Is there a street date for the new site prep guide? I've come across a few sites now discussing un-documented solutions that are looking for "approved" methods.
For example: Take a cabinet with a redundant CIOC carrier and redundant 24VDC bulk power supplies. My gut instinct would have me wire the bulk power supplies through OR-ing diodes and jump the Primary to the Secondary CIOC carrier power supply. That may be the wrong approach though if the two power inputs on the CIOC carrier are isolated and you may want something like that described in KB NA-0800-0089, where the bulk power supplies are not connected through OR-ing diodes to maximize reliability in the case of a high voltage failure.
The other hot topic seems to be shields. Interesting options appear when you run some of the migration extensions into several cabinets, from different CIOC busses. Although I know the shield drains to be on the CIOC column terminators, I can't find a single document showing the terminations.The new 11.3 SP1 media disks just arrived and are the first media to include the white paper on landing the shields to the DC (Reference Ground). Rumor was that this White Paper was rescinded (common impedance coupling!!!). Is that official somewhere?
Thanks,
Travis
Travis, I sent a note around on this. Here's a partial answer on the redundant COIC part:
Best practice would be to supply the Primary to the Secondary CIOC carrier input with two independent power supplies (OR-ing diodes, jumps, etc.) and feed them from at least two different fuses or even better from two different sources (buffered and maybe non buffered) The CIOC will tell you that one of the sources failed and you will have time to fix the issue during both CIOCs will stay powered, so from a reliability point of view, this is the best architecture you can choose.
I haven't yet heard back on the date of the latest site preparation guide or power & grounding info, but will add a comment when I hear back.
In reply to Jim Cahill:
Thanks Jim.
In reply to Travis Neale:
Travis,
there was released 2 new documents regarding CHARMs:
DeltaV™ S-series and CHARMs Hardware Installation
DeltaV™ S-series and CHARMs Hardware Reference
Try to check info inside ( I am not sure you have them ). Regarding power supply - I think you can follow up concept for controller power supply...
In reply to Radomir Pistek:
Hi Radomir. Yes I have those documents. Unfortunately they don't show the things I'm asking about. Take redundant controllers for example: The KB article I mentioned discusses an approach where you connect one bulk power supply to one System Power Supply and the second bulk power supply to the second System Power Supply. For M-series System Power Supplies (with one connection terminal), this makes sense because a high voltage failure of the bulk power supply will only take out one controller. In this scenario the bulk power supplies are NOT connected together with an OR-ing diode. It’s not clear how to translate this philosophy to S-series. S-series System Power Supply carriers have primary and secondary input power connections for each System Power Supply. If you connect your bulk power supplies together for redundancy with OR-ing diodes, why would you need two connections on the carrier? And if you don’t connect your bulk power supplies together to protect from a high voltage failure, will the carrier keep the two isolated? This takes another turn with the CIOC carriers, as a redundant pair is fed by a single carrier with primary and secondary input power. It’s not clear how the system will respond on loss of an power input, or high voltage situation, with the existing documents.
Don´t wait for guadline and use your experience. Only for confirmation - we will discuss now about solution for power supply of 2-position carrier ( not important whether simplex / redundant variant ).
To "choose" right solution you can define scenarios first. Than there can be 2 scenarios:
- bulk power supply is inside DeltaV cabinet / CHARM cabinet
- bulk power supply is outside of DeltaV cabinet ( better there is longer distance / connection between bulk power supply unit and "carrier" ).
- for both scenarios we will use Phoenix Quint PS units ( alliance partner products )
First scenario is easy - you will never get "High voltage" on carrier - Phoenix bulk power supply units have internal "protection". There is not necessary to count with this "issue". There is necessary to protect incomming feeders only ( input voltage for bulk powe supply units ). The same we can apply for CHARM solution. There is short connection from Bulk power supply to carrier ( then minimal influence regarding induced voltage and similar cases ) - this solves bulk power supply internal protection.
For all variants ( with / without diode modules - redundant / simplex bulk powe supply ) you will have to also apply any kind of protection between carrier and bulk power supply ( i.e. fuse ) because of "short" on the way / connection to carrier and carrier itself ( not only 2-position, but also 8-position in case of "common" unit ). Without protection bulk power supply units switch-off automatically ( internal overloading protection ). Than better is to make "star" connection to each carrier instead of "power supply loop" between all carriers ( each carrier has own "fuse" / MCB and bulk power supply unit is able to withstand 6 times nominal current - means in case of "short" is unit able to continually work under "short" untill fuse blew up ( "short" disappear ) - this is unique function of Phoenix bulk power supply units. Above is description for "M-series", S-series carrier has "fuses" integrated into carrier - 8-position and I assume also any kind of diode module for 2-position carrier ( not possible to confirm - no data in BOL 11.3 ).
Second scenario can be used with CHARM solution ( power supply solution for "small" cabinets - 48+ HARM variant, but not for "big" cabinets ). There is necessary to count with distance and apply any kind of protection which depend on where cables are fixed ( outside of buildings / inside and so on ).
If you will have a problem to understand what I means exactly, than open BOL and 2-pos. ( 8-pos ) carrier´s sheets. I would like to only add that for incomer power supply feeders ( for carriers ) you will need to use fuses with higher current than fuses used directly on carriers. In case of solution "fuse per module" all fuses will be the same.
Travis, I saw an email about an updated version coming in December or early January of the DeltaV power and grounding document for proper design and control system electrical power and ground networks preparation. Changes should include:
Thanks Jim! Andre Dicaire did a great job covering some of those items in his Nashville presentation, which piqued my interest in the first place, so if anyone else is reading this that wants a glance of what's to come you can download his presentation from the Exchange site. Some people look forward to the next Twilight Saga book. I look forward to the next Site Preparation manual.
Thanks Radomir.
I agree, use experience! Sometimes though, I visit a site and I'm appalled at the cabinet wiring. The customer is sometimes to blame and may disagree. When I have the vendor's recommendations in print, I can usually save myself a day of arguing.
I may start a new post, or write a blog article about the dilemma I see above. I just need to find the time. High voltage on the output terminals is one of the "common failure modes" for switched mode power supplies, which I believe was the premise of the KBA. While the carriers may have fuses in them now, I would hate to blow both primary and redundant controller power carrier fuses on one bulk power supply failure! That's for another post though.
Thanks again.
-Travis
Travis, I'm hoping your Thanksgiving will be a little more exciting than the site prep manual... ;-) I did pass along your comment so we hopefully have the new version sooner rather than later.
Travis, I also added you as a blog author in the DeltaV group if you want to post it or future ones here in the community.
Yes, solution for 8-pos carrier is clear and there is a question regarding power supply concept for controllers ( but I dont assume there can be any issue on the bulk power supply side - of corse in case of using "intelligent" bulk PS unit ). I think that more important is "changed" grouding concept - there is missing updated documents - regarding this can arise more issues....
Of course -- if you're looking for decades of continuous operation then you need four dc sources for the 2 pair of OR-ing diodes --- along with diversity upstream of the dc sources as far as practical. If decades needed exceeds 3 then no UPS is diverse enough to be more than one of the four, and it is better if the UPS(s) has(have) a diverse source(s) for bypass power, etc.