Virtual TUBE TESTER

[Colas LeGrippa]

Good morning ( to those who don't sleep ),

Everybody knows that if a 6L6 at idle, with 400V on the screen and plate and -20V on the control grid ( fixed bias ) draws 10mA, it shows a lack of conductance and therefore should be discarded. Is there a chart that would indicate : ''GOOD, WEAK, REPLACE'' according to different electronic conditions ( V, I ) of a given tube ?

Colas

[HotBluePlates]

I don't have my hickok manual at hand, but I *think* the dividing line between "good" and "replace" for output tubes was ~60% of rated Gm.

I'm not aware of a handy chart anywhere, but I haven't really looked. You could start with plate curves for the tube, but you'll have to do a lot of math to adjust the plate current for values of screen voltage other than what the graph shows (often around 250v). G.E. data sheets will help, as it has separate graphs for G2=250v and G2=400v.

The real test for Gm, though, is using an a.c. signal applied to G1. I suppose there is a chance the tube will idle a bit low, but still have acceptable Gm.

[Colas LeGrippa]

 Yeah right, tube data charts often show typical operation of 250V. In reality, it is 400-450V most of the time.
I plan to use an old amp as a tube tester. ( 12ax7 and octal type ).  By varying the voltages at the plate and grids, I'd see how's the tube is reacting and I'd throw away those with ''out of the blue'' specs. I have plenty of old tubes and I suspect they haven't been used that much but I need to test them before a new project, or selling them.

Colas

[Colas LeGrippa]

could be, but I heard tube testers weren't as reliable as a real tube circuit with HV, what d'you think ?

[HotBluePlates]

Somebody out there might a schematic on the old tube testers. 

The operation of tube testers is often not obvious from the schematic. You'd need to dive into a theory of operation detailed in a book like Tube Testers and Classic Electronic Gear by Alan Douglas.

Bottom line, most transconductance-type testers apply an a.c. signal to the grid, and use the meter to monitor plate current. The meter is "calibrated" in micromhos (often), which means that is how the scale is marked. The designer ensures that meter deflection due to current corresponds to a given Gm marking on the meter scale.

You could figure the correct probable plate signal with a given grid input based on the tube's characteristics, and apply your own test signal in a real amp.

I plan to use an old amp as a tube tester. ( 12ax7 and octal type ).  By varying the voltages at the plate and grids, I'd see how's the tube is reacting and I'd throw away those with ''out of the blue'' specs.

Well, you can get an idea of roughly what idle current an output tube should have with a given bias voltage, right? Or, take a known good, new tube and plug it in the amp. Make a note of the grid bias and the idle current. Plug in your unknown tube. Measure the idle current. Compare to your known good tube.

Preamp tubes should probably result in a bias across a cathode resistor similar to what is printed on the schematic. Or, plug in a known good tube and note the readings as before.

If you want to measure gain of a tube, apply a known test signal to the grid, and note the output on the plate. You need to know something about the stage in question to avoid applying so large an input signal that the output is distorted and gives a false reading of the gain. Divide the plate signal level by the grid signal level to calculate the gain. Gain is, of course, dependent on circuit design, so you oughta know up front what the stage is designed to deliver, or measure a known good tube first to use as a benchmark.

Power tubes could be measured by using a resistive load in place of a speaker. Apply an appropriate input signal to the grids of the output tubes (or maybe the input of the phase inverter, if you can't supply a big enough push-pull signal). Using the anticipated output power of the amp, calculate the RMS voltage that should appear across your resistive load. Measure the actual voltage across your load resistor, while using an o'scope (or your ears) to ensure the output is essentially undistorted (which will give a falsely high output power indication). Compare your reading with expected output power and/or a known good set of output tubes. Many of the old tube books claim power output tubes are best evaluated on the basis of their output power (rather than gain or Gm); which makes sense because you're judging them on the basis of what they're expected to do.

It's not that hard, really...

[Colas LeGrippa]

Easy, I'd say. What kinda signal would you inject in a PI ? 400Hz, 1K Hz, ( or else )  square or sine wave ?

[HotBluePlates]

Inject whatever frequency suits you. Maybe something within the guitar fundamental range would be good (80Hz to ~2kHz).

Use a sine wave to make sure you can easily calculate peak voltage vs. RMS (unless you have an o'scope, in which case you can directly observe peak voltage). The interest in peak voltage is that you can apply a way-to-big signal easily if you don't consider the bias of the individual stage and how much peak input signal that implies the stage can handle. Too big input -> distorted output -> maybe wrong guess about the stage's gain.

A scope could also show you if the stage can handle a large signal in one direction (maybe in the negative direction), but a much smaller signal in the positive direction. You'd like your input signal to be fairly small while testing gain, so that the input is in equal directions about the operating point, and avoiding distortion in the output. "Avoiding distoriton" is, again, only about maintaining accuracy in your measurements, not about the sound you'd like to achieve from the amp.

[Colas LeGrippa]

thanx, you're a pro.
Colas

[stratele52]

could be, but I heard tube testers weren't as reliable as a real tube circuit with HV, what d'you think ?

Yes tube testers are ( most of them) not made for really test tubes working on high voltage , almost all of them. Also the tube don't have to work hard on tube tester like it do on amp.

Salut Colas,
Ton projet de fabrique un tube tester est intéressant, j'y ai déjà pensé il y a longtemps. Mais après réflexion j'en suis venu a la conclusion qu'il est plus simple et plus "vrai" , plus sûr de tester mes tubes live sur un ampli tout simplement.
C'est peut-être plus long que l'utilisation d'un tube tester , peut importe le modèle mais plus certain.

En effet les tubes testers ne f onctionne pas a des voltage élevés ( 24 -48 volts ) en plus les tubes ne sont pas vraiment sollicités, ne travaillent pas forts. Résultat ,ils sont toujours bons sur un tube tester alors qu'en réalité ce n'est pas le cas. En plus , un tube microphonique  ne seras pas détecté, ça prend du son , un ampli pour l'entendre.

A part pour les court- circuit et pour vérifier les filaments un tube tester est inutile. J'en ai 3 à la maison et  m'en sert rarement.

[Colas LeGrippa]

Merci Stratele, c'est bien ce que je pensais. Une lampe ne peut être réellement testée sans haute tension.
Je commence à me faire des amis francophones sur ce site et j'avoue que ça fait drôle parler français ici !
Bonne journée

Colas

P.S.: 1.ooo

[stratele52]

Colas  le Grippa,
Oui un voltage élevé est nécessaire pour bien faire opérer les lampes ET aussi les lampes doivent être soumis a un travail qui ressemble a leur utilisation normal ,un ampli,par exemple.

Un tubes tester ( la très grosse majorité ) ne fait que faire travailler très légèrement les lampes.  Dans ce cas, à moins que les lampes soient complètement usées ( et tu l'aurais remarqué dans l'ampli, il ne fonctionnerais pas ) , le tube tester te dirais toujours qu'elle sont bonnes alors que ce n'est peut-être pas le cas.
Donc si le tube tester n'est pas fiable on ne devrais pas s'en servir.

Google translate;

Yes a high voltage is necessary to do AND also operate lamps lamps shall be subject to a work that resembles their normal use, an amp, for example.

A tube tester, (the very large majority) that do work very slightly lamps. In this case, unless the lamps are worn out (and you would have noticed in the amp, it does not operate), the tube test would tell you tube  is always good when it is perhaps not.
So if the test tube is not reliable it should not be used.