From: Mark Raymond [m.raymond@imperial.ac.uk]
Sent: 17 May 2005 14:45
To: Ken Bell
Cc: Claire H Shepherd-Themistocleous; Tony Lodge; Martin Sproston; Dave
Cockerill; Matt Ryan
Subject: Re: More protection diodes ???

Hi Ken,

The answer is I don't, but I have a few comments on this, some of which arise from a chat
with Jordan.

The diodes we have looked at do far from the viewpoint of radiation hardness seem to be
up to the task.

They appear to do the job of protecting the MGPA/VFE from voltage transients. Ideally
the maximum excursion at the MGPA input should be no more that 2.7 V. The protection diodes
can't achieve this, but probably in conjunction with the on-chip protection the chip is 
protected enough to survive. I think this could do with a bit more study, but we need more
devices to do it.

The diodes are not manufactured or qualified for radiation hardness, and therefore
there is no guarantee. The devices we have looked at up to now come from a particular
production batch, and it is possible that devices from a different batch might not
have the same characteristics. A way round this might be to negotiate with Philips 
to acquire all our production devices from a single batch, and perform some irradiation 
QA tests on some samples from this batch. I think we would need to do this to satisfy our 
CMS colleagues, and ourselves, that we have minimised any risk by introducing this component.

Perhaps it is worth discussing these points with a Philips representative. 
Certainly it would be good if we could get some more samples. If we are forced 
to a bulk order then we should consider whether it is worth a gamble on buying the 
full amount now (for the complete endcap) if they can assure us that they will all 
come from the same production batch.

This is probably a more complicated answer than you were expecting....

Cheers, Mark.



At 12:24 17/05/2005, Ken Bell wrote:
>Hi Mark,
>Tony keeps (justifiably!) pointing out that as part of umbilical & HV 
>card finalization etc, we should fully equip & test 3-4 supercrystals 
>at CERN.
>To do it completely,
>this means we want 100+ protection diodes.
>Martin Sproston had a devil of a job to get hold of (directly from 
>Philips, if I remember right) the few we have at present.
>Do you have decent contact with suppliers who might be able to give us 
>more - my memory says standard order is a reel of 10000 !
>Thanks in anticipation... - Ken.

------------------------------------------------------------------
  Mark Raymond
  Blackett Laboratory
  Imperial College       phone: +20 7594 7799
  London SW7 2BW         fax: +20 7823 8830
------------------------------------------------------------------