View Single Post
  #10   Report Post  
Posted to rec.audio.tubes
Big Bad Bob Big Bad Bob is offline
external usenet poster
 
Posts: 366
Default Tube Tester Calibration Needed

On 05/13/16 13:47, so wittily quipped:
On Tuesday, September 15, 2015 at 5:54:07 AM UTC-7, Peter Wieck wrote:
On Monday, September 14, 2015 at 8:07:04 PM UTC-4, Big Bad Bob wrote:


you've given me an interesting idea...


Much snippage.

There are several digital tube testers out there, and even a few that connect to a PC or laptop. Some use existing technology - outboard additions to a Hickok Cardmatic, for example, and some that are stand-alone products. Some are incredibly sophisticated, far beyond the actual need, but for those who believe in them, quite nice. A few here, with reviews:

http://www.jacmusic.com/Tube-testers...pare-Index.htm

http://www.tubesontheweb.com/matic.htm

None of them cheap. If you could come up with a practical DIY solid-state tube tester with the ability to show performance curves and at a reasonable cost, likely you will do quite well. Software is available for existing hardware, you may even be able to derive that as well without having to program from scratch. Given the thousands of tube variants with many levels of "Quality" within each family, and to satisfy many irrational expectations, your software would have to be very robust.

Of course, we are discussing vacuum tubes here - devices that are about as variable as ears of corn, and not hardly precision devices. Generally, I try to explain to tube enthusiasts that there is only one valid test for a tube - and that is the piece of equipment it goes into. A tester will tell one a little bit about *that* tube on *that* tester at *that* moment, but it is only indicative, not fully predictive, of how it will work in-circuit. If you could better that consistently, you would really have something. Take the 6AQ8 detector tube as one example. I have two Dynaco FM3 tuners and several 6AQ8s. Two (2) will work fine in both tuners. Three (3) will work fine in one or the other tuner, but not both. Put the wrong tube in the tuner, and it is silent, full-stop. *ALL* of them test just fine on my Hickok 539B. Go figure, and such are the mysteries of the Vacuum Tube.

The 539B is one of the 'better' vintage testers, and mine is properly calibrated - but nothing is perfect.

http://jimmyauw.com/wp-content/uploa...ickok-539b.jpg

All those knobs, bells, whistles and adjustments are impressive, and it *can* match tubes with the use of an outboard VOM, but it still will not explain why a tube that tests just fine will not work in one item, and work fine in another.

Again, if you can solve that mystery reliably, the world will beat a path to your door!

Peter WIeck
Melrose Park, PA




OK...here is you answer to your mystery. High impedance circuits are sensitive to leakage; AGC, PLL, FM, etc. If the tube has inner element leakage, the circuit may or may not work.

Bud


this makes sense as one possible explantion, sure. there's also THIS
[which leads to possible intolerance of component value changes over
time, and tube curve performance changes]:

https://en.wikipedia.org/wiki/Muntzing

so taking out extra components that IMPROVE RELIABILITY or SENSITIVITY
cuts down on price, sure, but at the expense of 'you can imagine what'.


one of the first 'victims' that seriously fell to Muntzing is the
simplified color demodulator circuits that became EXTREMELY popular in
the 60's and 70's on just about every color TV model, because there was
"no perceptible difference" (or so they say).

I didn't even know about this until recently, when I saw a schematic for
the color demod (with explanation) in a very OLD TV set, that had
multiple tubes involved (all multi-element tubes too). No 'grounded
grid cathode feed' for the green demod on THAT one. It was all
separated, like an 'ideal demodulator' would be, and re-combined with a
resistor matrix to form the tri-color signal. The 'Muntzed' version has
a single 3 element common cathode tube with phase shifters [as I
recall], driving the common cathode with the amplitude signal, one grid
with the Red phase-shift signal, and another grid with the Blue
phase-shift signal [the result is a somewhat-balanced phase demodulator,
which typically will be a single compactron tube especially designed for
this, with 3 triodes - output drives the color grids, and the color
picture tube cathode gets the 'luma' signal directly, resulting in
'color TV']. In this case, green is assumed to be 'in phase' (adjusted
by the 'tint' control), and the other 2 colors have fixed phase shifts
applied so they phase-demodulate properly. The result gives you an
APPROXIMATE demodulation, but "good enough" for Muntzing, with a LOT
fewer parts. And so, there you have it.

for how it's done "old style", check this out:
http://www.antiqueradio.org/RCACT-10...sionDesign.htm

"Color Demodulation 2.x" section describes how the newer method works

[I expect that by the 1980's, solid state sets used color demodulator
ICs that did things "correctly" since it's cheap/small to build silicon
with a zillion active components on it, unlike a handful of tubes
compared to just one].