How We Test Our Tubes

    One of the most difficult tasks in reselling used tubes is to make sure they are fit for useable service for a long time to come. After reviewing many tube testers on the market, both used and new, Classic Valve Design decided what was best for quality control was a tester of our own design.

    The basic tube tester measures for shorts and emission. Better tube testers measure the actual transconductance of the tube. Transconductance is always measured with the dynamic AC tube parameters and are very accurate to simulate the tube in actual use. The fanciest testers will generate actual curve traces, or measure the dynamic DC parameters of the tube.

    AC and DC parameters are closely related. AC can measure the transconductance, usually with an AC signal of 50 or 60Hz (line frequency) or multiple thereof (100 or 120Hz is also used) or something higher, like 5KHz as in the famous B+K 747. While this is useful, it cannot measure for every instance of AC operation. NO TUBE TESTER CAN. The curve tracer testers are accurate in that they can generate a set of curves that can be compared against what the tube manufacturer has published.

    We have designed and used a DC grid-shift tester for tube testing and matching. The plate is tested at a current draw near the tube's design maximum ratings. This is a very accurate way of determining life expectancy. We look for a rock steady plate current draw that must equal or exceed 85% of the design maximum current for the tube being tested. Below is an image of the actual meter template used on our tester:

Tester Meter Template

    But emission testing by itself isn't good enough. We can have great emission, but a bad grid that is unable to control it. The grid-shift test allows one to set a negative voltage that corresponds to the manufacturer's specified cutoff voltage for the tube. This negative grid voltage is set according to the tube being tested, then the grid is shifted from 0V to a specific negative voltage and the tube must cutoff (the green area labeled "grid" on the meter template). This area allows for a 5% variation, given the wide variety of manufacturers and tolerances. Our tester also allows for more than one grid to be tested. This allows accurate testing of multi-control grid tubes.

    Since this type of test is basically a manual DC curve tracer, it allows us to match tubes for DC parameters. This is accomplished via multi-point matching.

    Multi-point matching is accomplished by measuring the cutoff characteristics of a particular tube throughout its linear range. The image below shows how this is accomplished:

Multi Point Matching

    A typical 12AX7A will be fully cutoff on out tester with a grid voltage of -2.5volts. The first meter reading bordering the grid green area will be -1.7 volts, the centre scale about -1.4 volts and bordering the upper green, -1.3 volts. When we declare a pair (or more) tubes matched, the DC grid voltage reading MUST be within 5% of each other. Typical tube variations can be 10 to 15%. These tubes are still good, but we will not sell them as matched.

Frequently Asked Questions

Is your tester accurate? How was it calibrated?
    Yes. When creating the charts for the tester, we used a large sample of each type of tube that we come along or have in stock. This is usually at least ten of a particular type, from various manufacturers. This allows us to create an average for "Good/Bad" and the matching points are based of experience. Also many of our tubes sold have been tested against a TV7/U tester and found the results accurate.

My tester gave a different reading, what does this mean?
    No tube tester of different brands will read the same. Among the lower quality testers (emission only types), there can even be a difference between manufacturing runs within the same brand. The reason for this is in the way the tubes are being tested. Our tester will read similar to any that uses a constant current source as a plate load sourced by 180 volts DC.

How does your tester work?
    It is based on the DC grid-shift principle.