Since 1993 Copyright Notice

The unexperienced buyer of course tries to prevent buying second class stuff, but has no other option than to look for the obvious things. Such as: Clean, colorful boxes, photo shopped pictures, particular details, shiny getters, a needle of some old tube tester reading in the "green", and the voices of anonymous "specialists" on the Forums, telling you "how it's done". All these things mean close to nothing. I have seen myself so often used tubes that look and feel like NOS. Just not TEST that way. Tube testing is the only way, and that means at least test for PLATE CURRENT and TRANSCONDUCTANCE simultaniously. Better is test additional also for GAIN, PLATE IMPEDANCE and GRID LEAKAGE. If all these things are good, then the TUBE CURVES are normally good without testing those. Besides, from tube curves you can derive little from the quality, but they are extremely useful for matching.

The following is no good way to identify an NOS tube.

Tthough it must be said a good tube passes this test of course. It is just you can't say if a tube passes at 100% it is an NOS TUBE. Because the question was about NOS TUBES.

Transconductance is 'strong' on a Hickok tester. What does that mean? Like any kind of 'patent' Bench tester, this means such a tube is expected to work in every application, and not fail quickly. So as you can already see, any fine, used tube will pass that. Exactly getting this result was the intended purpose of this tester. Tubes testing in the '?' range doesn't mean the tester doesn't know. It means, such a tube works in none demanding applications just normally.

Below that, the tube is rejected. So in the days before semiconductiors, from this '?' test result, always many tubes were taken out, and saved up by someone for later, knowing these work in most radios anyway. And even today, such boxes with 100's or 1000's of a kind, pop up sometimes, much to the happiness of the fakers.

The following is a very good way to differentiate NOS from used tubes

Go by parameters only!

This means the result of a tube test must be taken at the manufacturer's recommended settings, and no other. The outcome of such a test must be able to be verified by any other tube tester which can do the same testing. That is a solid result

So we have: Plate voltage, plate current and grid voltage.

Then resulting from this, we find the transconductance. So when the tester gives only transconductance, or when you can not set the working point to the data sheet value, the test falls back to the lowest possible level, just saying if some tubes can be used or not.

Here is to judge such results:

For an NOS tube there is always this balance:

Condition 1 + 2 are the best. Balance can be 20% off. So plate current is only 80%. That is fine, as long as transconductance is clearly above 100%. Reason for wanting to have this balance (so the one up, the other down) is that gain and plate impedance often are fine in that case, and these two produce the sound differences between tubes.

Condition 3 is fine when not more than 120%. Above 130%, gain and impedance start to become wrong.

Condition 4 is for used tubes. Below 70%, impedance starts to become wrong. Plate current is below 100%, transconductance is below 100%. Such a tube is useful, if not one of the two is below 70%. Just not NOS. If an unknown tube is in this condition, increase heater voltage slowly, up to 10% more. If now the tube moves to condition 1) or 2) it is a fine used tube. If NOS do this, they are not fine at all. Theese are suspected to have cathode poisoning problems from being over stored. This may be removed by burning in, with a good device. Like this one. The lot should move to 1) or 2).

Lot picking

Sadly, today this has become the standard procedure with wholesale dealers. So they pick out the best tubes from the lot, and put the rest back in original factory cartons. Today, almost ALL tubes sold are like this.

Leaving the seal on, by opening the box from the bottom, or other ways, like melting a self adhesive seal with a hot air gun, and glue it back together.

This is actually easy to prove. If you suspect this, retest the whole lot at the data sheet setting, using fixed grid voltage. Not, that this is the best way to test, but it is the fastest, and SO your seller was using this method.

Put the results carefully in a graph, it is less work as you think, and you have to test them anyway. On the horizontal axis, you write groups of plate current, like when testing 6SN7 when the tubes test from (say) 5,5 to 11mA, make groups like 5.5-6mA, 6.5-7mA, etc, until 11mA. Then on the vertical axis write the number of tubes which fall in these groups.

Such a graph ALWAYS MUST shows a bell shaped curve, with most tubes in the middle. Not with 5 tubes, but with 100 tubes, ALWAYS. You can exactly see then if they picked the best out.

I do that without a graph, I just sort 100 tubes like that, in their boxes, on a table, and that also shows me the bell shaped distribution. I really can recommend this. Once you recognize a picked tube lot, and you know then, you have been fooled at some level, you won't do it any other way any more.