static vs pulse measurement

[Sándor Filotás]

Hello!

 I would like to ask those who have more experience in using utracer.

What are the experiences if a tubeis measured statically, and then with utracer in pulse operation, how different are the measurement results?
Also, how accurate is the characteristics? If I then put it in another pipe tester that measures statically, what are the results?

Thanks

[Hoeberlin]

Hi, in Germany we made a Comparison Test with a Set of Tubes, which were tested with various Tube Testers.

The Tubes ranged from small triodes ( EC92, ECC83, ), Pentodes, ( EF80 etc ), Power Tubes ( EL41, EL84), and up to EL34. We had NOS Tubes and also used and weak Tubes in the Portfolio.

Tube Testers ranged from Roetest( Weigl, Germany, called the Mercedes of modern Tube testing , tests by using real Current ), μTracer ( pulsed Current )

as well as older Testers with free adjusting of all Parameters ( Neuberger 370  and AVO Mk4 ), Tube Testers with preprogrammed Test Parameters ( i.E. Funke W19 ), and more simple Emission Testers, like Neuberger 270 and some others.

Results in compact Form: Neuberger 370, Roetest, Avo Mk4 are very close together within apx 3% Variation, and μTracer also tests in this Range. We made the Observation, that some used and weaker Tubes climb in Current after some Minutes at nominal Current, because they get warmer, which helps the Emission to get a bit better. This is not the case with the μTracer, as there is only minimal Heat produced at the Anode.

Funke w19 and the Predecessor  Models lack a precise Heater Voltage, and have only limited Voltages ( and only +150 is regulated ), this leads to significant Measurement Errors depending on Tube Type and Condition.

Emission Testers like Neuberger 270 can test Tubes, and with some Experience a Tube can clearly be sorted as good or bad. That is, what they are made for. 

You may be aware, that the μTracer 3 has a chance of underheating the Tube, many people put a separete Power Supply for heating for better Results.

Best Regards, Henning, Berlin,Germany, μTracer User since apx 10 Years

Von der neuen AOL-App für iOS gesendet

--
You received this message because you are subscribed to the Google Groups "uTracer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to utracer+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/b4e4b2ce-9d8a-4796-8910-29add31dbf6fn%40googlegroups.com.

[Sándor Filotás]

Thanks, then how reliable is the selection of a tube according to characteristics with utracer?

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/613895990.1601973.1682716863110%40mail.yahoo.com.

[Hoeberlin]

Hi, as long as the Heater Voltage is correct ( You cannot measure the Heater Voltage of the μTracer with a DVM or an Analog Meter !! ) Measurement of a Tube is precise. I  would estimate that to +-2% . I would recommend using an adjustable DC Power Supply for heating the Tube, and give the Tube enough Time to heat, then doing a few tests until the values do not climb any longer.

Many People don't know, that Tubes can have fairly high Tolerances, and everything between 65% and 130% of nominal Current is stated as good! The Circuit has to be designed to be able to handle those Tolerances.

For specific Circuits you may need matched Pairs, and these can easily be tested with the μTracer.

But even then you may run into a Situation, where a matches Pair is no longer equal, when one Tube gets tired earlier than the other.

BR Henning

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/CAC9EHy4xT46_nuQcLbktTfYvpLXKwukEAwoofix157JViGMFqw%40mail.gmail.com.

[Sándor Filotás]

Hi!  Thanks, I'll write tomorrow because it's too late today, thanks for the answers so far!!

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/1543657757.1620052.1682718368130%40mail.yahoo.com.

[Martin Manning]

Hello, BR Henning, Was there a report of some kind produced from this comparison test? That would be some interesting reading for uTracer users!

[Sándor Filotás]

Hello! Today I'm doing a test with an e88cc, an ecc83, and a 6as7g large dual triode.

First with the utracer 6, with external dc heating, test, characteristics, etc
Then with the Russian l3-3 tester, I let the tube warm up for at least 10 minutes.


