forward annotation/back annotation problems

[david_c_partridge]

Using 10.1.

OK so I have a design, I've done the layout in Ultiboard, and added some holes to the PCB for mounting it, and done some pin swapping to clean up some routing.

I realise I need to make a change to the circuit, so back annotate the changes.   The pin swaps and addition of the holes are reported as "not supported" - why?

I had to manually disconnect the parts and reconnect the nets to match the pin swaps I'd done - I shouldn't need to do that!

So I've now done the changes I wanted to back annotate by hand (at least the pin swaps - I couldn't work out how to place a mounting hole in MS)., and make the other changes I wanted to do in the first place.

So I now forward annotate - this mostly works, BUT my mounting holes are removed - and I have to place them again!

This really doesn't feel right - the whole thing about integrated schematic capture and board layout should be that a change to the layout can be reflected back to the schematic, and vice-versa.   The whole flow is disrupted if this doesn't work right.

Am I hoping for too much here?

[lacy]

Multisim, to my knowledge, doesn't have any mounting holes as components so when you did your forward annote there were no mounting hole in the new version therfore it seems they were erased. I think what happens is that on a forward annote the existing Ultiboard project is replaced with the newly modified  file. Since your new file had no mounting hole, due to the fact that Multisduim doesn't have them as a component, the new file overwrote the old file and they were removed from the design. This is my take on this and I could be wrong. If I am right then there needs to be a better system on the forward annotes that doesn't erase the old Ultiboard file in favor of a new Ultiboard file.

Try to lock the mounting holes on the PCB in Ultiboard before doing a forward annote to see if that keeps them from being eliminated. This is the only way I know that could possibly help this situation.

On the back annote issue, I will have to investigate this further before I can comment on it. On this issue could you possibly post a file for examination? It would really help isolate what it is that is causing the problem and maybe we could figure out some kind of workaround. In the meantime maybe soemone else may have already run into thuis situation and could help further.

[david_c_partridge]

Hi Lacy,

Thanks for coming back on this.

The mounting holes were locked in Ultiboard.  I understand how the problem came about, but think that forward annotation could be a tad smarter and recognise there are parts on the PCB that aren't in the new netlist (the mounting holes were of course not in the old netlist either), and maybe should ask what to do with them (delete, keep, move off board).   Smarter brains than mine can probably come up with a better approach.

The components that had pin swaps rejected were simple old Rs and Cs.  I'd done the pin swaps because when I initially sent the design through to Ultiboard, I'd rotated these parts (some CW, some CCW, depending on the topology)  to keep the layout tidy.  However I got a tad fed up of rotating these parts in different directions, so enabled pin swapping for these parts in Ultiboard, and swapped the pins so that I would only ever need to rotate a part thru 90 degrees CW.

That's when I found the problem I described.   This should be pretty simple to reproduce.

On a side note.   I rather wish that PCB layout packages would recognise that one end of a resistor, most types of capacitor, and some inductors is the same as the other end, and dynamically swap the pins to ease layout as the component is moved around the board.  Naturally this doesn't apply to polarised caps or to some polystyrene types where there is a difference (outer foil end to be connected to ground or low impedance side of circuit).   OTOH I don't know of any software that does recognise the difference for the polyester types.

Maybe I've missed something here, and EWB is smart enough to do this, but if so I haven't "grokked the fullness" of how to make it work.   Maybe I need to enable pin swapping of each R and C I place in MS, before I transfer to UB?   But if so, shouldn't this be the default for most types of R and C at least rather then requiring me to enable it on a component by component basis?

Cheers

Dave

[david_c_partridge]

Oh I forgot to mention the holes I placed were put there using "Place/From Database", rather than using Place/Hole - this may impact the behaviour.

Dave

[kittmaster]

While Lacy covered a good majority of it, this is how I always taught PCB design.....

While forward does work properly, back has always been twitchy and unreliable, especially when it redraws busses, offpage connectors........what a mess it makes.....put placeholders in your design as 1 pin components for the holes, give it a footprint....done.

Message Edited by kittmaster on 07-25-2008 08:23 AM

[david_c_partridge]

There you have the advantage of me - totally self taught, so none of the "tricks of the trade", and "don't do that's" got passed on to me :-(.

Dave

[lacy]

Chris, you are just full of good ideas. I like the placeholder for holes deal. That was great.

On another note. I have found that the forward annote does not always work as it should either. I have had times when I made changes and Ultiboard would not accept these changes and just ignore them or parts were eliminated that I didn't want eliminate just like this persons holes. But with the placeholer method you outline this last situation may not happen to me again.

Would it be too much of a hassle if you could post an MS schematic demonstrating this?  I think a visual reference would help.

Just for this users information. I am totally self-taught on Ultiboard and Multisim with over 6-7 years experince afterward. So it just takes time to become profiencient at anything and I still don't consider myself an expert. Being self-taught is not a bad thing.

Message Edited by lacy on 07-25-2008 05:34 PM