DIY Tube tester by Grant Willis

[rayandkerry]

Hi, all.

I'm going to build the tube tester described at this page: https://valveheaven.com/2015/03/an-inexpensive-easy-to-build-diy-valvetube-tester/

Here is a link to a PDF that has the schematics: https://valveheaven.com/wp-content/uploads/2015/03/An-inexpensive-easy-to-build-diy-valve-tester.pdf

It seems like a pretty cool design given it should be able to measure Gm (indirectly) and test for shorts and gas leaks.  In the schematic he shows the PT wired in reverse so he can get a 6.3v feed for the heaters being produced using the 30v and 24v windings, along with 15v and 30v feeds that are run through voltage multiplier circuits for screen and plate voltages up to 250v.  Neat!

But I have a number of questions.  Some on building, some on usage.

First off, I could not find the same transformer as in the design (i.e. a multitap PT with 30v, 24v and 15v taps).  And I Hammond has discontinued their 30v center tapped transformers.  So instead, I got a 28v center tapped PT that is overspec for current rating (it's rated for 3A, the original schematic has a PT rated for 2A at 30v).  I'm hoping since the PT won't be fully loaded it will put me close enough to the original 30v.  I wrote the article's author and he seems to think it's OK.  He's a nice guy, but it's difficult to converse much via the webpage's comments section...hence I am asking folks around here.

I also got a 12.6v center tapped PT for producing either 6.3v or 12.6v feed for the heaters.

Attached is a modified version of the original schematic.  I was hoping some other eyes could check it.

Basically, I've changed the PTs (as described above), added a switch for selecting 6.3v/12.6v heaters, and added two more multimeters to show the plate voltage and heater voltage (in addition to the grid voltage and plate current ones from the original design).  I figure these additional meters will be helpful in ensuring I've got all the switches set right for a given tube.

So...

Question 1: Does the schematic look right given the different PTs, the dual voltage heater feed and the extra 2 DMMs?

Question 2: The original schematic shows a diode and capacitor being used on the heater feed.  I did not include in my schematic.  Is that necessary?  If so, why?  All the heaters I've wired up were simply wired straight from the PT...no diodes or capacitors.

Question 3: How, after taking readings, do I determine the life left in the tube?  I've done a fair bit of searching on how to test tubes and always get stuff relating to using a specific tester (i.e. do these specific steps on the Hickok TV-3).  And all the tube testers I've seen have a meter that reads from good to bad, sometimes with a % reading.  There's none I found that provide a raw Gm reading and then explain how to translate that to a % of useful life for a given tube. 

Does anyone know how to determine the life of the tube once I've calculated the plate current (with the screen and plate voltage set to what I see on the tube's datasheet) and transconductance (by varying the grid voltage by 1v and getting the change in plate current)? 

For example, page 8 in the PDF with the schematic shows a 6v6 tube with -12.5v on the grid and 55.7ma of plate current. I’m assuming the tester is configured 250v on both the plate and screen and that the author dialed in -12.v on the grid because the 6v6 datasheet shows all those values in the single tube class A section of the datasheet.

But the datasheet says zero signal current is 45ma and the max signal current is 47ma. So how do I interpret getting 55.7ma which is above the datasheet values?  Does that mean the tube would read 124% on a commercial tester (because 55.7 / 45 = 1.24)?  Honestly, I have no idea on how to usefully interpret that.  I wrote Grant, but he just said "Hi Ray, the data that you need to refer to is the standing or non driven currents in the valve data. When the valve is driven the plate current will increase".

I also asked him the following about understanding the transconductance reading in terms of tube life:
"Also, let’s say I calculate a Gm of X by varying the grid voltage by 1 v and getting the difference between the current readings. What does it mean? In the 6v6 example, the data sheet says transconductance in class A single tube operation with 250v plate and screen current with -12.5v on the grid is 4100 umhos. Let’s say I calculate 3500 umhos? What would that mean? Does that mean I have a tube with 85% of it’s useable life (because 3500 uhmos is 85% of the 4100 uhmos the datasheet listed?"

But he didn't answer directly.  Can anyone else help me understand how to translate the readings into useful information about tube strength and life?  I think there must be some general rules on tube testing out there somewhere I am just not finding in my internet searches. 

Any thoughts folks can offer would be greatly appreciated!  Thanks!

[PRR]

Plate mA and Plate Volts meters are in parallel. One of them must be wrong.

Not clear why D11 is here.

[shooter]

Not clear why D11 is here

agree
appears it would be conducting if plate vdc is < PS vdc by a volt or 2 

[sluckey]

You need a crystal ball, not a tube tester. 

[rayandkerry]

Plate mA and Plate Volts meters are in parallel. One of them must be wrong.

Not clear why D11 is here.

