Help us design our next Do-more platform!!

[cyounger]

Thanks

[BobO]

More details...we are leaning toward 3 sizes, as follows...

*Model*	*Discrete In*	*Discrete Out*	*Analog In*	*Analog Out*	*High speed in*	*High speed out*	*Ethernet*	*Serial*	*Expansion*	*Power Supply*
A	6	4	0	0	6@100KHz	4@50Khz	None	??	None	DC
A++	6	4	2	0	6@100KHz	4@50Khz	None	??	None	DC
B	10	8	0	0	8@100KHz	6@50Khz	None	Yes	8 cards	AC, DC
B++	10	8	2	0	8@250KHz	6@50Khz	Yes	Yes	8 cards	AC, DC
C	20	16	0	0	8@100KHz	6@50Khz	None	Yes	8 cards	AC, DC
C++	20	16	4	2	8@250KHz	6@50Khz	Yes	Yes	8 cards	AC, DC

Each model would have several P/S and I/O options as appropriate, like DC/DC/DC, DC/DC/Relay, AC/AC/Triac, AC/AC/Relay, AC/DC/DC, AC/DC/Relay, etc.

The high speed inputs and outputs would operate similarly to DL06's, but more flexibly and Do-more-ish. Current thinking is 4 high speed counting/timing channels similar to the CTRIO/2, but with inputs assignable from any of the onboard DIs, and outputs assignable to any onboard DOs...not fixed like CTRIO or DL06...and hopefully fully hardware with virtually no latency between an input change and out output firing. Also leaning toward all inputs being capable of firing interrupts, and some number of fast hardware timers that could fire interrupts. Motion would be akin to the CTRIO2's Axis mode, with perhaps 4 define-able axis blocks (which are only as useful as your available I/O and speed) that would be controlled via an instruction set derived from the CTRIO2's axis instructions.

Again...super preliminary...brainstorming...*will* change. In fact y'all can help change it. :hint,hint:

[DLTimmons]

B++  10  8  2  0  8@250KHz  6@50Khz  Yes  Yes  8 cards AC, DC 

C++

20

16

4

2

8@250KHz

6@50Khz

Yes

Yes

8 cards

AC, DC

 Also leaning toward all inputs being capable of firing interrupts, and some number of fast hardware timers that could fire interrupts. Motion would be akin to the CTRIO2's Axis mode, with perhaps 4 define-able axis blocks (which are only as useful as your available I/O and speed) that would be controlled via an instruction set derived from the CTRIO2's axis instructions.

Again...super preliminary...brainstorming...*will* change. In fact y'all can help change it. :hint,hint:

Add the analog out to the B++ the C++ would be more I/o than need. Use the Ctrio2 software as a motion profile generator, add some basic motion commands and a flexible Serial communication module for communication with things like laser distance measurement and you would have a low cost motion controller that would cover 99% of the stuff I have to deal with. Use Peerlink to communicate with a master PLC I would be in good shape. Let input interrupts handle over travel and thing like that. Let the hardware interrupts give you a steady time base to calculate position correction on say 1 ms. the analog out give you a command signal(velocity or torque) High speed inputs for encoder and analog inputs for position back options like LTV, laser, ultra sonic, ETC.

[BobO]

Add the analog out to the B++ the C++ would be more I/o than need. Use the Ctrio2 software as a motion profile generator, add some basic motion commands and a flexible Serial communication module for communication with things like laser distance measurement and you would have a low cost motion controller that would cover 99% of the stuff I have to deal with. Use Peerlink to communicate with a master PLC I would be in good shape. Let input interrupts handle over travel and thing like that. Let the hardware interrupts give you a steady time base to calculate position correction on say 1 ms. the analog out give you a command signal(velocity or torque) High speed inputs for encoder and analog inputs for position back options like LTV, laser, ultra sonic, ETC.

We'd probably have to give up some other I/O to add an analog output. Not completely sure how I feel about that, but we can evaluate it. If we were able to, it would likely only be one channel.

[DLTimmons]

We'd probably have to give up some other I/O to add an analog output. Not completely sure how I feel about that, but we can evaluate it. If we were able to, it would likely only be one channel.

8 inputs and 6 out would be a workable combo with a single analog out would have to be bipolar +\- 10 volts as most drives take that. a single axis would be fine as long as you could do a following on quad input and a second quad for encoder feed back. you would not want it handle a large amount of I\O any way.

[BobO]

I think we could make it work by dropping 1 DI and 1 DO...for 9/7/2/1. We'll kick it around.

[Controls Guy]

On the lower I/O count models, consider making the analog inputs capable of withstanding 24V.  Then the software can be made to use it as either a DI or an AI.  I've seen several brands do this.  Should be able to be done with outputs as well.  (I've actually done this in a cheesy way when I needed one more DI, and had spare AI's but not DI's, but let's not go there!)

One approach, since you have so much more resolution than before, would be to make it actually a 0-24V analog input.  Or, you could just read analog to 10 or 12V, but be able to take 24V.  You don't necessarily have to be able to differentiate it beyond "over 12V" if you're never going to use that level except as a DI.

[deep6ixed]

I hate to resurrect a dead post, but has there been any updates on this?

[plcnut]

has there been any updates on this?

I received a survey from A-D a week ago, where they were asking for customer input on the new platform. But they obviously are not giving any time frames yet.

[BobO]

This is mostly certainly moving forward. I don't want to oversell it, but it seems like every time we jump into some aspect of Do-more, we find ourselves doing far more than we thought we could in ways we never imagined. I don't want to give details, but an example is interrupts. When we did the H2-DM1E, we took the approach that it was fast enough that we didn't need them. We changed our minds and decided that we wanted them for the next platform, and ended up going way beyond the conventional approach. Again, no details, but envision two tiers of execution with tight I/O and motion integration and you'd be close. We've said it time and again, but it is absolutely true: Do-more is just getting warmed up.

The survey is just to make sure we're addressing the needs before we get so far into the design that things can't be changed. If you have specific care-abouts, we are definitely listening.