Re: CDC Interconnect Concepts

["Fernando J. Barbosa" <barbosa@jlab.org>]

Hi Gerard and Yves,


Gerard Visser wrote:
> Hi Yves, Fernando,
>     I cc to list, this should be on list I think... The referenced 
> drawings could perhaps be posted?
>     Anyway, some comments below -->
>
> Yves Van Haarlem wrote:
>> Hi Fernando,
>>
>> Thanks a lot for your drawings. I agree that a meeting would be very
>> useful, Wednesday/Thursday would be best for me. I like the Al shield 
>> and
>> the grounding rods. I still have to think about the two options but I 
>> have
>> some questions/comments already, see below:
>>
>>
>> On Tue, 10 Jun 2008, Fernando J. Barbosa wrote:
>>
>>> In particular, I find the cable length from the straws to the preamp 
>>> card of
>>> about 8" to be excessive and with limited shielding, from outside 
>>> and between
>>> channels. 
>
> Me too. This is an issue not of ease of construction, but of 
> performance (EMI sensitivity). Yves, maybe it is in fact impossible to 
> do better, but it should at least be recognized as risky and cannot be 
> lightly dismissed just because "it only has to be connected once." 
> This is _not_ Fernando's main point, I believe.
>
Right. I am really concerned about this issue.

> The task of connecting the cables between the straws and the
>>> connectors seems also very difficult.
>> True, but in principle it needs to be done only once.
>>
>>
>>> Also, the EDAC connectors (series 341) are rated at 1200 VAC rms 
>>> which is not
>>> safe.
>> It was proposed by Gerard, maybe he has a comment :-) We also need to 
>> test
>> how easy it is to make the connector gas tight. I looked for alternative
>> connectors (not specified for vacuum because too expensive) but I was 
>> not
>> very successfull.
>
> Not sure exactly what you mean by "not safe," Fernando. In case we in 
> the end use a connector rated at 1200 VAC rms ie. 1700 VDC, it could 
> be necessary to get special approval from your safety committee. A 
> high-pot test at 200% of our working voltage would probably be passed, 
> and considering the very low power level here I would think that is 
> quite enough.
>
> But, on the other hand, what part number are you looking at 
> specifically, Yves. I just now checked EDAC website, for instance this 
> connector looks fairly suitable: 
> http://www.edac.net/downloads/306-316-356%20Series%20Spec%20Sheet.pdf
> It has a rating of 2545 VDC between adjacent contacts! It comes with 
> card guides as an option. Looks good to me.
>
> [Disclaimer: I've not searched exhaustively for all available "old 
> style" card edge connectors to find the one I think is best. The above 
> is just an example of what should probably work well.]
>
The HV rating is borderline for this connector but also keep in mind 
that the spacing between contacts on the PCB side has to be even 
smaller. In any case, you suggested alternative connectors which have 
the required ratings. Also, the HV supply has programmable trip levels, 
so this is more a matter of operation rather than personnel safety.

>>
>>
>>> Presently, according to Yves' drawings, the 26 contact connector
>>> connects 24 straws and 2 contacts are used for grounding the preamp 
>>> carrier
>>> PCB to some ground point, presumably the end plate.
>> Maybe I did it wrong in the drawing but we do not plan to use the two
>> remaining contacts for grounding for the reason you mention.
>
> No, Yves, I think you do... Or at least, I think you should. The 
> preamp ground to the end plate is as critical, more critical than the 
> anode wires connection.
Right.
>
> A connector of 26 pins, single row, 0.156" would work well, use the 
> two end pins for the ground and use the middle 24 for your signals. 
> Or, perhaps better for reliability, if it fits which I imagine will 
> not be a big problem, use a 28 pin connector, 24 signals in middle, 
> skip one pin on each end (let it float on connector, have no/minimal 
> contact pad on pcb, not connected), and one ground pin on each end. 
> This will give some extra clearance against surface leakage or 
> breakdown on the pcb board.
>
Yes, this should work. The grounding from the preamp card to the 
end-plate should be made through this connector. Otherwise, we would 
need additional wires connections to the end plate from the preamp - 
more connections, higher impedance, etc.

>>
>>> The HV to a group of 24
>>> straws  may still be operational even if a wire breaks in a straw 
>>> and this
>>> may create a condition where there is a 3 kV difference between 
>>> adjacent
>>> cables and adjacent contacts on the EDAC connector.
>> I agree, but how do you get to 3~kV, I can only think of <2~kV?
>
> I'm also a bit confused on your point, Fernando - the straw bias 
> resistors are not being sized to withstand 3kV constinuously... So, 
> this must not be an issue, or there are wider implications.
Sorry, my mistake. I should have written 2 kV, not 3 kV.

I think we need to review all this in detail. Perhaps it would be 
advisable to have a conference call next week instead of this week so we 
could all look at the various issues. I am going to ask Slava to 
distribute the pertinent drawings. Is there a good time to have a 
conference call next week?

Regards,
Fernando




>
>     - Gerard
>
>>
>>
>> Cheers,
>>         Yves
begin:vcard
fn:Fernando J. Barbosa
n:Barbosa;Fernando J.
org:Jefferson Lab
adr:Suite #10, 12B3;;12000 Jefferson Ave.;Newport News;VA;23606;USA
tel;work:757-269-7433
version:2.1
end:vcard