# How Apex Matching® Works

Apex Tube Matching® is performed entirely in-house on our custom-built tube matching system using our custom-designed software and testing methodology. What this means is we are not matching your tubes on inferior matching hardware or using outdated equipment. Rather, we are using state-of-the-art equipment designed and built to our specifications using our years of experience in the music industry. Our designed systems are highly robust and extremely accurate, measuring current and voltage to provide the best possible matching for our tubes. These systems are a result of many, many years of experience and an incredible amount of planning, design, prototyping, and testing resulting in the best tube matching available in the entire industry. We know you'll find our tube matching exceeds your expectations in every way!

## Plate Characteristics

When we match vacuum tubes using the Apex Tube Matching® method, we begin by running tests on all of our tubes to gather data on their Plate Characteristics. A vacuum tube's plate characteristics are a visual representation of the way plate current responds to changes in plate voltages at a constant screen voltage. These changes are represented as a curve on a graph. A full set of plate characteristics consists of multiple curves - each one representing a different value of grid voltage. By graphing these multiple curves, you can get a good idea of the way the tube should perform at various voltage values. A generic example of a tube's plate characteristics is shown below:

As you can see in this example graph, there are 10 lines, each representing a different grid voltage, with grid voltages closer to zero volts creating higher lines. For example, Grid V1 may represent 0 volts, where Grid V10 might represent -50 volts.

## Determining Valid Plate Characteristics

With vacuum tubes, there is much variation in the way the plate characteristics actually perform - this is the reason why we match tubes. With variations like these, there will always be vacuum tubes that fall too far outside of the expected variation to be considered quality tubes. Tubes like these do not perform to the standards that we expect at Apex® Matching; we perform in-depth analysis of our tubes not only to perform excellent matching, but to be certain that all of our matched tubes fall within acceptable ranges of variation. It wouldn't matter if your tubes were matched if none of them performed properly!

In order to determine what is an acceptable tube, we begin with manufacturer-provided specifications. This gives us a very good indication of the plate characteristics we can expect to see from our tubes at set voltages. Using these specifications, we begin formulating our test points and voltages (as described below). Using these test voltages, we can check our resultant plate characteristics against the manufacturer-provided specifications to be certain they match. In addition to confirming our tubes match what we should be seeing, this allows us to verify that our testing equipment is functioning properly and up to specifications.

After we determine these baseline settings, the next step is for us to determine what we consider "within variation" for providing a quality matched tube. We strive to provide only matched tubes that perform to the levels set forth by the manufacturer and which are expected of our tubes. Before matching even our first tube, we first run hundreds of tubes through our matching process at our determined test values. This gives us a very good baseline to begin analysis. Using this baseline, we can begin calculating the standard deviation of our tube tests. Standard deviation is a measurement used to quantify the variation of a set of data values. A very low standard deviation indicates the value lies very close to the expected value for the tube, where a high standard deviation indicates the data points are spread out over a wider range of values. Below is a textbook example of the plot of a bell curve of values. Each band in this image has a width of one standard deviation.

We expect to see a very similar bell curve of our test values. Below is an example of the type of distribution we see with one of the test values. This is a plot of the hundreds of tests we perform at a single point on the graph.

Using this analysis, we can determine the suggested minimum and maximum value for plate current at these test points. Anything falling outside of this range can be considered "outside of variance". This does not necessarily mean the vacuum tube is "bad"; in fact, the vacuum tube will still perform perfectly well. However, it does mean the tube is likely not qualified for matching as it falls too far outside of the standard deviation. We employ additional analysis to determine what is considered a "bad" tube. These rejected tubes are returned directly to the manufacturer, removing any risk the customer might receive a tube which operates poorly due to manufacturer defects.

Once this analysis is performed, we are ready to begin matching our tubes. However, the process does not stop here. As we continue to test tubes, we can refine our standard deviations even more closely, as well as detect any major fluctuations in "expected" values from our tubes. This allows us to spot any changes in the quality of the tubes - if we see a large change from the "normal" values from the tube, we take extra care to clarify the reason for this change with the manufacturer. This prevents any large-scale manufacturing issues from affecting the quality of our matched tubes.

