Schaublin 125-CNC retrofit.

anfänger wrote: What resolution do you need for the encoders?
There is one converter with an nicrosontroller in the wiki:

wiki.linuxcnc.org/cgi-bin/wiki.pl?ResolverToQuadratureConverter

That never really worked.
Certainly not up to the standard that you would want on a Schaublin.

I ended up using a 7i49 instead.

Hi Mark,

i have took a quick look at the 7i49 manual. From the hardware layout (Ananlog INs and Outs ) it should be capable to interface Synchros.
But for this a firmware mod is needed. don't know if this card has an own subprocessor FPGA/DSP or everthing is done by the mesa master card. if the hardware site fits moding the detection in vhdl should be no big problem.
www.researchgate.net/publication/2875822...ing_CORDIC_algorithm

i have no overview of the hostmod firmware, because i haven't need it before.

Maybe PCW could tell if this would be a time consumption task or an ten-liner....

You could simply replace the synchro resolvers with a conventional encoder and drive the motors with a DC servo drive. I have the cncdrive DG4S servo drivers in my 1983 Shizuoka mill . They work fine with the SEM brushed motors and Heidenhain clone encoders originally fitted. The original tachos are no longer required with modern digital drivers, but I left those where they were due to the way they are fitted within the motors.

The original controller in the Shizuoka drove analogue servo amplifiers and DC brushed motors but the Micon control software dealt in digital units of course. So the initial output from the controller was step/dir, with a "positioner" module to compare that with the encoder feedback and generate an analogue (+/-10V) signal for the amps. So step/dir interfaces in analogue servo systems are not unheard of.

db1981 wrote: i have took a quick look at the 7i49 manual. From the hardware layout (Ananlog INs and Outs ) it should be capable to interface Synchros.

I am using a 7 wire Synchro Resolver with a 7i49 as the spindle position feedback on my lathe. That has 4 primary coils and a centre-tapped secondary. I managed to get it to work but can't remember the details.

I think that the STMBL hardware is capable of driving a DC servo. And it works with conventional resolvers, so might be capable of doing what you need with only a minor software rewrite.

Of course the really cool solution would be to port LinuxCNC to the Data General Nova.

Of course the really cool solution would be to port LinuxCNC to the Data General Nova.

LMAO That would be a serious port. Stuff Linux into 16kb !

Here is the latest video:


I am going to be seriously short of space once I cram everything into the pneumatics space. AC servors direct use of single phase is another advatage, so I wont need transformers. I am tending towards AC servos interfaced with analog +/-10V, so I can use a 7i97 and add linear encoders.
Mark

LMAO That would be a serious port. Stuff Linux into 16kb !

That's stuff for an "old rabbit"

in 1997 i coded an complete FAT32 file system / atapi and ide managment in an ATMega with 4Kb .... to play mp3 through an texas 320 dsp.
mempage switching due that you only had an 8 bit data bus, write own mem compression routine because of minimal ram .....

most of todays guys sadyl code memory waste from morning till evening ....