happening through the glass fibre, now happens as through as onto both sides. If length 10 cm thickness of copper 35 micron, gap to the gnd area 0,1 mm then C=0,1*35E-6*8,85E-12/0,1E-3*2 pieces=0,7 pF. Its just the resonant tank with resonance frequency about 10 cm*10 nH/cm=100 nH what results in f(rez)=600 MHz - its detrimental need to tick with a finger.

The highest intentional frequency on that board is 38kHz for IR remote control. The two tracks on the right hand side connect to a 2x AA battery holder. What happens at 600MHz I really don't care about. If I was making a down converter for satellite reception I wouldn't be designing or making the PCB - I wouldn't know where to start. Most of the stuff I do is PIC and/or analog and the boards produced by my mill are perfectly suitable for the purpose. I also use the mill for making things other than PCBs and these are parts which are custom designed for the purpose. I couldn't buy these parts and if I couldn't make them, I just wouldn't have them.

I am retired and this electronics stuff is my hobby to which I can now devote plenty of time. I spent quite a lot of time adding stuff to the mill (limit switches and probe connections) and writing the software (linked to above) which makes milling PCBs really easy. That time spent is not a cost it is just an enjoyable part of my hobby time.

 

Nobody is milling a gap of 0.1mm at home. Try your calculation again with a more typical gap of 0.4 to 0.6 mm. You'll see the capacitance is quite low. Even with your estimate, the calculated capacitance or even 10x your capacitance would not impact my designs.

And in any case the tracks that have that 'massive capacitance' are the connections to the battery

 

And in any case the tracks that have that 'massive capacitance' are the connections to the battery

Ooh, a way to save part count.
Soon we'll have a calculation for the capacitance with the ground plane on the other side of the board.

 

As a retired machinist, I have something to say. I see people saying things about accuracy and flatness when talking about these small machines. I think, not knowing for sure but from experience that a lot of it is caused by the owner/operator of the machine.

Accuracy of the cut width staying consistent across the whole of a part can be as simple as the machine never being adjusted, trued to it's self. Any mass produced machine doesn't come from the factory "ready to run". Be it one of these, a table saw or anything. It needs to be adjusted so all axis of movement are parallel or perpendicular to each other depending on the axis. And the spindle motor needs to be squared to all of them. If this isn't done you will never, never have the accuracy that a machine is capable of. Even very large very expensive industrial machines require this to be done. And as a operator of a machine it is the very first thing that is done before starting to machine a job. It's one of the first things taught in an apprenticeship or on the job as just an operator.

As to flatness, the tables of these small machines aren't flat or any I've looked at aren't. Many aren't even made of one piece of metal. They are usually made up of a few extrusions bolted or otherwise joined together to make it. PCB stock is way, way more consistent in thickness than the table of the machine is flat. The fastest fix for this would be to clamp a "spoil board" to the table and leave it there. The spoil board can be MDF board. After bolting it to the table it needs to have a "skin" cut taken off of it, using the largest cutter the machine spindle will hold, and by taking overlapping cuts over the surface that will then be used to hold the PCB. A second advantage of the spoil board is allowing the component holes to be drilled through the PCB and not hurting the actual machine table. That's why it's called a 'spoil board', it gets spoiled rather than the machines table.

 

Spoils board is an excellent idea but there is software to "flatten"your PCB.

https://www.autoleveller.co.uk/


To get really good results in isolation routing, the process is far too painful for me .

Many people mill a spoils board custom for the board being milled. That is even more pain. But imagine that it has some merit as I would imagine the spoils board can warp slightly over time.

 

You need a spoil board for making PCBs - drilled holes, mill the board outline, and dowels for locating double sided boards.
It doesn't need to be exceptionally flat as, with a probe, the software can compensate for un-levelness. I wrote my own but there is this which is available free: https://www.autoleveller.co.uk/

 

You need a spoil board for making PCBs - drilled holes, mill the board outline, and dowels for locating double sided boards.
It doesn't need to be exceptionally flat as, with a probe, the software can compensate for un-levelness. I wrote my own but there is this which is available free: https://www.autoleveller.co.uk/

Oops that is right. I posted the wrong link above. Correcting now.

 

 

Toner transfer I've done some good ones but where I live don't think they let me do that you can fart next door and smell it.
@AlbertHall looks great to me

Have you taken a look at this post, and at this thread?