LPKF Protomat S62 PCB router-Setting universal router milling depth

[djsb]

I have just produced my first board on this machine. However the track widths on the PCB do not match the widths in Kicad.
I have made a test line on the board and measured it with the microscope but the width of the line depends on the position of the tool in the tool holder. Position 1 (marking holes) has the same 8 mil universal router bit as position 2 (milling top layer) as has been requested by Boardmaster and yet they both have to be set up independently. The setting ring on the head is surely a global setting and effects all the tools place in the head. Does the router bit change it's depth whilst the board is being routed dependent on the insulation width needed or is the ring on the head the only way of setting this depth. Also does the depth depend on how the insulation is set up in CircuitCAM?
Can someone enlighten me?
Thanks.

David.

[shanes1007]

Hi David,

Contact LPKF support at 503-454-4229

[free_electron]

any deficinecy in flatness of the pcb material will cause a shift in isolation / routing width...
you need absolutely flat pcb material and position it on top of perfectly flat carrier material. if the board can flex you are in trouble...
the depth / width of a bit is set by pushing the ring up down on the shaft of the milling bit. the milling software uses the numbers you feed it. if you tell it you set it to 8 mils while in reality it is set to 8.5 anything you mill will be 0.5 mils wider than expected ...

that is the problem with these machines. you need ot be so precise and babysit the darn thing to get anything out of it that is even remotely usable.. i had a 95S ... never succeeded making one working board with it. sure , simple stuff with 12 mil tracks .. try a board with full rubout and 6 mil track and gap ...  full disaster. it runs for 24 hours per side and craps out halfway through.

these machines are generally useless.

[Neganur]

24 hours per side?

What kind of board size is that, I'm pretty sure my latest job ran 30 minutes for about 5 cm x 10 cm with a fairly decent component density. Also, 6 mil traces is a bit borderline for a milling machine. Gaps are limited by the width of the tool - 8 mil is absolutely doable (can only speak for the S62). Big FR4 boards (more than A4/letter) are indeed a pain if they're not brand new and flat. The vacuum bed can only do so much.

If you want to do a full rub out for a board with 6 mil traces and gaps, sounds a bit like a weird layout to me - what's keeping you from using copper pours? That track size sounds like digital bus lines. The machine uses whatever end mill you have available to do the rub out, for example 0.5 mm, 1 mm and 2 mm (19 mil, 39 mil and 78 mil). Anything narrower and it has to use one of the universal cutters that depend on the depth they're adjusted to - and yes, they're slow as hell

Maybe the milling machine wasn't the right tool for the job; bit harsh to claim that the machine is useless in general and craps out. Sure, a tool may wear out but you would have to replace and adjust it, then continue the program. For a one-off prototype it's still faster than sending the file to a board house I guess. Would etching have been faster? (considering the sawing, possibly drilling, printing the photo positive film etc)

[Neganur]

The setting ring on the head is surely a global setting and effects all the tools placed in the head. Does the router bit change it's depth whilst the board is being routed dependent on the insulation width needed or is the ring on the head the only way of setting this depth. Also does the depth depend on how the insulation is set up in CircuitCAM?

I think I got used to adjusting the tools after every tool change (except for when the head picks up spiral drills and the contour router). The V shaped universal cutter wears out during longer sessions (about 30 minutes I'd say) so it's a good idea to check the width every now and then if the design is critical. The insulation in CircuitCAM affects which end mill is being used iirc, but has nothing to do with the depth. Not sure if it's a good idea to use the 8 mil cutter for the marking holes phase, a spiral drill should be better unless I misunderstood you.

[free_electron]

24 hours per side?

Full size PCI card ADSL modem. multiple big TQFP's ( 240 pin , 144 pin ). Full rubout is mandatory around the entire analog frontend to reduce capacitive loading / crosstalk and to get the isolation standoff required for telco's. We wanted to run the board on the machine exactly as it was going to be fabbed... the idea of buyin this machine is so we could run different layouts , find the best one and then have that mass produced.

Never succeeded in getting one working.. i had the plating system and the multipress as well. metallisation is so finickey it woudl invariably fail on a few via's. and you had to use their special prepreg with double foil. the top foil had to be peeled off after drilling. in the end it was cheaper just to run the board on 2 day turnaround at the local shop.

to make a 4 layer board was a 2 day job if the machine didnt freeze halfway through a mill job. it would suddenly stop, spindle full throttle and head down. and just sit there. ctrl-alt-delete was all you could do and restart the job. you had no idea where the machine wa in the cycle. they blamed in on the rs232 link. they swapped motherboards two times and in the end even gave us a completely new machine as it had succeeded in bending its own toolchanger ...