If I succeed, I will share my experiences.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/0a4e149e-fcf0-4716-a779-ce6d9d8dd2a3n%40googlegroups.com.

[Sándor Filotás]

Hi!

I did a test, that's all I had time for today.
The tube is a 6080 dual triode, utracer 6, heating external power supply DC.

I measured at the following points:

Ua: 75v

Ug1: -20v

Utracer result:

triode 1 :  Ia  98.55        triode 2: Ia 103.3

S= 6.25                            S= 6.33

L3-3 result:

                Ia 102             Ia  106

            S: 6.7                  6.7

(The tests were carried out on the L3-3 device in such a way that the tube was at anode voltage for about 20 minutes at the set parameters.)

In the next test, I ran a characteristic of the 6080 tube with the utracer, then selected a lower and upper point of a curve, and measured it with both testers.

The two point:

G1  -40v

1: Ua  79.8v           2: Ua 117.6v

Utracer:

1: Ia 18.4

2: Ia 112.2

L3-3:

1: Ia 16

2: Ia 134

The utracer is calibrated, quite accurately

[Hoeberlin]

Hi, i think the Comparison with the L3-3 is quite close.

For matching Tubes, the absolute Values are not as important. Most Important is that the Tubes have a similar Current, no matter, if both Tubes show + or - 5%.

To me, measuring a tube with a 1% Measurement Setup does not make much sense, as Tubes have high Tolerances, and wear out over Time.

I have had a Radio with weak Reception on FM. I checked all Tubes, and all Tubes ( ECC85, ECH81, EF85, EABC80 and EL41 ) were between 20 to 40%. I replaced all Tubes with NOS Tubes, aut that did not make it better.

Best Regards, Henning

Von der neuen AOL-App für iOS gesendet

https://groups.google.com/d/msgid/utracer/9df1bd27-efb8-42fd-ad71-6e952db2dfcfn%40googlegroups.com

.

[Sándor Filotás]

I understand, but what's the point of doing a characteristic on a "cold" tube if you put it in a static tester and put it under anode voltage and there will be a 10-20% difference?  It would be great to pair with Utracer, but it's not easy

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/1035527676.1884916.1682808734658%40mail.yahoo.com.

[Hoeberlin]

Hi, Sandor, i talked about the Comparison Test we made in Germany many Years ago, there was also a L3-3, among with Neuberger 370, and other "real" Testers. The μTracer (3+ Version) was within 3% of the tested Values.

To me, it makes no sense, to ask for more, as i use the μTracer for testing tubes used in Radios from the 50's mainly. And if one Radio drives a EL84 at 48mA, and another Radio with PP Output has a Summarisation of Anode Current of just 40mA ( 20 per Tube ) then you have even more Variations.

And if You compare two Instruments ( Your L3-3 and the μTracer ), then how do you know, that your L3-3 is absolutely precise.

You might want to use a 10kOhm Resistor, ( 1%, 10Watt ), and connect this to Your L3-3 Anode to Cathode, and you schould get a reading of 25mA at 250V. this schould be precise. If not, Your L3-3 is out of Calibration.

You can use the same Resistor with Your μTracer for Comparison.

Hope this helps.

Best Regards,

Henning


Von der neuen AOL-App für iOS gesendet

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/CAC9EHy5FwHoBemDFdHRxkS5nG%2BbpJ2AJZPv1zW864y%2BVh2DvKA%40mail.gmail.com.

[Sándor Filotás]

Szia! Mindkét készülékem pontos, sok időt töltök azzal, hogy a műszereim pontosak legyenek. Enélkül minden mérés értelmetlen lenne. De egy dolog impulzus üzemmódban mérni egy csövet, és másik dolog l3-3 ba anódfeszültség alá helyezni és 10 perc múlva megmérni. Anódfeszültség alatt még nagyon felmelegszik a cső, itt egy nagy triódáról beszélek, aminek a típusát mértem 6080-as, kicsi csöveknél valószínűleg kevésbé jelentős ez a jelenség.
Hiszen ha beteszem a 6080-at az l3-3-ba, feszültség alá helyezem és hallom, hogy hőtágulás miatt hogyan pattan a cső, akkor én is szeretném, ha vakon hinnék az utracer mérését. De teszteltem, az biztos, van elmélet és van gyakorlat.

