How do you identify unmarked tubes?

[Uncle Erik]

Long story short, I came across a box of around 100 tubes for a dollar and could not pass them up.

Problem is, none of them are in boxes and 95% of them have had their markings wiped off.

Is there any way to identify what's what? Lots of octal, lots of miniature tubes, and quite a few rectifier and regulator tubes, as well. If possible, I'd like to use them. I doubt all or most are good for amps, but I do have tube radios and wold probably pass along anything I couldn't use.

If this is a lost cause, any suggestions for what to do with them? I've thought about getting a bunch of cheap sockets and aluminum, then wiring up the filaments to make a lamp or something. Any ideas?

 

[dBel84]

I have some similar tubes - I can make out that they are triodes, diodes or pentodes by comparison with tubes I know the ID of but as to which specific triode ? Some of my European tubes have factory stamps (etched into glass) and there is a code book to try and figure it out but I have generally given up with my lot.

If there is an answer to this, I will be very interested myself..dB

 

[Roam]

Some tubes you can probably identify by their structure, a tube with "mouse ears" is almost certainly a Tung-Sol 6SN7. If you're lucky you should be able to recognize a few tubes this way.

With VR tubes it should be pretty straightforward, put together a simple PSU and use an inline resistor to limit the current, then hook up the power transformer to a variac and slowly increase the voltage until the tube lights up. Measure the voltage across the VR tube and you're done.

With the rest of them it gets tricky, the first step would be to sort them out according to type, in most cases you can tell if it's a triode, tetrode, or pentode from the number of grids it has. Then sort each group by filament voltage by measuring the resistance across the heater pins and comparing to a known tube. Then you can light up the tubes and put them in a curve tracer, and compare the curves to those found in various datasheets.

 

[dsavitsk]

Unfortunately you are not likely to find many mystery "mouse eared" tubes in your box. The reality is that you need to take a look at the internal structure, find the heater with a multimeter, and once you have a guess, hook it up to a power supply and some components. From that, you can estimate some bias points, etc. Some of the guys at the tube asylum know them well enough to tell you what some of them are. Even that may not tell you much as the differences between a lot of tubes are pretty subtle. A tube tester is helpful, too, if you can lay your hands on one.

I recently sorted through a couple thousand old tubes. The NOS ones were still mostly boxed and thus easy, but the pulls were unboxed, many without labels, and many bad. For every Mullard 12AU7 I found (1, and it was dead) or similarly valuable tube, there were a couple hundred utterly useless tubes (let me tell you about my 1U4 collection), and a couple hundred that I couldn't identify. In the end, I did find quite a few good tubes, but I was not able to identify more than a handful of the label-less ones. The ones I could were common and easy -- 6dj8, 12AX7, etc.

Quote:

95% of them have had their markings wiped off.

The 5% that are marked are likely a good hint as to the identity of the other 95%

 

[Cognomen]

Unfortunately you are not likely to find many mystery "mouse eared" tubes in your box. The reality is that you need to take a look at the internal structure, find the heater with a multimeter, and once you have a guess, hook it up to a power supply and some components. From that, you can estimate some bias points, etc. Some of the guys at the tube asylum know them well enough to tell you what some of them are. Even that may not tell you much as the differences between a lot of tubes are pretty subtle. A tube tester is helpful, too, if you can lay your hands on one.

I recently sorted through a couple thousand old tubes. The NOS ones were still mostly boxed and thus easy, but the pulls were unboxed, many without labels, and many bad. For every Mullard 12AU7 I found (1, and it was dead) or similarly valuable tube, there were a couple hundred utterly useless tubes (let me tell you about my 1U4 collection), and a couple hundred that I couldn't identify. In the end, I did find quite a few good tubes, but I was not able to identify more than a handful of the label-less ones. The ones I could were common and easy -- 6dj8, 12AX7, etc.

Quote:

95% of them have had their markings wiped off.

The 5% that are marked are likely a good hint as to the identity of the other 95%

Put em in the freezer for a few minutes, grab em by the pins with hemostats, and shine an led flaslight on the condensation from a side angle while rotating the tube. Sometimes the tube id quickly appears and disappears. Also try freezing longer 6 mintues or so 10 minutes or so and even longer for any that don't form condensation well without more freezing. If you discover the area where the id is but cannot fully read it, mark the closest pin for the next attempts. I unsheath lamp cord and put the sheath on small pins. For larger pins I use a longer section bent around the nearest pin and held in place by the two adjacent pins. If freezing doesn't work, skin oil might. Rub the tube on the oily sides of your noise, forehead if oily, hair if oily. Shine the LED flashlight on it while slowly wiping in one direction. Sometimes the id appears. I've had more luck with this method with octals, loctals which often have the id on top, and coke bottle tubes, some of which have their id above the index. I've had some luck with grapeseed oil, WD-40, and 90 weight outboard engine gearbox oil. Usually the trick is to get a very thin coating. Hands quickly become too oily with these methods. While I did not have fluorescent light handy, I've read some people have good luck with it as well as with magnifying.