The original schematic only had the plate mA.  I added the Plate Volts meter myself, hoping to be able to simultaneously read  both.  The basic idea is to use low cost multimeters as a display.  I could just use 1 multimeter for both Plate mA and Plate Volts, but thought it'd be  cool to have both at the same time.  It's only another $6 for a DMM. ;-)

As to why the diode is sitting in front of the HV multimeter, the author says the following: "The 200mA plate current meter has a 1N5404 power diode wired across it to protect the meter from accidental shorts or overload."

You need a crystal ball, not a tube tester. 

LOL!

[PRR]

> thought it'd be  cool to have both at the same time.  It's only another $6 for a DMM. ;-)

The way it is connected will NOT work.

But hey, that's cool! More meters is more impressive! Post pictures!

[rayandkerry]

The way it is connected will NOT work.

Do you think it's possible to take simultaneous voltage and current readings with two DMMs?  How would I need to connect the DMMs? In series?

Also, any ideas on how to interpret the plate current and Gm readings?

[PRR]

> Do you think it's possible to take simultaneous voltage and current readings with two DMMs?

Of course!

Maybe my picture didn't turn out.

[Sonny ReVerb]

I always wanted to build the RAT tube tester. I have most of the parts, but I have a Sencore that works pretty well (an eBay buy I had to rebuild/recap). I'm not sure how it compares to the one you are considering, but maybe you can use some of his design. Anybody build a RAT?

[rayandkerry]

Maybe my picture didn't turn out.

Oh I see it now.  Sorry I must have read your original post on my phone and missed the pic.  Thanks!

[rayandkerry]

I always wanted to build the RAT tube tester. I have most of the parts, but I have a Sencore that works pretty well (an eBay buy I had to rebuild/recap). I'm not sure how it compares to the one you are considering, but maybe you can use some of his design. Anybody build a RAT?

Wow.  Very cool.  I'll need to have a think about that.  Thanks!

[HotBluePlates]

... page 8 in the PDF with the schematic shows a 6v6 tube with -12.5v on the grid and 55.7ma of plate current. I’m assuming the tester is configured 250v on both the plate and screen and that the author dialed in -12.v on the grid because the 6v6 datasheet shows all those values in the single tube class A section of the datasheet.

But the datasheet says zero signal current is 45ma and the max signal current is 47ma. So how do I interpret getting 55.7ma which is above the datasheet values?  ...  I wrote Grant, but he just said "Hi Ray, the data that you need to refer to is the standing or non driven currents in the valve data. When the valve is driven the plate current will increase". ...

Assuming the photo was taken with plate & screen voltages at 250vdc, you would interpret the result as a tube that passes higher-than-average plate current.

I have a Russian tube tester, and the card for setting up a test for the 6π6C ("6P6S" or what is generally sold as a Russian 6V6) shows the same info as page 3 of this 6π6C data sheet: plate & screen voltage of 250v, bias of -12.5v, with plate current falling 33mA≤45≤57mA.  What does that mean?  While the American 6V6 data sheet has a single "bogey" (average) value shown of 45mA, the Russian sheet says that 45mA is the average value and new tubes land +/-12mA of that average value.

They mention the same kind of thing on the transconductance, where the average value is 4.1mA/volt (same as 4100 micromhos), with an acceptable range of +/-1.1mA/volt.  Said differently, the range from 3000-5200 micromhos would be acceptable for a new tube, with the average being 4100 micromhos.

I haven't gone through all the data to find norms, but the above reflects a tolerance of almost 27% above/below the published bogey value.  In this post, I point out that Hickok's manual for one of their testers says the left edge of the green "Good" portion of the English scale is supposed to correspond to 20% below bogey for small-signal tubes, and 35% below bogey for power tubes.  I could have sworn the manual for my Hickok 533A had a figure more like 40% below bogey for tubes before the needle strayed into the "?" or "Replace" portions of the scale.

...
Question 3: How, after taking readings, do I determine the life left in the tube?  ...  There's none I found that provide a raw Gm reading and then explain how to translate that to a % of useful life for a given tube. 

Does anyone know how to determine the life of the tube once I've calculated the plate current (with the screen and plate voltage set to what I see on the tube's datasheet) and transconductance (by varying the grid voltage by 1v and getting the change in plate current)? ...

The same post I mentioned above tells how to perform Hickok's Life Test.

Beyond that, no one and no test can tell you the amount of life remaining in a tube.  That point was made pretty clear in Getting the Most Out of Vacuum Tubes (written by an engineer at a tube manufacturer).  What's more, tubes that measure solidly in the "Replace" section on a Hickok will still pass sound if you plug them into your amp.  And you may decide they sound just fine.  Which is why all the tube manuals, and all the books, and all the tube tester manuals wind up saying, "A tube has reach the end of its life when it fails to give satisfactory service."

Yes, it sounds like a cop-out, but everyone has to decide for themselves how dead is "too dead to keep using."  I've got a few 12AX7s (acquired in a large lot that were mostly wonderful) that measure super-dead on my tester that I may just try popping into the old Vibro Champ to see if they still work.