I stand by my comment. these machines are useless for anything larger than a 5x5 board or a board with 'old style tech' ( 10 mils and larger track and gap ). try to do anything at the 6 mil and below level and they are junk

[Neganur]

Aye that sounds terrible.

[djsb]

I've been using the S62 for 12 months now and have managed to do around 1000 pcb's on it. Apart from the occasional tool change screw up the machine has been OK. We produce 16 fixed sized pcb's on a standard 9 x 12 panel for our classes. Each panel takes around 1 hour to do using the standard default isolation width. I changed the isolation width to double the default value and it took 5 hours to do 32 boards on a panel. We have found that we have to increase the isolation width to cater for students who have never soldered before and who tend to bridge the tracks easily if the width is too narrow.
We are using this machine to produce boards over the summer each year and will hopefully start to produce individual student boards later on (after we have sorted out the panelisation and how to precess the boards).
We have found 2 sources for universal router bits that are half the cost of the official LPKF bits, Midwest Circuit Tech in the USA and Mega Electonics in the UK.
One good thing about this machine is that it can be left to get on with the job once set up and it's much easier to respond to local demand for boards.

David.

[free_electron]

I've been using the S62 for 12 months now and have managed to do around 1000 pcb's on it. Apart from the occasional tool change screw up the machine has been OK. We produce 16 fixed sized pcb's on a standard 9 x 12 panel for our classes. Each panel takes around 1 hour to do using the standard default isolation width. I changed the isolation width to double the default value and it took 5 hours to do 32 boards on a panel. We have found that we have to increase the isolation width to cater for students who have never soldered before and who tend to bridge the tracks easily if the width is too narrow.
We are using this machine to produce boards over the summer each year and will hopefully start to produce individual student boards later on (after we have sorted out the panelisation and how to precess the boards).
We have found 2 sources for universal router bits that are half the cost of the official LPKF bits, Midwest Circuit Tech in the USA and Mega Electonics in the UK.
One good thing about this machine is that it can be left to get on with the job once set up and it's much easier to respond to local demand for boards.

David.

Agree, for this kind of stuff it may work very well. Single sided ? 10 mil track and gap ? works perfectly fine on these milling macines.

Try doing the 'fineline examples' in their marketing bullshittery and scale it up to a 'real' board and they grind to a screeching halt...

[brett_cgb]

I've found that using the cutting bits as supplied tends to be a frustrating experience. The tool bit depths are all over the place.

Tool bit depth is set by the distance between the top of the plastic ring to the tip of the tool on the other side of the ring. In other words, when the tool is oriented in its normal working position, the distance to the tip from the spindle is set by the top surface of the ring. The spindle is in contact with the top of the ring.

Before I realized how variable the ring position is on the tool as-supplied, I was constantly adjusting the depth limiter on the mill.

I've since had plastic block milled with holes the tools sit in. These holes are at different depths, and extend through to the far surface of the block. Cutters and end mills sit in one hole, drills and routers in the other. I first make sure the tool bits do not extend through the far surface. Then using a piece of scrap circuit board, ohmmeter, and drill press vice, I slowly push the tool through the block (and ring) until the tool just touches the circuit board as indicated by the meter. This very accurately positions the ring on the tool. I'm only really careful with the cutters and end mills, but once the rings are correctly positioned, adjustment to the working depth limiter in the mill are greatly reduced, and are sometimes not needed.

Once I have the working depth limiter set for the cutters, everything else just ends up where the block/vice/meter positions the rings

[Kevin Liu]

Hi,

I recently got an LPKF S62 to replace my old C30s, S62 is convenient for prototyping PCB.
It saves a lot of time on tool change and readjustment. Especially after I installed a DIY camera.
The tool setting depth can be clearly identified through the vision.
I am also making a tool ring setup tools, It can be easy to set up all different tools depths at once.
All tools setup depth on the same depth level so just only complete 0.2mm universal cutter depth setup on the S62 other tools are all aligned.
It is easy to let S62 use third-party tools. because there are many third-party tools lengths between 35.5mm to 38.5mm.

If there are some LPKF ProtoMat users interesting in the third-party application just join my Facebook group.
https://www.facebook.com/groups/lpkfuserclub