Distortion Pedal Testing

[matt]

Part 1

This is a post to walk through the setup of an automated test for a guitar pedal. To start, we’ll aim for the following tests:

• Gain with pedal bypassed
• Pedal noise floor
• Gain with pedal active an all controls at midpoint
• Frequency response with tone knob fully counter-clockwise (CCW)
• Frequency response with tone knob fully clockwise (CW).

For this test, we’re using version 1.81 of the QAAnalyzer application and version 0.99 of the Tractor application. QAAnalyzer is located here and the Tractor application is located here.

When finished, it will probably take about 2 minutes (including operator change-over and interaction) to test a pedal with the tests we’re building here. This is about 30 pedals per hour, or around 240 pedals per day. Keep in mind this is not optimized. A more realistic test in a very high volume environment might involve putting all knobs at 10 o’clock position, measuring frequency response, gain and distortion, and then replicating those tests with all knobs at 2 o’clock position. Test time for that scenario could be around 30 seconds. The primary aim with this tutorial is to familiarize you with the task of creating tests. How you test will depend on a lot on your business.

Onto the task…

First, start the QAAnalyzer application. To ensure a clean startup, let’s begin by restoring the default settings:

Next, start Tractor. Every time Tractor starts it begins with a new project. Let’s save the default empty project as DistortionTest_1:

After saving, we can see the project name in the title bar.

Connect your pedal to the QA401. This is done as show below. Note the output of the QA401 goes to the input of the pedal. And the output of the pedal goes to the left input of the QA401:

You also need to short the unused negative input of the QA401. This will allow the QA401 to operate single-ended. The arrow shows a BNC shorting plug applied. These are usually 50 or 75 ohms. Either will work.

Now, with the pedal active, let’s go back to the QAAnalyzer application, press button at the bottom of the control section marked “Press to Run” and measurements will start:

Since we have no input to the pedal, this is effectively a noise floor measurement. From the display above, we can see the total RMS “power” from 20 to 20 kHz is -93.4 dBV.

Next, turn on the generator by pressing the GEN1 button.

Next, let’s change the level of the generator. To do this, press CONTROL and click on the GEN1 button. That will open the following context menu. Change the amplitude to -35 dBV. That is a good value for approximating a level from electric guitar pickups.

Now, we see a display that appears similar to below. On this screen, we can read two important parameters. The gain is about 8.5 dB, and the distortion is around -8.6 dB. If you are using a different pedal, these figures could be very different.

Next, turn the pedal LEVEL setting up to max. The gain should increase. In the plot below, the gain is shown to be 15.6 dB at max level/gain.

And if we set the pedal level/gain to minimum, we get the following plot where we can see the gain is at -85 dB:

.

Lastly, bypass the pedal and note the gain is around 0 dB, regardless of the knob settings on the pedal:

Thus far, we’ve done the following:

1. We’ve measured the noise floor of the pedal (
   -93.4 dBV from 20 to 20 kHz)
2. We’ve measured the pedal gain in bypass mode (-0.33 dB)
3. We’ve measured the pedal gain when active at max Level (15.6 dB)
4. We’ve measured the pedal gain when active at midpoint Level (8.63 dB)
5. We’ve measured the pedal gain when active at minimum level (-85.03 dB)

Asking Tractor to make these measurements

Tractor can make these measurements very quickly for us. Let’s assume the operator will perform the following steps:

1. connect the pedal to the QA401
2. Run the tractor test
3. Depending on the pass/fail, the pedal will either continue to be processed for shipping (if it passes) or it will be sent for troubleshooting.

When the operator first connects the pedal to the QA401 and starts the tests, let’s first measure the gain with the pedal bypassed.

We’ll start with an operator instruction to center all the knobs on Tractor and put the pedal in bypass mode. To create this step, click on Add New Test:

The test we want to add is called PromptA00. It is the Test Category “Operator”:

Select OK and you’ll see the test appears in the main screen:

For the Prompt Message shown above, enter the text “Set pedal to bypass and center knobs” and then press OK.

