Here is a small sampling of the advertising for Mercury tube testers. The Model 1000, and the Model 2000, are the preferred units. Both are transconductance testers, and are more portable than many other brands. The Model 2000 can test television CRT's with the optional adaptor. The prices above are in Canadian dollars, making Mercury a real bargain for its time, compared to the Jackson 648s we saw earlier. And since this is Canadian advertising above, they should be available North of the border, making any eBay frenzy for one seem silly. These testers were never as expensive as Hickok, Jackson, or even B&K. And they were certainly never as 'rare', although from the number available for auction, I wonder just what the definition of 'rare' really is. Below is a full page advertisement for the Model 1000, courtesy the February, 1962 issue of Electronic Technician magazine. Let's revisit this announcement of this 'newest' tube tester from Mercury.

Looking over the ad; the first claim is the capability to test 'for true dynamic mutual conductance'. From my investigating, Mercury is using a 1VAC test signal; a welcome improvement over the 5VAC used by many Hickok tube testers. However, it appears a tube like the 6V6 is tested as a triode (much like the Hickok 799), and some high-power output tubes (like the 6L6) are tested for emission only! The Grid Emission test is always a good idea, and I have caught 'gassy' tubes more than once. Lastly, there is the claim that '...Automatic line voltage regulation' is a feature of the Model 1000 and Model 2000, although better testers usually have 'Line Adjust' capabilities. As a side note, I always check the filament voltage after the tube has warmed up sufficiently, to 'calibrate' the 'Line Adjust' setting! This is because the line voltage today is not the same as it was forty years ago, and some power tubes draw enough filament current to load down a light-duty power transformer. Without working on a 'live' tester, you could always set the filament voltage slightly higher than the 'normal' value. Taking 6.3VAC as an example, I would check that the 'Line Adjust' gave me at least 6.5VAC with no tube in any socket, or else compensate by adjusting slightly higher than the '12 o'clock' position. The Model 1000 and Model 2000 are very compact; much smaller than the B&K 707, which in itself is pretty compact! However, this compactness comes at a compromise. There is no selection for an appropriate micromho 'scale' as one example. You have to squint sometimes to figure on an 'exact' reading from only two scales on the meter, and know which scale you are supposed to be reading! This is not really that big of a deal, as you should always get a reading 'in the Green', and have a sixth sense of which scale is appropriate. I still like Mercury tube testers, and would not hesitate recommending them to anyone thinking about buying one. Just make sure it's one of the transconductance testers.

I decided to put my Mercury 1000 through a few experiments, just to 'see' the test parameters, and how these parameters compared with other tube testers. This idea for the whole experiment was brought on by a reader who also owned a Mercury 1000. So without further ado, here was what I did.

• I took one (1) NOS Sylvania 'big bottle' 6CA7/EL34, and one (1) NOS JAN 6L6WGC, manufactured by Philips.
• I set up the Mercury 1000 exactly as described in the operator's manual. What is strange is that the manual instructs you to test any 6L6 for emission only, but there is a transconductance test setting for a 5881 and a 7027. I decided to try testing the 6L6 'both ways', using 6L6 and 5881 settings. See the chart below for a summary of all test results.
• The tubes were allowed to warm up for a few minutes, and the reading were duly noted.

From the chart above, we can see a few points worthy of note. First, for high current filament tubes, like a 6CA7/EL34, the filament voltage sags a little. Not a big deal, as I have had tube testers that sag far worse. The lower plate resistance of the 6CA7/EL34 also makes the Plate voltage sag just over 5%; again not that big of a deal. Testing the 6L6WGC as a 5881 loaded down the test signal(!), a phenomenon I have yet to understand fully. At the Plate and Screen notes for the 6L6 tested 'by the book' we see a '*' designation. This is in reference to the fact that as an emission test, all the grids are tied to the plate, and a small AC voltage is applied to this virtual diode. The meter shunt is vastly different for the two 'tests', in respect of the fact that as an emission test we are measuring the rectified voltage at the Plate (16.8VAC), and in the transconductance test we are measuring (at the Plate) the amplified AC voltage which was applied to the Control Grid. The 6CA7/EL34 had an AC Plate voltage of a measured 39.2VAC, which indicates the tube is amplifying quite nicely. And that is exactly why you would use a tube tester in the first place; not to impress your friends with the sight of your $1,200 Hickok 539c, but to tell you what the condition (or the 'quality') of your 6L6 is.