[Halld-tagger] wrong component

[Fernando J Barbosa]

Hi Richard,

I verified that the BOM indeed has the correct part number. Very unusual 
that the manufacturer used a different part without contacting us for 
permission and considering that these parts are not fully compatible. I 
will be contacting the manufacturer tomorrow. Please send me the codes 
etched on the two devices.

I am glad Alex resolved the problem.

Best regards,
Fernando




On 3/6/2014 3:00 PM, Richard Jones wrote:
> Dear colleagues,
>
> Kudos to Alex Barnes, who has been struggling for almost two weeks now 
> to understand the reason why our preamplifier boards have an ugly 
> oscillation in the signal baseline at 100 kHz, as soon as we set them 
> up and bias them for readout of fibers.  We traced the problem down to 
> one malfunctioning component which is a voltage regulator on the 
> preamp board that generates one of the critical dc voltages for the 
> amplifier.  That part was generating 100kHz oscillations on our board, 
> but they would go away under certain weird conditions, which are not 
> compatible with actual running, unfortunately.  Chasing down the 
> particulars of those weird conditions turned out to be a red herring.
>
> Finally Alex decided that either all of the parts are burned out and 
> malfunctioning in the same way on all of the boards, or else it is 
> actually the wrong part that was installed by the manufacturer.  The 
> second turned out to be correct.  We ordered some of the correct part 
> ourselves from Digikey and swapped it out, and voila the oscillations 
> are gone.  After extensive searching online, we have determined that 
> the manufacturer installed the TLV431BSNT1G instead of the NCP100SNT1G 
> that we specified in our parts list. This is confirmed both by the 
> markings on the chip and by the operating characteristics. Both of 
> these are shunt regulators, but they have reference voltages that are 
> different by almost a factor 2.  Now that it is understood, we are 
> back on track with setting up for QA of fibers.
>
> -Richard Jones

-------------- next part --------------
A non-text attachment was scrubbed...
Name: barbosa.vcf
Type: text/x-vcard
Size: 265 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-tagger/attachments/20140306/072e17c3/attachment.vcf