You received this message because you are subscribed to a topic in the Google Groups "uTracer" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/utracer/6ILutTorLOQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to utracer+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/1483231796.1973813.1682849343989%40mail.yahoo.com.

[Sándor Filotás]

sorry:

Hello! Both of my instruments are accurate, I spend a lot of time making sure my instruments are accurate. Without it, all measurements would be meaningless. But it's one thing to measure a tube in pulse mode, and another thing to put it under an anode voltage of l3-3 ba and measure it after 10 minutes. The tube still gets very hot under anode voltage, here I am talking about a large triode, the type of which I measured is 6080, this phenomenon is probably less significant with small tubes.
After all, if I put the 6080 in the l3-3, apply voltage and hear how the tube snaps due to thermal expansion, then I would also like to blindly believe the utracer's measurement. But I tested it, for sure, there is theory and there is practice.

[Hoeberlin]

Hi, Sandor,

i never made the observation, that a tube makes a mechanical noise due to getting hot. But if, i think it is, because mechanical Parts in the Tube start to move or expand. In that case, yes you are right, you will have different measurements and results.

Best Regards,

Henning

Von der neuen AOL-App für iOS gesendet

https://groups.google.com/d/msgid/utracer/CAC9EHy7NUOpTmvNBZRR04Rt%2B-Xvj4yYebS1F3XoeU2V%3DAmN2%3DQ%40mail.gmail.com

.

[Sándor Filotás]

Hi!

In this case, the question is, if I select a pair of tubes, the characteristic that is "cold" is good, after 10-20 minutes at the anode voltage, how much will it differ from the first "cold" measurement?

I like the utracer because it's a good device, but I'm thinking about something like this, I want to measure it as accurately as possible.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/1430059592.1956942.1682851129472%40mail.yahoo.com.

[Hoeberlin]

Hi, Sandor, if You want testing under real load Condition, precise, and Computer Controlled, then you may consider the "Roetest". This Instrument is available as Kit, and i have seen it life. It's impressive, but it's an investment as well.

For my needs it's just too much of everything, but everyones needs and wishes are different.

You schould find the Website by searching for "Roetest". There you will find all information needed.

BR Henning


Von der neuen AOL-App für iOS gesendet

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/CAC9EHy4FBWiRKWzmGeTRPHN2%2BBTdZmZzHbTAtebDsG1ZcWzYtA%40mail.gmail.com.

[Sándor Filotás]

Hello!

One more note, I did the same test with a small tube, an ECC81. The difference here is very small, this is a low-power tube, what I wrote probably occurs with high-power tubes.
If anyone is interested, I will also describe the measurement results, but here the two testers are really almost the same, which I admit I am happy about.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/619162149.1972102.1682854146810%40mail.yahoo.com.

[Martin Manning]

Here are a couple of trials I ran on my uT3 that show the change in the readings with time. First for a 6L6GC, and then with 2x 4k7 resistors to see the drift in the uTracer itself. The tube heater was started from cold and left on for the entire run. Note it takes some 35 minutes for the uTracer to stabilize, and ~2/3 of the variation after 5 min was accounted for by the instrument.

[Sándor Filotás]

Hello! Thanks for the answer, it all makes sense, the parts reach their operating temperature in about this time. It is true that this is not as big a difference as what I measured, but it is important and useful information!

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/c07b613f-3b5e-48d4-bb0a-d5043cfa5323n%40googlegroups.com.