## Testing Points to Validate Plate Characteristics

When performing Apex Tube Matching®, we perform 5 tests on each tube in order to gather data and to confirm the vacuum tube falls properly within expected plate characteristics:

• Emission Test Point - this test point is used to confirm the tube is performing properly at a high emission point. In order to test this, we use a low plate voltage and a grid voltage closer to zero for the tube. We then check the resulting plate current to make sure it is within range (represented by the bars above and below the test point). If this plate current value is out of range, we reject the tube for performing poorly under high emissions. This test is not performed on traditional matching systems.
• Control Points 1, 2, and 3 - these test points are used to gather baseline plate current values at common plate/grid voltages for use in amplifiers. This allows us to be certain that a tube is matched at any point that a guitar player may be using it. If the current falls outside of an expected range, we reject the tube. We also use the "Transconductance Point" as a "Control Point 4" for our matching. See below for more information. The plate currents measured at these three test points will be later used for matching (see below.)
• Transconductance Point - this test point is used to gather a plate current which can be later used to calculate transconductance for matching purposes. This plate current is measured at the same plate voltage value as control point 1, but at a more negative voltage value. If the current falls out of the expected range, the tube is rejected. This point can be thought of as "Control Point 4", but we specifically use it to test transconductance. See "Matching the Tubes" below for more information on transconductance.

Each of these tests is conducted once the tube has reached a stable plate current, meaning the current values have normalized and we are not detecting any fluctuations in the current. In order to do this, we warm up our tubes before we begin our test process. Each tube is given a significant operational time on the tester in order to warm up and stabilize itself before any measurements are taken. Once this "warm-up" period has passed, we begin taking measurements of the tube current.

We take several measurements of the current to be certain all measurements fall within a certain tolerance of one another. If the current values detected are falling outside of this tolerance (current is still variable), we continue to take measurements of the plate current at the test point until the current stabilizes. If the current does not stabilize after a certain time period, the tube is rejected as unstable. In this way, we can be certain the current values measured at each of our test points are accurate for the tube. Below, you can see a visualization of how we determine a "stable" current.

This image represents multiple measurements of current. You can see the current does not become "stable" until measurement taken at C11. At this point, the previous 4 measurements (C7, C8, C9, C10) all fall within an acceptable tolerance for the measurement. Once this occurs, we can take the current measurement as the average of the measurements falling within the tolerance (C7 + C8 + C9 + C10 + C11 / 5). This gives us a stable and accurate current measurement. As you can see, by following our process to carefully evaluate each of our tubes correctly and accurately at these 4 test points, we are able to confirm the vacuum tube is operating within its expected plate characteristics. Any vacuum tube that does not fall within these expected values is removed from the Apex Tube Matching® process, guaranteeing you see only quality, high-performance results. Once our vacuum tubes have passed this process and we have gathered the appropriate data, they are passed on to matching.

## Matching the Tubes

Once tubes reach the matching stage, we have already gathered and stored test data for each tube. Using this data for a very large volume of vacuum tubes, we are then able to pair tubes together which are very, very similar in their plate characteristics, giving you an extremely high-quality matched set of tubes. In order to perform this matching, we evaluate the vacuum tubes for multiple values:

• Plate Current at 4 points - The value for the tube’s plate current is matched at the "Control Point" tests - Control Point 1, Control Point 2, Control Point 3, and the Transconductance Point.
• Transconductance - The tube’s transconductance is defined as the change in plate current between the "Control Point" and the "Transconductance Point" divided by a corresponding change in Grid Voltage. For vacuum tubes, this value is calculated while holding the plate voltage constant. The resulting unit for this calculation is a mho if the current is measured in Amps and the voltage measured in Volts. In most cases, mho is an extremely small unit. Thus, we use the more conventional unit μmho for our transconductance values.

#### $$\text{Transconductance} = \lvert \frac{\Delta I_p}{\Delta V_g} \rvert$$

On the plate curves, this can be interpreted as the vertical "spread" between constant grid voltage lines at a particular plate voltage and current. The wider the spread, the higher the transconductance. All tubes show an increase in transconductance as the plate current goes up, which is why the grid lines are further apart near the top of the curves. Most tube testers are unable to measure the transconductance at a specific operating point, making our methodology much improved.

We evaluate these values and match them with other vacuum tubes having identical values within a very, very small tolerance. The result of this process is a set of tubes which you can be confident are very nearly identical in their plate characteristics.

Here is a simplified example showing the improvement of Apex Tube Matching® over traditional matching systems

Tubes in a traditional matching system (single point) are considered matched as they are identical at the matching point. There is no additional test performed.

In this simplified example (showing only 2 match points rather than 4), Tube #1 is not matched with Tube #2 - the tubes match only at the first test point, but not the second. This means you can be confident that tubes matched through our process will match much more closely than in traditional matching systems, as the matching has been performed at multiple points along the tube curves, ensuring a real match at all voltages.

For a simplified understanding of how the Apex Tube Matching algorithm works, check out our video on the Apex Tube Matching algorithm with Daniel Moberly, one of the system engineers behind Apex Tube Matching.