Testing Details
Notes and photos from Mitch about how he builds the amps.

Various checks and tests are done during and after a build. Once I have the power supply wired, I'm able to test the amp for the first time. I check all my wiring several times, both visually and with an ohmmeter to make sure everything that is supposed to be connected is connected and there are no shorts. I also make sure the polarity is right on the diodes, if the amp has them, and on the electrolytic capacitors.

Then I apply AC power. No tubes are in yet, of course. I check to see that the pilot light comes on and I watch for smoke.

Smoke Test

If it passes the "smoke test" (they always have so far!), I measure the voltages. I check the AC high voltage coming out of the power transformer, also the AC heater voltage. If the amp has solid state rectifiers and they are already connected, I check the DC high voltage too.

If the amp has a tube rectifier, I plug it in now and check the DC high voltage.

Voltage Test

After the power supply checks, I switch the amp off and wait for any capacitors that have been charged to fully discharge. This can take several minutes, but the amp isn't safe to work on until the capacitors are completely discharged. If there isn't any load on the power supply at this point, I'll clip a 220K resistor across the main power supply capacitor to help discharge it.

After the chassis wiring is done it's time to go through the final checks. I check my wiring very carefully several times, again visually and with an ohmmeter. I apply AC power and check all the voltages. They will be somewhat high because no tubes are in yet, but there should still be no smoke.

If everything is still good I power it down and wait for the capacitors to discharge. Then I plug in the preamp tubes. I power it up again and watch for them to glow normally. There should still be no smoke. I measure all the voltages again. This time they should be closer to the expected values.

After the preamp checks, I power it down, wait for the capacitors to discharge, then I plug in the power amp tubes and a dummy load. You should never run a tube power amp with no load. It can damage the output transformer. I apply power again in the Standby mode. If all the tubes glow normally, I flip it out of Standby. If there was ever a time for smoke, this would be it!

I've never seen a smoke show yet, though. Things to watch for now are red plating on the power amp tubes which probably means the bias is set way too high. Also in an amp with negative feedback there's a 50/50 chance the output transformer primary is wired backwards. If that happens there will be some loud squealing noises because the amp is getting positive feedback instead of negative. Effectively it's become an out of control oscillator.

There's no easy way to predict which way the output transformer should be connected to the output tubes. The phase relationship between the primary and secondary windings varies from one transformer to the next, even the same models from the same manufacturer. The easiest way is to test it in action. One way the amp will squeal, the other way it's quiet.

Once the output transformer is wired properly and all the voltages have been measured and found to be within spec, the next thing is to set the bias. This sets how hard the power amp tubes will work. One way to check how hard a tube is working is to measure the cathode current while the amp is idling. Some amps have jacks on the back for this purpose. On others I use "bias probes" that I made using the 8-pin plugs from a couple of relays, a couple of 8-pin tube sockets and a couple of 1 ohm precision resistors. These aren't as fancy as the commercially available bias probes, but they let me easily measure the cathode current and the plate voltage.

Bias Probe

For example, say I have a push/pull amp using a pair of EL34 output tubes and there's 450 volts DC on the plates. EL34's should idle at around 17.5 watts operating Class A/B in a push/pull amp. Ohm's Law says Wattage / Voltage = Current. In this case that would be 17.5 / 450 = .039, so the cathode current should be 39 milliamps.

Cathode current can be measured by reading the voltage across a resistor placed in series with the cathode and applying Ohm's Law to calculate the current. My bias probes put a 1 ohm resistor in series with the cathode as well as providing an easy way to measure the voltage at each pin of the tube without having to get inside the chassis. I use a 1 ohm resistor because it has no effect on the amp's operation and it makes the current calculation really easy. Ohm's Law also says Current = Voltage / Resistance. If the resistance is 1 ohm, then voltage and current are equal.

Since I'm looking for 39 milliamps of cathode current, I would adjust the bias control until I read 39 millivolts across the 1 ohm resistor on my bias probe. I check it for both tubes. The bias current should be the same if they are a matched pair.

Adjusting the bias can affect the plate voltage so I check it again after setting the bias. If the plate voltage has changed, I have to recalculate the cathode current so I get 17.5 watts with the new plate voltage. I'll readjust the bias and check the plate voltage again. Sometimes it takes a few of these adjustment cycles to get the bias right. The plate voltage changes less and less each time until eventually it doesn't change and the tube is idling at the proper wattage.

Bias Probe

Bias Probe

Next comes the soak test. This is where the amp runs for 24 hours with a dummy load and a steady input signal to make sure everything is stable. The bias and the high voltage shouldn't drift. There should be no red plating on the output tubes and of course there should be no smoke!

Soak Test

Once it passes all these tests comes the ultimate test - plugging in a speaker and a guitar to hear how it sounds. I like to take my time with this test!

Home              Construction Details              About Markin Amps              Contact E-mail: message@markinamps.com

© 2011 - 2017              Valid CSS!     Valid XHTML 1.0 Strict