[Ihor Smal]

Just another 5 cents, the current in the theater of 6080 is 2.5A.. you cannot have the same heating voltage with the Russian (I assume AC heating) and external DC heater for uTracer. With such current there will be unavailable voltage drops everywhere, even on the connecting wires, so as a result you will have different emission (due to different RMS voltages) just because if the heater temperature, the measurement will never be the same and getting measurements so close is already very good outcome:)) 

Also running for absolute accuracy with tubes is very ambitious:) the variation in emission for NOS tubes of the same model already varies 10-20% and then with time, after hundreds of hours or using them, the emission wil drop 20-50% more. So accurate numbers are good for selling matchet tubes, which over time will 100% run into unmatched and will require rebiasing. 

Ihor

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/c07b613f-3b5e-48d4-bb0a-d5043cfa5323n%40googlegroups.com.

[Sándor Filotás]

Hi! Did you think that I didn't think of this for an accurate measurement? The tube went from a laboratory power supply during the utracer measurement, at l3-3 the heating was AC but it was exactly 6.3. I set all voltages with a digital multimeter and checked them during measurement.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/187d298eff8.27fb.16a52cc106b662770bf54050d014535e%40gmail.com.

[Ihor]

I just wanted to share more detailed measurements with uTracer in this case. The tube is NOS Sylvania 6080. The heater voltage (On the tube's pins) is exactly 6.3 DC. The heater voltage just before 10cm long wires that go to the tube pins is 6.70V, so there is drop of 0.4V at the current of 2.5A (which leads to the resistance in the 20cm (total) wire of 0.15Ohm). For such hungry tubes one has to use really thick wires, I typically do not measure such tubes so no problems, and I can always bump the voltage a bit. 

Now back tot he measurements with uTracer.; 6080 was measured at Va=Vs 80 and 125V, with the Vg -25 and -40V. For each of those two conditions 30-50 measurements were taken and plotted (see the plots). Ideally everything should be at the same spots but because of the resolution of ADC for measuring Va and Vs (which influences even charging the boost caps to the set voltage, because PIC always measures of the set point is reached), one can see that the measurements are "all over the place". The error between the measurements is about 3-5% (the stated accuracy for L3-3 in the manual is 1.5-2.5% of the max measured scale). The internal averaging is already turned on on the PIC/uTracers, still the sampling of the the current leads to what one sees. The up-going slope totally makes sense because when uTracer slightly overcharge the caps, the applied and measured voltages will be higher as well as the corresponding currents, so there is correlation which is nicely visible in the plot. 

The classical tube testes which do not sample will never be able to produce such "noisy" measurements just because of the mechanical ammeter construction which with high inertia (compared to the scale of changes), so one will read a kind of average when uTracer in the plots would produce). So just keep in mind that if someone report any measurement from uTracer, it is just a one point from those plots, and if one would redo the measurement 10 times one would get 10 totally different points distributed within the "clouds" of plotted points.

[Sándor Filotás]

Hi!


I measure the same tube every precisely with utracer, the same value the next day, the deviation is very small. What I want to get out of the whole thing is that if you insert this tube into an old tester, eg l3-3, the continuous anode voltage will heat up the tube significantly, in addition to the temperature produced by the heating. That is why the measurement may be different, but the pipe will be under tension in its final location, that measurement is closer to the actual use. As I wrote, the two instruments are very accurate for low-power tubes.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/53ab5f9c-b58c-44ed-9393-12669d2a6a82n%40googlegroups.com.

[theold...@gmail.com]

Interesting discussion, but…..

 

When I was studying we spent a fair bit of time on the subject of error / tolerance calculations. Everything ever produced, however small or big, has a tolerance. Nothing is exact. When you have multiple items, and you fit them together, the final assembly also has a tolerance, but the question now is how much is it, what are the limits of that assembly?

 

As a real simplistic example (and keeping the math simple), take two resistors, each with 20 % tolerance (which was once common in the old tube equipment). If you put these two resistors in series, the total tolerance limit will be 40%! If you put these same resistors in parallel, the resultant tolerance limit is 10%. Older capacitors also had a typical tolerance of 20%, transformers (the better ones) 10%, the tubes? Well, 30% is not uncommon. Put all these components in a circuit, it is a virtual miracle it actually works!

