Barring any needed repairs beyond that, I always go to
the limit with a full proof-of-performance check and some
calibration adjusting, along with a thorough cleaning and
These checks will show any other problems that need
addressing. For the most part, I find this enjoyable and
rewarding work. I mention this so if you do have a minor
problem on your purchase, don’t be surprised. Many
sellers offer a 10 to 30 day return and refund if you’re not
satisfied with the purchase.
The half dozen TE units I mentioned earlier are the
basic items you will want no matter what direction your
specific interests take you. When making a decision on
any of these units, give some thought to the level of
quality you need for current projects and also the level
you may want down the road. There is no sense in paying
for expensive features you may never use. Of course,
accuracy is one important criteria and this subject is worth
dwelling on for a bit.
Accuracy without stability is meaningless. I can adjust
any given crystal oscillator to a single digit of a PPM (parts
per million). Is that accurate? Of course it is! However, 10
minutes later, it may have drifted out of specification; 24
hours later, it may not even be in the same ballpark
because of lack of stability. So, when the seller says its
accurate, that accuracy has to be defined and may have
different definitions for different situations.
For most of our projects, 100 PPM will be more than
satisfactory for some parameters and even 10% is good
enough for many others. On the other hand, for high level
research center projects, any error greater than fractional
PPB (parts per billion) might be far too inaccurate. Usually,
you can find an accompanying manual to download to
check out the factory specs for suitability to your usage.
The general appearance of the TE in available photos
can tell a lot about its history. Are there damaged controls
on the front panel, large dents on the housing, or a totally
grimy and gouged panel? It’s probably been abused in its
Beware that it may need more than average repairing.
Look at the metal connectors for corrosion and/or deeply
embedded dirt and mildew — this most likely means it had
long term storage in a damp environment (again, a
warning sign of possible circuit board degradation). Some
photos show units with so many stickers and labels on
them, it gives them a beat-up look. However, these may
be calibration, departmental, and company IDs that just
accumulated over the years; a careful eye can see through
this for assessment.
Read the seller’s assessment carefully and take
everything with a “grain of salt.” Some folks are honest
and some are not. One common copout is “we do not
know the condition due to lack of apparatus to test it.”
At any rate, these are the guidelines I have used in my
purchases and I’ve been quite successful. I have only been
stung once due to an unobtainable custom band switch,
but I bought the unit for such a cheap price that I kept it
as a part’s donor for a similar unit.
Other items may be advertised as “will not power-up.”
I love these sales as the price is usually vastly reduced.
Granted, it could be something as serious as an expensive
burned out power transformer, but the odds are it is more
likely in the 120V primary section that only needs a quick
and simple fix.
After looking over an ad thoroughly and deciding to
buy, I determine ahead of time what is the maximum
amount I am willing to bid on it. I leave eBay and let it do
the automatic bidding for me, so I don’t have to play
games with other bidders.
Another piece of advice is don’t get caught up in a
bidding frenzy. It is humorous to see folks go crazy with
this. On several occasions, I have seen two identical items
that had very comparable images and descriptions. One
started at $100 and the other at $9.99. The cheaper one
attracted an army of bidders, and the bids escalated
rapidly. As the auction reached day 3, the frenzy was on
and the bidding was up over $150 while the other unit still
sat there at $100 with no bids. The auction week passed
and the initial $9.99 unit winning bid was $225. The other
unit closed without a single bid — go figure!
Let’s touch bases now on the initial items I mentioned
that I considered essential, and give you my favorites for
each category. Starting with the DMM, although there are
many quality ones out there, my favorite has always been
Fluke Aside from being high quality, they just seem to take
a beating and bounce right back. Every industrial
environment I have been in, Fluke was the meter of
choice. When purchasing a DMM — be it bench or
handheld — I would not recommend buying one with
more than a 4-1/2 digit readout. The reason is that these
will suffice for today’s lower standard supply voltage ( 3. 3
VDC and 1.8 VDC) powered chips, and still have all the
accuracy you would normally need. To go beyond that,
they are getting into extreme accuracy and resolution, and
require regular calibration checks to maintain any
confidence of that high level (which can get expensive).
As to power supplies, I have never bought a
commercial supply. Right from the start, I have designed
and built my own, beginning with a high voltage vacuum
tube model and several lower voltage “stick built” solid-state types. Not long after that, in the early ‘70s with the
advent of the 78xx series of regulator chips and the
LM317 ICs, they became one of the simplest and easiest
circuits to put together — especially if you had a junk box
of salvaged power transformers.
All-in-all, I built 15 regulated supplies ranging from
lower voltage/500 watts to 300 volts/30 watts. They have
come and gone over the years, and currently I have
settled on just four in my lab: 0 to 300 VDC at 100 mA;
12 VDC at 10 amps; dual 0 to 20 VDC at one amp with 5
VDC added; and one small low voltage auxiliary supply.
These have served me well in all my projects in the
last 10 years. All but the 12V/10 amp units are shown in
26 October 2015