On Jun 7, 10:41*am, "William Noble" wrote:
"Jim Adney" wrote in message

...

I recently got out my trusty Heathkit TT-1 tube tester to check a
bunch of tubes that had accumulated here. I've stopped using its roll
chart, because it's starting to show its age, and I assume the roll
chart is not replacable. If it IS, I'd love to hear about it.

this does bring up an interesting thought, not totally unrelated - with
today's inexpensive computer interfaces, why not build/market a tube tester
accessory that would plug into a USB bus - all you would need is one of each
kind of socket on the unit, and a couple of power sources (filament and
plate/bias voltages) - so a set of SCRs to choose filament voltage and apply
it to the proper pins, 4 or four cheap D/As to create the voltages (maybe
with an HV op amp to create higher voltages), and op amps and A/D with a mux
to scan voltages and currents on all the pins of every socket - this would
probably take no more than 50 to 100 parts and a small PC board and you
could have the tube info read from a computer database and have the test
results displayed graphically - transconductance plots, leakage, emissivity,
all those esoteric parameters.

Done as a labor of love, where the NRE is not amortized, it could be
profitable at the $150 to $300 price range - wouldn't this be a good thing?
it would take less space, be more accurate, faster and less error prone than
using a 40 to 60 year old largely mechanical device.

so, who's gonna make it?

** Posted fromhttp://www.teranews.com**

Hunh, it should be so simple!
I suppose a wide range switching supply for the filaments would work,
but it needs to be able to supply between 1V@50mA and 120V@100mA as
well as up to 3-4A at around 6V. Likewise you would need 3 variable
DC supplies for plate, screen and grid bias as well a source of AC
signal for the gm measurements. Also some form of 'free-point'
switching to connect the various sources and measure inputs to the
socket pins [at up to 4A and 500V]. Then you need the detectors as you
say. Methinks the parts cost would be a bit above your estimate. Then
the program to run all that stuff and display the results and,
finally, creating the tube test data tables. Sounds like a couple of
man years work.

I am in the process of adapting a Heath IT-3121 output to test tubes,
and that simply needs a bias amplifier to drive the control grid, an
external supply for the screen and a filament supply. It looks like
the IT-3121, an IP-17 and a socket box will do the job along with the
bias amp. That's only about a week's work.

Neil S.