[time-nuts] Weird Stuff WareHouse shutting down

[jimlux]

On 4/9/18 5:20 AM, Dave Daniel wrote:
> Responding to Jim's post ( I can't find his original post), a 
> significant advantage to owning "vintage" instruments is that, in 
> general, they may be repaired more easily than later model instruments. 
> This fact was my guiding principle when setting up my lab, and that was 
> based on Jim Williams' "There's No Place Like Home" article that appears 
> as chapter 17 in his book entitled "The Art and Science of Analog 
> Circuit Design". The more recent the design of the instrument, the more 
> highly integrate is it's circuitry. In many cases, that integration 
> manifests itself in the use of VLSI ASICs of one form or another that 
> cannot be found, and if one is able to find one to replace a failed 
> component, the techniques and tools required to swap it out are 
> advanced, perhaps quite advanced. For a corporate enterprise, these 
> facilities may be /de rigeur/, but for the home lab, they are, for one 
> or more reasons, not feasible and the home lab owner must ship the 
> instrument off to some company which can perform the repair or 
> calibration at significant cost. I can repair a Tektronix 7104 
> oscilloscope. I'm pretty sure I can't repair a Tektronix TDS7104.
> 


This comes up on time-nuts a lot...
(and at work at JPL, because for the most part, we buy equipment, not 
rent it, and the original project ate the whole cost - there's no 
concept of "amortization and depreciaton" - it's just how the govt works)

Yes, the ability to repair is useful, and I'll not deny the pedagogical 
value of older, less "automated" instruments - nothing beats a slotted 
line  (or lecher wires) to understand VSWR, etc.

At JPL we have piles of HP8663 signal generators - a fine piece of gear, 
but all more than 20-25 years old - and they're all slowly dying off. 
Likewise, we have tons of HP portable spectrum analyzers, all with 
screens that are hardly readable. But that creates a problem - we have 
racks of equipment that *depend* on the idiosyncracies of those 8663s, 
and because no project wants to redesign the entire test campaign, we 
keep scrounging to find the last few working ones.

If your time is "free", then the hours it takes to track down a 
replacement transistor with the right properties, solder it in, etc. 
makes it a good deal - *if and only if* you didn't have something else 
more interesting to work on.

I'm fairly busy these days, and while resurrecting test gear was 
something I did when I was younger and poorer, today, I don't think I'd 
make that choice.   I am more interested in making the measurement, than 
in learning more about the innards of 30 year old test equipment.  And 
I'd rather spend my time learning the idiosyncracies of modern 
inexpensive equipment - that $900 SignalHound is a pretty nifty piece of 
gear, even if it is "Windows only" and I have to run a VM to talk to it. 
  Likewise, my $500 Ten-Tec VNA won't go to 50 GHz like the 8510 I used 
to use at work (the one where the sweeper hiccups periodically and locks 
up, and is "not economically repairable), but it's a lot cheaper than a 
FieldFox and serves the need I had for it at the time (measuring mutual 
Z between antennas).