The above created our first test.

Now, let’s add another test so that we can measure the gain with the pedal bypassed. To do that, we add another test. This time, we’re looking in the LevelGain category for the test named GainA01:

Click OK. Now, you’ll see the Test Details as shown below:

The above are the defaults. Untick the Measure Right Channel since we’re only measuring the left channel. And then, change the minimum and maximum gain to pass. We know that in bypass the pedal shows -0.33 dB of gain. Let’s say that -1.5 is the minimum we can accept, and 0 dB is the maximum we can accept. You must change these based on your expected results. The Test Details with the edits made appears as follows:

Press OK. Your display should appear as follows. Note we have two tests in the test plan, and we have a Run Test button that is enabled.

Pressing Run Tests we’re greeted with the following:

Note above we see a list of Test Tasks, and we see the Start Button and the message indicating we’re ready to start. Here, the operator would push Start. The operator is then greeted with the first test task, which is a message instructing the operator to set the pedal to bypass mode and center the knobs:

After setting the controls as required, we press PASS. If for some reason there was an error (such as knob wouldn’t turn) we could push fail. But for now, we push pass. The next test above (GainA01-0) will run very quickly. The resulting screen is shown below: Then we’re greeted with the following:

From the above, look closely and you will see the GainA01-0 test ran quickly, and measured -0.34 dB. Because this was greater than -1.5 and less than 0 dB, it is considered a pass.

If we line up the Tractor app and the QAAanalyzer app side by side, the full operator test thus far appears as follows:

In the next post, we’ll add in more of the tests that were highlighted in the intro.

[matt]

Part 2

Above we created a first test for the pedal in bypass mode. Next, we’ll instruct the operator to activate the pedal, and we’ll verify the noise floor, and also the operation of the tone and level knob.

First, let’s add a test to tell the operator to activate the pedal.

Again, we want to prompt the operator to activate the pedal:

Next, with the knobs centered and the pedal active, let’s measure the noise floor of the pedal. This will ensure that the custom won’t be surprised with a noisy pedal due to a manufacturing defect. Recall above we measured the noise floor of the pedal at about -93 dBV (20 to 20 kHz bandwidth). At this point, we don’t know what the expected range of the noise floor might be. But we can use a window of +/- 2 dB and then run 10 units, check the mean and sigma of the results, and then adjust as needed.

Testing Noise Floor

To measure the noise floor, add another test. The test we want is RmsLevelA01

After adding the test adjust the parameters as show below. As with the test above, we want left channel only (as the pedal is mono), and we want to specify our “pass” levels as discussed above.

Testing LEVEL Pot

When testing the LEVEL pot, there are really two extremes we probably want to check: fully counterclockwise and fully clockwise.

Remember above, the gain of the pedal when the LEVEL was max (fully CW) was 15.6 dB. And when it was fully CCW it was -85 dB.

To capture these tests, we’ll first prompt the operator to move the LEVEL knob fully CCW and we’ll make a gain measurement, and then we’ll prompt the operator to move the knob fully CW and we’ll make another gain measurement. We’ll use +/- 2 dB of the limits for these measurements. Both of these tests will use the GainA01 test we used previously.

So, let’s prompt the operator to set the knob fully CCW:

And then ensure the gain is -87 to -83 using the GainA01 test.

And then do the same for the CW test: Prompt the operator and measure the gain.

Next, let’s take a look at the tests we have so far. On the left, we can see the tests that are running, but it’s getting a bit confusing what each test does. We can change the test name to make it a bit easier to see what is going on. To do that, simply edit the “Test Name” for each test in the Test Details section.

With the names edited, our Test Plan now appears as follows and is much easier to understand.

And we can run the test as follows in the video below. Note the test time is about 30 seconds, including operator interaction.

Next, we’ll take a look at frequency response masks as we test the tone knob on the pedal.