Been manipulating the program I'm using so I can save some manual labour, and keep the profile as close to the original design as possible. Sanding does remove a slight variance to originality.
So in keeping with the program I've added time to the finish, but sacrificed cost of finishing.
After my disaster where a bearing seized and fried a stepper motor driver, I'm finally getting my collective sh1t back in one sock.
If it wasn't for utube I'd be buggered.
Not completely sure that the neighbours appreciate the song of my people but it's of little consequence right now as I'm in my zone 
Nice one Paul, have to give you 10/10 for persistence.
Are you taking all that out in one sweep, or several thin slices?
There's a roughing stage where it leaves deeper lines that need to be addressed. Each pass with a 3.17mm bit is limited by choice. I use the recommended 1/3 of the diameter at 1.1mm. It reduces the load and risk of breakage. When I replaced the motors, I took a page from tim the tool man and increased the nm from 1.2 to 2.3 which basically overdoes the original motors. I'd rather have the extra power and not need to use it than the alternative, the steppers using the optimal amps as needed.
The program takes about 5 hours
2 for roughing and about 3 for a parallel finishing pass. There is a waterline and pencil finish which I found to be a waste of time. It's a great learning curve that can be challenging but the rewards are worth it in my opinion.
This is what happened when the ball screw bearing seized.
Ok i can see the fin shape. But what am i looking at?? Did cutter rip a whole section up?
Select to expand quoteolskool said..
Ok i can see the fin shape. But what am i looking at?? Did cutter rip a whole section up?
Yep, the ball screw bearing seized and refused to allow the spindle to lift. The result is an increase in depth at each g code command. Eventually going right through the job.
A trophy Stu? Maybe the "Fangy limestone finders fail" 
It's certainly a good reminder to check the bearings and peripherals constantly.
Pepe, I am wondering whether the Sound of your People could be enhanced with Angle grinder on Aluminium backing tracks. I am happy to come over and jam a few riffs so your neighbours learn to love when you are at work 
Mate, I'm sure they love that anyway. But I've got a subwoofer that I'm not using right now that would amplify the sounds of swearing and dropped tools...
In between ball nose endmills that take forever and regular endmills that leave a poor finish, you might find cornered radius endmills to be a convenient middle ground.
Select to expand quoteJulienLe said..
In between ball nose endmills that take forever and regular endmills that leave a poor finish, you might find cornered radius endmills to be a convenient middle ground.
Thanks Julien, I'll see whats available, as long as they're carbide and don't cost the earth. Bringing the final pass down to .2mm helps, but does take time.
Corner radius are usually mold making tools and very pricey, especially to wreck on fiberglass, but the shop linked below really helped. They have a nice selection of diameter/flutes/radius at great prices and the endmills were, as far as I remember, good.
www.shop-apt.co.uk/carbide-end-mills.html
I recall Mal Wright complaining about going through a lot of expensive carbide bits when he was making G10 proto fins way back around '06.
Same thing, He had to go really slow or chew up lots of expensive carbide bits.
I have a box containing a couple of half cut fins that were wrecked when something went wrong or broke, 
It's funny you should mention that as I found that the best bits are the cheap chinese ones. I've tried expensive 3 flute coated bits, 2 spirals, titanium coated bits at around 15 bucks a pop, but the ones that suit cost about 2 bucks each. I do one side, flip the job and replace the bit. Bin the bit and we're off again.
The only thing to watch is finishing.
As it wears down, so does your z-axis zero. If your talking .15mm finish pass it may not be exact. It's reducing as you're milling. Fun
Edit: challenging
What does the carbon dust do to the keyboard? I can imagine all those little switches shorting out at the same time.
I won't ask how come it looks so clean!
The carbon dust gets sucked up by that black hose by the spindle, which is attached to the vacuum cleaner. The finer particles are drawn in and filtered by the overhead dust extractor.
You may be able to see a brush attached to the bottom of the spindle.
That's a cool looking machine .
A ball nose cutter would be best for wear apposed to a flat .Also a good quality solid carbide ball nose cutter should last ages and could run it at a zillion revs. Coated cutters won't be good on glass .
What rpm are u cutting at , can you use a larger diameter for roughing .? If u could go to one quarter inch carbide , it would be easy to hand shape a cutter on a bench grinder and sharpen it yourself .
I'm running the spindle at 370hz and found that it is at it's quietest at that speed. As I said, I've tried the coated cutters and settled on the 3.17. It's by far the quietest and as you increase the size of the cutter, so does the noise.
Realistically the noise could become an issue if it's not regulated. I have used a 3/16 3 flute carbide cutter but it was not good.
Select to expand quoteImax1 said..
That's a cool looking machine .
A ball nose cutter would be best for wear apposed to a flat .Also a good quality solid carbide ball nose cutter should last ages and could run it at a zillion revs. Coated cutters won't be good on glass .
What rpm are u cutting at , can you use a larger diameter for roughing .? If u could go to one quarter inch carbide , it would be easy to hand shape a cutter on a bench grinder and sharpen it yourself .
I have to say, anything that I would grind would have NO chance at 1500 rpm, let alone 21000. 
What is the machine originally designed to do , and what are the green hoses going to the fans , is the motor water cooled ?
It's a cnc machine. What it cuts is up to you, and the power of the spindle.
The spindle is water cooled, the fans blow through a radiator, further cooling the coolant.
I understand it an CNC milling machine , ive been around and used them for over twenty years . I have never seen one this small or one that is water cooled . Thats why i was wondering if it had some unique purpose , especially the water cooling . Possibly making it quieter running than an air cooled motor ? Or possibly an air cooled motor cant rev to 21,000 rpm ?
I wouldn't be running a 3mm cutter over 10,000 rpm on something soft let alone carbon or glass. Id be thinking half at most or you cutter will melt without an air blast , ( which you dont want ) , or coolant which you cant use .
What revs are u using the cutter ?
It can run at 24000 rpm but after a while I found it's peak reliability at 21. Water cooling just increases the amount of dependability as opposed to air cooling.
I had used a reservoir of 6 litres but increased it to 9 and added a radiator which decreased the temperature by 15deg.c. The pump and radiator run independently from the system for safety.
I use mach3, meshcam and freecad to design, convert and cut.
I have completely rebuilt the controller using md542 stepper motor drivers, db25 breakout board. Didn't like that , so I took it apart again and used bootlace ferrules to attach the drivers and used shielded cable to stop emi, by earthing at each end. Now it runs fine, without losing steps and even sounds smoother.
Edit: normally I will be cutting at 9000 rpm through the laminates with a 3mm bit designed to run a lot faster than that. The karcher also pulls a fair bit of air past the bit.
That may be small, but this one (for the logos) is smaller at 3020.
Incidentally, my triton router runs at 18000 rpm, and the porter cable laminate trimmer runs at 14000 rpm.
Edit: at that speed, its not the small bits that worry me, its the large ones. Like a 1" quarter round.
Wow , 9,000 rpm cutting without cooling through carbon and glass for three hours seems very fast . Do you smell smoke cutting at that speed ?
No not at all, and by watching the path of the bit you can determine if its burning by the colour of the dust, brown-turn it down. White, it's all right.
It's more like 5 hours. 10 with a ball nose cutter.
I did smell smoke once though...
