Retrofitting a 1986 Maho MH400E

Make sure that the AB sensors are 90 degrees out of phase, not 180.
(Don't ask me how I know that's a possible mistake to make)

When one sensor sees a gap, the other needs to be at the tooth edge, not tip.

Oups… good point, I had them wrong (180 degrees). But I have made one of them adjustable to fine tune with Halscope.
Thank you.

It's good that you can learn from my mistake

The problem with my Y axis encoder was low voltage. I adjusted the supply up from 4.75V to 5.1, and now it is working reliably.

Z has been cleaned and reinstalled, but I haven't got it working yet.
Mark.

I can confirm that this will be a good way to put the encoder for rigid tapping. It is in fact similar to what is originally done on my 600E (1990 vintage).

What is done on that is that there is a shaft that is rigidly connected to the shaft going from the gearbox and forward. This shaft goes all the way through the gearbox, and is sticking out in the back. And on that shaft there is attached a 1024 (I think.....) lines/rev encoder with index.

This index is also needed when orienting the spindle for tool change (mine has that too:-)).

For this to work out correctly one have to adjust a parameter in the setup to make sure the orientation is correct. And this will need to be adjusted after the head have been disassembled (like I have done). Tha allignment will be off when you reassemble since you will most likely engage the angeled toothed gears differently when you assemble it back together.

But you can rest assured you are going about this the same way as Maho is doing it on later model mills, so you should be quite safe.

Althoug I would suggest having a higher frequency response in your readers. Remember you are reading the phase relationship continously. And that means you need more than 1/2 frequency margin. I'm not saying it won't work, but I'd be more comfortable with a higher margin than what you have now.

Anders

I spent some more time swapping around EXE and encoders this evening. It is no so easy to cheap the wave patterns on the HAL oscilloscope when either X or Y has no encoder connected, as they both move off a a moderate speed. Z is very stable even without feedback, and only creeps up slowly.

The good news is that the Z encoder is working fine ( at least A dn B channels, I'll need it tuned to look for the index. I isolated the problem to the third axis of Exe. Ordered a 602 D stand alone EXE off Ebay this evening. The seller is even in Austria. Fro €64 including postage, I hope to have this issue closed next week.
Mark