Mr Rettinger
for your information,I have retrieved the info on the ISDN receiver 
sensitivity test in the TREG data base,
1994 message
best regards
Pierre-Marie 
***********************************************************************
From: Pete Van Raalte 
List-Post: [email protected]
Date: Thu, Oct 13, 1994 4:35 PM 

John may be able to apply the method we use for ISDN Primary rate. John asked 
about the receiver
sensitivity test in BABT 340, Table 3, test 17. We do test 18, which is the 
equivalent for PRI, and we
definitely have a K-Mart equivalent to the expensive setup suggested by BABT. 
(Pardon the commercial
reference.) We loop back through a 6 dB attenuator, and run a test program 
that assures we get back
what we sent out. The software tests all 24 channels for about 10 seconds. One 
bad bit constitutes a
failure. While this test has not been formally presented to BABT, our UK 
consultant feels that it is
defensable as a reasonable substitute. I would appreciate hearing from anyone 
with opinions on the
acceptability of this procedure. 

From: Nick Evans 
List-Post: [email protected]
Date: Fri, Oct 14, 1994 2:31 PM 

>From the work we've been involved with in this area I can confirm previous 
comments that the ISDN
receiver sensitivity test for BABT PQAA certification can be much simplified 
from the type approval
requirements specified in the ETSs. 

For the test BABT has suggested to us a link with 6dB attenuation to simulate 
a maximum line length,
there is no need to introduce further impairments such as noise or jitter. We 
have also been instructed that
the test must evaluate receiver performance at layer 1, ie with no higher 
level error correction distorting
the results. One way we have looked at doing this is to set up a low level 
loopback on the Equipment
under test (as specified in I.430 I am told). A pattern can then be fed into 
the receive path and measured
on transmit path. There are reasonably priced bits of equipment around which 
can do this (although I am
not sure the price falls into the K-MART league). 

I know other companies have also looked at assigning one sample of their 
actual product with slightly
modified software to act as the reference Pattern Generator and bit error rate 
detector. Their proposal in
this case would be that this modified equipment becomes a QA reference sample. 
We get involved with
QA samples a lot when it comes to testing things like Call Routing Apparatus 
(CRA) so I don't see why
any reasonable proposal in this respect would not be accepted by BABT. 

From: Dave Clement 
List-Post: [email protected]
Date: Tue, Oct 18, 1994 8:48 AM 

We do are doing test 18 the same way and it has been accepted by BABT via 
approval of our "Quality
Plan". As a matter of fact most of the tests in Table 3 are not conducive to a 
production line when
compared against the standard refernces. We use the "Quality Plan" as the 
vehicle by which we describe
how we are going to meet all the supplemental requirements in part 2 of 340 
and have submitted it to
BABT for approval. Any revisions are submitted for approval in the same way a 
product design change
is. This approach effectively eliminates the need to "defend" our methods at 
audit time. 

Revised: May 24, 1995

Reply via email to