 

As things get smaller, tolerances become more important, and in modern digital equipment a tolerance of 20% would be a killer. It is the mindset we work with, and now the tolerance of a tube becomes a mental issue more than a realistic one. Trying to measure a tube with 5, or even 1% becomes irrelevant with respect to its intended use or functionality. So even though it may be an interesting exercise from an intellectual perspective, it has limited practical value. I absolutely refuse to believe anyone can hear the difference in a sound produces with a tube that is 5, or even 10% below its nominal specification.

 

The title of this thread is static vs. pulsed measurement. So I thought the premise was that (with everything else being equal) that there could be a different result between these two methods.

The uTracer uses a measurement pulse of 1 ms. Now bear with me, from hereon in some of it is my own interpretation of how I understand this to work.

Since this is 1/1000 second, and if viewed as a pulsed DC, there would be a 1 ms gap, followed by the next 1 ms pulse. That would be 2 ms for a full period, which would translate into an equivalent 500Hz AC signal. Very few tubes are used in a pure constant DC current, so the tube normally deals with an AC component superimposed on a DC current. In that context 500Hz is nothing that would stress any tube at all. My point with this being that I do not think the pulsed measurement approach of the uTracer should have any influence on the measured result vs. a static measurement.

 

I would also like to contend that discussion of a few % accuracy is aside from academic interest totally irrelevant in the real tube world.

 

But I stand to be corrected on any or all of this…….

 

Bill v.

 

From: utr...@googlegroups.com <utr...@googlegroups.com> On Behalf Of Sándor Filotás
Sent: Sunday, April 30, 2023 4:08 PM
To: utr...@googlegroups.com
Subject: Re: Re: Re: Re: Re: Re: static vs pulse measurement

 

Hi!


I measure the same tube every precisely with utracer, the same value the next day, the deviation is very small. What I want to get out of the whole thing is that if you insert this tube into an old tester, eg l3-3, the continuous anode voltage will heat up the tube significantly, in addition to the temperature produced by the heating. That is why the measurement may be different, but the pipe will be under tension in its final location, that measurement is closer to the actual use. As I wrote, the two instruments are very accurate for low-power tubes.

 

Ihor <iste...@gmail.com> ezt írta (időpont: 2023. ápr. 30., V, 21:49):

I just wanted to share more detailed measurements with uTracer in this case. The tube is NOS Sylvania 6080. The heater voltage (On the tube's pins) is exactly 6.3 DC. The heater voltage just before 10cm long wires that go to the tube pins is 6.70V, so there is drop of 0.4V at the current of 2.5A (which leads to the resistance in the 20cm (total) wire of 0.15Ohm). For such hungry tubes one has to use really thick wires, I typically do not measure such tubes so no problems, and I can always bump the voltage a bit. 

 

Now back tot he measurements with uTracer.; 6080 was measured at Va=Vs 80 and 125V, with the Vg -25 and -40V. For each of those two conditions 30-50 measurements were taken and plotted (see the plots). Ideally everything should be at the same spots but

.

[Bill van Dijk] <SNIP>

[Sándor Filotás]

I think you didn't interpret what I wrote correctly, which could be because I'm using a translator, I'm Hungarian, and I don't speak English well.  You don't understand what I wanted to say, the discussion is over for me.

--

You received this message because you are subscribed to a topic in the Google Groups "uTracer" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/utracer/6ILutTorLOQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to utracer+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/00d101d97c32%2434f76da0%249ee648e0%24%40gmail.com.

[Félix Menéndez]

Hi guys,

Nice discussion (at least for heating up our neurons!)

Let us not forget that valves’ manufacturers themselves state that specs and curves are “typical” or “average”… so, as long as the reading of a given valve is close enough to manufacturers’ figures (say +-20%) that valve (IMHO) is “typical”…

Regards,

Félix

EC2ALV

--

You received this message because you are subscribed to the Google Groups "uTracer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to utracer+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/00d101d97c32%2434f76da0%249ee648e0%24%40gmail.com.

