DIY tube tester

[lowell]

I've read about tube testers being overpriced and unnecessary. I still would like to know the life of a tube because sometimes customers wanna know how old and worn their tubes are if they bought an amp used.

Is the lowering of the heater voltage a legitimate way of doing this? I was thinking that I could simply scope the plate while reducing the heater voltage to test tube life. Does that make sense to you all?

Also if there's a test or two that are the most common and necessary I'm all ears. Up till now I've been doing just fine with putting tubes in an amp to see if they work. I guess the usage a tube has seen would be a nice measurement to make.

[Enzo]

Lowell, I don't mean this unkindly at all, but I have to admit I have no idea what you mean. Lower heater voltage to test tube life? I can imagine maybe looking to see if emission holds up when the cathode is cooler? Not sure that really tells anything about how much longer a tube will work.


Here is the deal, power tubes wear out just doing their jobs, like tires on a car, we expect that after so much use, they will be worn out, and we replace them. The small tubes can last a lifetime. Yes, they can fail. SOme do, but many don;t. They do not generally wear out like power tubes do. Nothing they do is very demanding. MAybe a 12AT7 reverb driver works a little harder... maybe. Now they can go noisy or microphonic, they can even stop working completely. But not usually.

I rarely admit it, but I do have a tube tester. It is under the bench and I get it out every few years for some odd reason or other. In the repair biz, time is money, and by the time I can get it out, set it up for a tube, put the tube in it and warm that up and then run through the test sequences, I could have just stuck a new tube in its place and known instantly if that was the problem. And a tube tester really tells only part of the story. What they do pretty well is verify a bad tube is bad. If it tells you there is no emission or that the tube is shorted, I believe it. What it can;t tell you is that the tube is good. It tells you that the tube functions, but it does not put the tube under much of a load or present real world conditions. It won't tell you a tube is noisy or microphonic, or that it has lost its tone. And it won;t tell you tube life remaining either. I have a 12AX7 and it reads a bit lower emissiion than some others. Shall I assume it is fading into old age? or did it come from the factory with a bit less gain than another? Who knows?

You can spend stupid money on a tube tester. But you can also find them at ham fests and yard sales for next to nothing - like $10. SO I don;t go along with overpriced. Like anything else, deals are out there. Like buying musical gear, eh? If you set one on your counter where guys can use it, you might sell some extra tubes they don;t really need. I don;t think that is ethical, but others may feel different.

Let's go back to cars. Look at the tires on yours. Now tell me how long they will last? How about the transmission or the differential? A mechanic can strip them down and tell you if there is damage or if it looks good, but can he tell you when it will die? And how about the battery. Many of them are sold as "5 year battery" or some such. Well, if you have a 5-year battery, do you replace it on the first day of the sixth year? I sure don't. I expect a battery to last the life of the car. Some don;t make it but most do. So what if I told you your battery was four years old so it had a year of life left?

WHen a customer asks me about tube life, I just explain what I said above - power tubes wear out and small tubes tend not to. Expect to replace the big ones now and then. I tell them that replacing the entire "set" of tubes in their Marshall is a waste of money. I tell them to get new power tubes and see if that makes it sound better, and they can always go back and get preamp tubes later if they still want to. I have no way in the world to determine how long his tubes will remain healthy. The average customer is usually pretty happy he got a straight answer from me.

How long will my tube last? I don;t know, how long will your A string last?

Keep common tubes in stock, then do what you been doing, slap in a new one and see what happens. If you are worried about screwing up new tubes, have a few good used tubes handy. How about that odd tube whose matched partner shattered on the floor? You know you are saving it for that single ended something that comes in needing just one tube. Well, you could wait a long time for that.

[lowell]

Enzo thanks. You're confirming what I thought. I had a guy call here pestering, yes pestering me about having a tube tester. I told him the best way is to try new tubes in the amp, not to test them - he wasn't having it. Good to hear it from a reliable source.

[R.G.]

I really wish I could stop doing this, but I can't.

I keep thinking that the world needs an automated tube tester. Given that I can now buy a full, working computer used for less than a single new tube in many cases, it strikes me that with a little work, you could get a massively more effective tube tester than was ever available.

In many cases this is because we only use a half dozen tube types. The 9A pinout duotriode about covers the preamp tubes, and almost all octal power tubes fit the same socket.

So what's needed for a tube tester?

