This is a good FAQ post to keep on your computer.
___________________________________
Posted by Jim McShane (D) on August 29, 2013 at 15:51:50
I have been deluged lately with people asking about issues with their tube gear that they have decided, for whatever reason, must be the tubes. Occasionally, it is, but in my experience, at least 75% of the time it is not. I wanted to share some tips to help people out.
1. The #1 cause of noise problems with tubes is poor contact between the tube pins and the socket. There are a wide variety of possible causes: dirty pins and/or sockets (in the case of a noisy new tube, the socket is the PRIME suspect), socket contacts that have been stretched out or worn out and don’t make good contact, or in a few cases, bad solder joints on the socket itself.
Tube sockets — even excellent ones — are not sophisticated, high-reliability connectors. You won’t see tube sockets used in critical applications because they aren’t reliable enough.
Do yourself a favor — clean the tube pins and sockets every time you change tubes. A bottle of DeOxit D5 and a handful of good old-fashioned pipe cleaners work great for octals and other large pin sockets; for the miniatures, D5 and the tiny dental brushes for cleaning between your teeth (another drugstore item) do a solid job. BTW, D5 is good to 400F, so don’t lose sleep over it because of something you read on the “Hi-Fi Hysteria” forum. And it doesn’t gum up, either. Spray some in a paper cup and let it evaporate— you’ll be able to see for yourself.
Sometimes, you swap a tube, and the new tube is okay. So the tube you removed must be bad, right? NO! Sometimes, the scraping action of removing and reinstalling a tube in that socket is enough to temporarily restore contact, fooling you into thinking the tube is bad. KEEP YOUR SOCKETS CLEAN!!
If you are installing new production tubes into vintage or older gear, keep in mind that most modern tubes have
slightly smaller pins than the original stock tubes did. You may need to re-tension (or, in the worst case, replace) the original sockets that can’t make good contact with the smaller pins.
Honest — only a few % of the tubes I get back for noise are bad. Most go on to play fine in another socket.
2. The metal cans surrounding the tubes serve as shields designed to prevent external fields from inducing noise in the tube. Would a manufacturer spend the money to put them on there unless they were worth it? Will the tube run a bit warmer with it in place? Possibly, but be smart about removing shields!
3. Would you like to use KT-120s in your amp? It’s more than just a current increase in the heater that’s an issue. I’ve had a couple of cases where excessive grid circuit resistance causes unstable bias. The KT-120 has a maximum grid circuit resistance of 51 k-ohms. If your amp is above that, using the KT-120 carries a risk. Just like the EH and Tung-Sol 7591As, some KT-120s will be compatible with excess resistance, but others will not.
4. Don’t fry those expensive new power tubes! If your amp has an adjustable bias, turn the bias current DOWN significantly before plugging in new tubes. The bias that was correct for the older, worn tubes may not apply to the fresh, new ones. And it doesn’t take much overcurrent to damage a tube or shorten its life. Adjust the bias to spec, and make sure the amp is fully warmed up, including both the tubes and the internal parts. The whole amp should be warm, which takes at least half an hour, in my experience.
Keep one last thing in mind — often, you see a voltage “spec” on a schematic at the control grid of a tube. DO NOT TRY TO SET THE VOLTAGE AT THE GRID TO THAT SPEC! Set the negative voltage at the tube grid to the value required to achieve the specified cathode current. If the tube is passing the correct amount of current and the grid voltage is -20 volts, then to hell with the “spec” of -22.5 volts or whatever it is. Making the grid voltage more negative relative to the tube’s cathode is like reducing the pressure on a garden hose to restrict water flow. And if you let up the pressure on the hose, more water flows — that’s the same thing that happens if the tube grid is less negative.
5. Most tube testers do not test power or other high-voltage/high-current tubes at a sufficiently high voltage to be very meaningful. Current matching is pretty much out of the question for all but the few “lab grade” testers that can apply high voltages at reasonable current levels to the tube. What your tube tester says with 150 volts on the plate will differ significantly from what my rig says at 465 volts! Our old pal Steve Melkisethian (proprietor of Angela Instruments) used to say he didn’t even CARE what your tube tester said! It’s a little harsh, but not entirely unreasonable. Remember, a tube’s test results depend on the conditions under which you test it.
6. Other components, besides the tubes themselves, can make your matched tubes seem less well-matched. Power tubes may be matched for transconductance, but it is imperative that if you need matched tubes, they are matched for current - some say plate current, I say cathode current.
Why cathode current? Because between the cathode and ground of many (most?) power tubes, you will find a resistor in the circuit — low value; 10 Ohms is common — and that resistor’s purpose is to make it possible for you to measure the voltage across that resistor and use Ohm’s
Law to calculate the current. Measuring the current with an ammeter requires breaking into the circuit and inserting the ammeter. However, with a voltmeter, we don’t have to break into the circuit; we can measure the voltage across the resistor and determine the current flowing through it. For instance, if you measure 0.5 volts across a 10-ohm resistor, you can compute the current by dividing the voltage by the resistance. .5 volts divided by 10 Ohms = .05 amps, 50 milliamps.
However, keep in mind that not only the plate current but also the screen current flows through the tube cathode to the ground (of course, triodes have no screens, so this doesn’t apply). We’ll ignore situations where the grid is driven positively and current flows through it as well for now.
So that’s 50 mA. Did you just read the example above? That’s the plate current and the screen current combined (unless the tube is a triode). And that combined current is what is causing the voltage across the resistor. By knowing the total current, the cathode current, and the voltage across the tube from the plate to the cathode, we can easily calculate the tube’s total power dissipation. That’s important, since excessive dissipation (measured in watts) kills a tube.
Why would the readings be different on “matched tubes”?
a. The tubes are never perfectly matched — some tolerance exists
b. The screen and/or plate voltages are slightly different across the different tubes due to variations in the resistances of the output transformers’ primary windings.
c. The cathode resistors are never the same value.
d. Etc.
7. Variance in the brightness of the tube heater is normal and not an issue. When using tubes, it is essential to distinguish between the orange glow of the tube heater at the tube's center and the reddish glow of the large internal metal tube plate. The picture below shows one of a set of tubes with a reddish-orange tube plate. That is a problem for sure!
Now, look at the picture below. It shows some variation in the brightness/glow of the tube heaters, both at the top and bottom of the internal tube structure. This is entirely normal and results from minor variances in production. There is no cause for concern.
The amount the heater protrudes beyond the cathode sleeve varies from tube to tube; if one protrudes more than the other at the top, it will be brighter. It can also happen at the bottom of the tube; it depends on whether the heater protrudes more (as it does at the top). Don't worry! It is not a malfunction and is no reason for concern at all.
I just wanted to share a few things that maybe will help people struggling to understand why they had "problems" with their tubes. One of the points mentioned above is almost always the case!