[Ihor Smal]

Indeed I think there are multiple interpretations for this discussion but the difference between the static and pulsed ways of measuring (indeed the frequency of pulse measurement is nothing compared to what tubes can handle in terms of transient processes) can be due to the fact that in one case (static) anode/plate is hot and in the pulse measurement it is cold, and that’s it. It is true that those conditions will be different. Theoretically there should be difference only due to secondary emission from the plate.

Ihor 

On 1 May 2023, at 15:38, theold...@gmail.com wrote:



--
You received this message because you are subscribed to a topic in the Google Groups "uTracer" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/utracer/6ILutTorLOQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to utracer+u...@googlegroups.com.

[Sándor Filotás]

This is exactly what I'm talking about, I stopped to measure the tube because I heard this on a Hungarian forum.  I thought it would either support or disprove their claim.

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/9B9512A1-200E-4E2F-BCD7-44C18AE5D2E6%40gmail.com.

[theold...@gmail.com]

Agreed.

 

My apologies to Sandor Filotas if I came across as a wet towel, that was definitely not my intent.

 

What I was trying to achieve was some of my own clarity about the objective of this discussion. What I am gathering so far, and again I stand to be corrected, is that there could be a difference in the two measurement methods because a static current of some magnitude will increase the temperature ( it will be pretty hot regardless) of the plate, and create a secondary space charge around the plate which would possibly impede the desired electron flow. That sounds like a reasonable possibility, since severe overcurrent can actually cause “red plating”. The emphasis there is on severe, and I would assume the effect of (over) heating of the plate should be included in manufacturer’s posted charts since I assume they used static current sources for their tests.

 

From my own personal experience using both the 3+ and V6 uTracer, when I compare my own generated charts with those published by the manufacturer, I see (visually) little to no difference. Where I sometimes do see a noticeable difference is between similar marked tubes from different manufacturers. So in terms of this discussion it seems to me that unless we also include the manufacturer in the equation, we’re not comparing apples to apples. In order to eliminate that variable we would have to go back to an earlier proposed initiative (which to the best of my knowledge never actually started) where one tube would be sent around to different testers.

 

I still stand with my earlier conclusion that this is all very interesting from an academic perspective, but of limited practical value. Conversely, having said that, we would not have known all this had it not been brought up and discussed!

 

Bill v.

--

You received this message because you are subscribed to the Google Groups "uTracer" group.

To unsubscribe from this group and stop receiving emails from it, send an email to utracer+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/9B9512A1-200E-4E2F-BCD7-44C18AE5D2E6%40gmail.com.

[Sándor Filotás]

Hi!  When measuring statically, the cathode is hotter, doesn't this cause more emission?  If you raise the heating voltage of a tube, it also raises the anode current, so it would be logical.  I'm not saying that utracer measures inaccurately, but that high-performance tubes can behave differently during a static test.  But there is theory and there is practice, it is necessary to measure both ways and share experiences.  I did it... If someone else does it, I'm just happy for it

To view this discussion on the web visit https://groups.google.com/d/msgid/utracer/014401d97c3e%243a2d6390%24ae882ab0%24%40gmail.com.

[Hoeberlin]

Hi, Sandor,

i made tests some years ago with the μTracer, with nominal Heater Voltage, with +10% and also with -10% Heater Voltage.

Some Tubes do not show a big Variation, some others of the same Type do show a very big Variation.

I tested Tubes from different Manufacturers and also new and used and also weak tubes. Many used tubes show very high Variations, where new Tubes (NOS) show only a very little Variation.

BR Henning

Von der neuen AOL-App für iOS gesendet

https://groups.google.com/d/msgid/utracer/CAC9EHy6Y%2BV43qnbYv7ntwk4dpfWSYNWYWCYAJHhk5TZ9JvNWog%40mail.gmail.com

.

[Martin Manning]

US made Hickok tube testers had a "Life Test" function which reduced heater voltage by 10%. A "good" result was a reduction in gm of less than 20%.