On Wed, 18 Nov 2009, Anders Wallin wrote: > Date: Wed, 18 Nov 2009 19:48:17 +0200 > From: Anders Wallin <[email protected]> > Reply-To: "Enhanced Machine Controller (EMC)" > <[email protected]> > To: "Enhanced Machine Controller (EMC)" <[email protected]> > Subject: [Emc-users] Mesa 5i20 vs 22 vs 23 ? > > Hi all, > > I have a mill which runs a mesa 5i20 card and I am now starting a > lathe build/retrofit (more on that later!). Compared to a few years > ago there are now three or four variants of these cards available: > 5i20, 5i22-1, 5i22-1.5 and 5i23. > Also, I know there has been some good work on the modular hm2 > fpga-code and emc2-driver for these cards. > > - How stable is the hm2 fpga-code + driver right now? anyone using it > routinely for 'production' ?
There seem to be at least several people using it in production. There are undoubtedly some remaining firmware and driver bugs, but the basics look to be pretty solid. (Thanks Sebastian!) > - Any benefits of the 5i23 (400kgate fpga) over the 5i20 (200k fpga) ? > - The 5i22 cards are more expensive, have a bit more I/O, but isn't > the large fpga overkill for a setup where emc2 runs the pid-loops on > the cpu anyway? The main advantage of the 5I22 and 5I23 is future growth potential. Here are some advantages of the 5I22 and 5I23: 1. The PCI bridge chips used on the 5I23 and 5I22 support DMA. The HostMot2 driver was designed at the outset to eventually support DMA to lower the overhead of card access. 2. The larger FPGAs will support fancier configurations, for example the resolver configuration I'm working on will not fit in the 5I20s Spartan2 chip (Mainly because theres no hardware multiplier in Spartan2) 3. Higher speed, the Spartan3 chips used on the 5I22 and 5I23 are about 50% faster than Spartan2 used on the 5I20. This is especially noticeble when processors are embedded in the FPGA. On the other hand, the 5I20 is probably overkill for controlling a lathe, if 72 pins are sufficient. (GPIO will eventually be expandable via SPI anyway) the HostMot2 firmware supports up to 12 axis of servo interface on the 5I20, so a 3/4 axis lathe is not taxing its capabilities. > > For the lathe I will have servos on the Z- and X-axes, not sure about > DC-brush vs. brushless yet, but Jon Elsons PWM amps work very well on > the mill so I am leaning towards them. Then the spindle will hopefully > be driven by a big 1.75kW brushless servo. The lathe project also > calls for a revolver-type toolchanger which needs one I/O bit for a > pneumatic cylinder, and one servo/stepper axis to rotate the revolver. > In addition I will need the usual jog-pendant I/O: MPG (thinking about > two, separate for X and Z), a few selection-switches, and some > buttons. Live tooling is a dream for the future :) > Looking at the schematic for the mill[1], I think the lathe setup will > be pretty similar and I should be OK with 72 I/O pins. > > regards, > > Anders W > > [1] > http://www.anderswallin.net/wp-content/uploads/2008/09/dc_servo_schematic_2008jan19.pdf > > ------------------------------------------------------------------------------ > Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day > trial. Simplify your report design, integration and deployment - and focus on > what you do best, core application coding. Discover what's new with > Crystal Reports now. http://p.sf.net/sfu/bobj-july > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users > Peter Wallace Mesa Electronics (\__/) (='.'=) This is Bunny. Copy and paste bunny into your (")_(") signature to help him gain world domination. ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
