So I stumbled over this bot on the net and was so amazed that I had immediately had to order one. Now waiting for it. This is so exciting! 
Anyway, I wonder if I can extend the arm to increase the area at the cost of accuracy. Plan is to draw and write A6 (14.8 cm x 10.5 cm) postcards in one go per side and with small borders. The video looks as if this should be possible easily. Any experiences?
Okay, I forgot to mention that I know that modifying one arm will give distortion that I will have to transform away in software…
Hi, Yes, we have experimented with bigger arms, although we’ve always made both arms bigger. As you say, you’ll get a bigger drawing area with a little less accuracy. Pen weight will also become more significant as you lengthen the arm so you may need to consider that. Also the weight of the extension will contribute of course. The firmware limits the arm movement so that it doesn’t hit the body of the machine, but with a longer arm there may be some areas where there’s a chance that the extension might hit the machine body so do be a bit careful with that.
I’ll be interested to hear how you do the mapping - we calculate the angles based on the pen position using the intersection of two circles (which in your case would be different radii) and you could then calculate the pen position for the ‘normal’ arm that you need to send to the machine easily from those angles, but it seems like there ought to be a more elegant solution than that…?
Let us know how you get on - what language are you planning to use?
Yeah, I expected that the really small size has a reason… I’ll try as soon as the device will arrive and will report about it. I’ve not yet thought about the position calculation. I don’t know the your firmware - not even the exact dimensions of the mechanism - so the first idea is to draw a grid, measure the distortion, and reverse it approximately e.g. using piecewise polynomial approximation.
However, in the meantime I saw a few videos with your line-us machine on Youtube and you give another solution: pantograph. Maybe I can find a solution with that, but I’ll have to do the pen lifting myself.
Oh, and I don’t know yet which language I’ll use. Probably Python and / or C / C++.
I think a pantograph done properly could be an interesting option. It would need to be lightweight and rigid - maybe carbon fibre rods…? The dimensions of the arms are both 50mm (so from the servo shaft to the bearing centre for the top arm (shoulder), and from there to the centre of the pen holder). The length of the bottom arm (elbow) isn’t relevant to the calculations.
An approximation approach to the mapping will probably work OK - Alternatively, having though a bit more the simplification of the intersection/angles approach (of course) is that once you know the angles you can just translate back along the log arm by the extension amount.
Looking forward to seeing how it goes!
Okay, I am still waiting for delivery. The device has been shipped immediately after order (thank you!) but it is still on its way from Britain to the continent…
First I’ll give it a try without additional hardware. Could work for my postcard task. I’ll give notice.
Then I came across a possible extension. What about a clothespin when keeping the pen not too heavy? I am thinking about gluing a pole to it which to fix on the arm instead of the pen and holding the pen with the clothespin. I’ll give it a try…
This is really exciting! 
Yes, that could work. Let us knowhow it goes.