1. Heater Power. No sweat, run it on 6VDC regulated. If you simply must dink with how much X varies with heater voltage, fine, run it from the output of a DAC into a voltage regulator.
2. Plate supply, variable. OK, DAC running to the reference voltage of a high voltage regulator.
3. Screen supply. Ditto
4. Bias voltage - DAC into an inverting regulator.
4. Measure plate current, screen current, etc. Resistor sampling into ADC input.
5. AC gain. Make a nominal gain circuit, plug it in. Apply a sine wave to the grid, take AC voltage off the plate through a cap/resistor divider arrangement then ADC it.
6. Noise. As in 5, but make the input zero. May even have to switch in an amplifier stage to bring the noise up to measurable.
7. Microphonics: As in 6, but turn on a surplus pager/phone vibrator motor to wiggle the socket.

Yeah, that's a tall order for the junior amp hacker, but for the life of me I can't figure out why some tube fanat... er, enthusiast! hasn't done this. The computer it's plugged into lets you program it in visual basic and spit back all kinds of interesting data. Shoot, a sound card doing the input and reading back the down-divided output lets you run distortion curves, response curves, impulse response, yada, yada, yada.

Matching tubes? Let the machine do it, including spitting out a sticky-tag with the measured tube parameters on it to tag the tube. Tube drift/aging? measure it new, measure it later, compare the tags. Maybe offer this as a high end service to amp nuts.

So many circuits, so little time. sigh.

[Gregg]

Steve Bench has some tube tester articles and schematics on his page.

[Albert Kreuzer]

Yeah, that's a tall order for the junior amp hacker, but for the life of me I can't figure out why some tube fanat... er, enthusiast! hasn't done this.

Somebody has:
RoeTest - Röhrenprüfgerät Röhrenmessgerät
In German, but you get the idea.

[R.G.]

That's rather more complicated than I had in mind, but it's that idea. I had in mind something that tested 12a?7 tubes and the common octals only. But yeah - that!

[Steve Conner]

There are a few other modern computerised tube testers. I remember seeing a nice one in a Peli case that cost about $2k.

Considering I tinker with tubes to get away from computers, I'm not about to buy one.

[stokes]

R.G.,take a look at the AT-1000 at alltubetesters.com way overpriced for a tube tester,but it looks like what you are asking for.

[R.G.]

R.G.,take a look at the AT-1000 at alltubetesters.com way overpriced for a tube tester,but it looks like what you are asking for.

Thanks, I'll take a look.

I'm not really in the market for a computerized tube tester. The idea just occurred to me and before I could quit thinking about it, it got semi-designed in my head. A computer driven tube tester needn't be as expensive or complicated as the ones I've seen. You only need a couple of high voltage DC amplifiers (that is, a modified voltage regulator) and some a-d and d-a. The test signal could come out of the sound ports of the computer, the output for noise and gain tests could go into the sound port, and the rest would be pretty simple. I'd be surprised if the cost of parts was $100.

I really do wish I could stop doing designs in my head just because I thought of something. It's a PITA sometimes.

[Chuck H]

I really do wish I could stop doing designs in my head just because I thought of something. It's a PITA sometimes.

Sometimes it gives me something to focus on so I don't have to think of all the other complications in life. It actually helps me fall asleep. But it can be a distraction during the day when I have other things to work on. If I had your head it might be worse.

Not that there's anything wrong with your head. Maybe it's just not a good one for me.

Chuck

[stokes]

I know what you mean,RG.I have often thought of putting together a simpler tester along the lines of what you refer to.I have a decent Triplet 3444,which is a bit better than most since it has some higher voltages than most of the testers out there,but it is what it is.I find it somewhat useful when I get a batch of NOS tubes for matching current and Gm for power tubes.For preamp tubes it is just informational,as knowing the gm for these tubes really tells you nothing.It has a "noise test" feature,but have yet to be able to detect a tube that ends up being noisy in actual use.I often think about using it as a base for something like you are talking about.

[Enzo]

We use this one for 12AX7s. One large wheel for each triode...




(stolen from DIY Audio)

[Billy R aka DynaFreak]

I am with Enzo on this but I do have 2 tube testers I use all the time. One I built myself, one was made by Fender. The one I built myself is for power tubes and it will run all the common octal pwr tubes at about 440 volts and has a variable bias voltage, and a 1 ohm resistor from the cathode to ground. Tip jacks run out from ground, bias voltage, and the cathode fo use with my Fluke. I use it to let me know if octal pwr tubes are working, and if they are reasonably matched. That is about all I need to help me fix amps. Most customers realize that the best time to get a new set of power tubes is when it is in for other service and if the power tubes have some age on them they often will get replaced. I often get requests to retube amps including the preamp tubes and I always tell the customer if they have working, quiet, non-microphonic preamp preamp tubes in the amp they are money ahead and not to change them.

The other tester is a Fender Champ. Every now and then I will run a batch of 12AX7's through it to get a general idea of their relative noise and microphony.