One other issue with these portable instruments is the filter passband
shape, which may not quite meet CISPR requirements.  Does the filter we get
 produce results that -- under the same conditions -- are produced by the
CISPR compliant instrument. IMO under _some_ conditions it will, and under
others, it won't, because for emissions which require the whole bandwidth,
the narrower peak and wider skirts of the portable instrument will deliver
less energy than the more expensive one... but for  emissions which do NOT
require all of the bandwidth to measure (which IMO except for SMPS noise
seems to be the case much of the time) the results for the two filters
should be the same, if thermal noise in the passband can be neglected.

When could we expect to need the perfect filter shape? Broadband emissions,
such as SMPS's, certainly.  Some modulated emissions.  Is this often enough
to stick rigidly to the CISPR filter spec?  I guess it depends on the
strength of the justification, and the amount of margin a product
demonstrates. A 10 dB margin won't be affected by a slight shape-factor
difference... and perhaps this CAN be used if it is taken account of during
the test, and documented in a TCF.

Just some random thoughts... engineering tests are of course another kettle
of worms (grin) entirely.  The matter of dynamic range is a thorny one for
many OATS, because location is determined by convenience and cost, rather
that RF environment. And how many sites use the ubiquitous 8447D _ahead_ of
the $60K instrument and preselector, negating their virtues?  Been there,
done that, and if folks can't be convinced it's the wrong way to go, it
doesn't get changed! That's another thread, I think.

Cortland

====================== Original Message Follows ====================

 >> Date:  19-Oct-97 11:13:22  MsgID: 1055-104952  ToID: 72146,373
From:  Jon D Curtis >INTERNET:[email protected]
Subj:  Re: HP pre compliance analyzers
Chrg:  $0.00   Imp: Norm   Sens: Std    Receipt: No    Parts: 1

Hi Sushil,

Good question.

You asked what's the difference between an pre-compliance system and a
more expensive receiver.  That's a very good question.

Unfortunately HP seems to be too busy walking a double edged market
segmentation sword to give a good answer.  They don't want to bad-mouth
their portible items and they still want to sell a few expensive
receivers.  An even better question might be when is it OK (i.e. you get
the same results) to use a pre-compliance system and when do you need a
receiver?  HPs answer that you need a receiver any time you are a test lab
taking final measurements doesn't sit well with me, I want technical
justification and to date I haven't got it.

Here's how I understand the argument:

The main diference between the HP859X (precompliance) and the Receivers is
that the receiver has a preselector.  A preselector is an upfront filter
which prevents large signals outside of the pass band from overloading the
mixing diodes and other active elements.  If you are testing radar systems
or other active transmitters you need to be real careful about overload.
If you are testing ITE/industrial equipment on an average test site
(NOT NEXT TO A LARGE PUBLIC TRANSMITTER LIKE WBZ),
then you are VERY unlikely to see overload.  Worried about it?  Put a 10
dB pad between the antenna and the HP859X; if the signal drops more than
10dB, you've got overload - increase the pad until the drop on the screen
matches the pad size - Presto: You've added a very frequency insensitive
preselector.  Pads cost about $20 each.

So what happens if you aren't careful and don't notice the
overload?  Spurious signals appear on the screen, harmonics appear HIGHER
than they are.  POINT:  You can't pass because of overload.  If you passed
and you had overload you still passed.

Now if preselectors are a way of preventing most foreseeable overload, why
not put them on everything?  Because the preselector degrades the accuracy
of the instrument.  It's a filter, It's not flat, It's another layer of
uncertainty.  To make the combination of the preselector and the measuring
device meet the CISPR 2dB accuracy takes in HP's words "a lot of
engineering".  Hence the 60k price tag.  But look futher, the receiver has
a specified absolute amplitude accuracy of 2dB, while the HP8591E has an
absolute accuracy of 1.5dB.  I've checked the HP8591E with my power meter.
Except for a 1.5 dB divit out at 60MHz or so, it's within 1dB across the
band.  It's an accurate instrument despite HP's precompliance label.

Let's answer that last question:  When can you use HPs "precompliance"
systems to take final measurements?  When you have proceedures to check
for overload and are testing low power or non-transmitting devices on test
sites with reasonable ambient conditions.

Lately HP has noticed that many in the industry are using the HP859X units
despite the precompliance label, thus cannibalizing the market for their
recievers. It will be interesting to examine the arguments they come up
with to discredit their instrument.  Stay tuned.

Jon D. Curtis, PE       
      
Curtis-Straus LLC             [email protected] 
One-Stop Laboratory for EMC, Product Safety and Telecom
527 Great Road                voice (508) 486-8880
Littleton, MA 01460           fax   (508) 486-8828
http://world.std.com/~csweb

On Thu, 16 Oct 1997, sgour wrote:

>   
> Greetings:
> 
> I have been looking for a HP Signal Analyzer for 
> compliance testing.  The HP 8590 series is only a 
> pre-compliance (non CISPR-16 compliant) tool, yet 
> it is being used in the industry widely for compliance
> testing.  Can somebody explain to me why I need to
> buy the 8542E or 8546A ($60K+) fully CISPR-16
> compliant receivers, when a pre-compliance analyzer
> seems to be the industry standard.
> 
> 
>   
>   RCIC - http://www.rcic.com
>   Regulatory Compliance Information Center
>   
> 
> 



**Primary Recipient:
  sgour INTERNET:[email protected]

====================== End of Original Message =====================

Reply via email to