[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: gEDA-user: ngspice shot noise

Bob Paddock wrote: 
On Sunday 02 October 2005 07:45 pm, Al Davis wrote:

On Sunday 02 October 2005 03:22 pm, Karel Kulhavy wrote:

2N3904's 

But if it isn't, I can return them to the shop because that's
deception of customer, isn't it?

No, and I really believe they all meet all published specs. 


I have not caught up with the start of this thread yet, so not all that sure
what is going on here.  However I could not let your statement above pass.

In the early days of ISO9000 when few had heard of it, Motorola
was making a big deal of them just getting an award for Six Sigma
defect management.

Virtually on the same day they sent us tens of thousands 2N2907 transistors,
labeled as such.  Days later when none of the thousands of boards out
of production worked correctly we discovered that in those 2N2907 cases
where really 2N2222 dies.

Of course Mot. was not going to pay us the big bucks that we wasted in
cleaning up their mess by fixing all of those boards by hand.


hehe. Reminds me of when I was chewed out by a production floor manager who said my design was poor because it failed when they used a different vendor for the op-amp. Upon further investigation I discovered they had stuffed my boards with a single comlinear very high speed current feedback opamp where the BOM in the production database clearly called out a part which was a _dual_ low speed general purpose part. Someone had mixed them in the parts racks in the factory.

On the 2N3904 front, I've never really had failures with them which were not due to a design (board, not device) flaw.

One thing you have to watch out for on parts with multiple sources is that they are virtually never the same. If you really compare all line items in terms of typical and min/max specs there is almost always something different. I haven't paid as much attention with transistors because I've always either had extremely loose specs on on the device or I've called out a particular vendor. I have seen some IC's though where reportedly identical parts from 2 vendors differed in terms of ESD. Back in the mid 90's when many fabs were running at capacity, I had to get a bunch of comparators which I had been buying from Motorola from another vendor. Too bad the other parts only had about 500 V ESD vs 2kV.

Now, the bean counters have caused me headaches with capacitors. "Oh, I'll buy this cheaper 100 uF 30V electrolytic". Too bad my part was chosen based on ESR with almost no consideration of capacitance.

-Dan