Calling all valve GURUS

[Wakefield Turntables]

Hi Andy, what tests do you want to do on the valves?

PS i wouldn't consider myself a valve guru..., just someone with an opinion

The idea was just to purchase a cheap, reliable piece of kit which would provide accurate assessment of the valves health. Ultimately I just something that could accurate grade my valves for future projects, then hopefully enabling me to use the best (healthiest) valves for the job. I understand that only so many tests can be performed upon a valve. I intended to then research each test and it's relevancy. I perform diagnostics on people every day to assess health so I figure I might have a fighting chance of understanding the test values and their relevancy to the valves health ( well that's the logic of a clinician)!!!! Hope you can see my head logic.

[da2222]

The idea was just to purchase a cheap, reliable piece of kit which would provide accurate assessment of the valves health. Ultimately I just something that could accurate grade my valves for future projects, then hopefully enabling me to use the best (healthiest) valves for the job. I understand that only so many tests can be performed upon a valve. I intended to then research each test and it's relevancy. I perform diagnostics on people every day to assess health so I figure I might have a fighting chance of understanding the test values and their relevancy to the valves health ( well that's the logic of a clinician)!!!! Hope you can see my head logic.

Very difficult to do unfortunately...You can find various emissions only types but they only tell part of the story. The AVO CT 160 is very good indeed and will offer mutual conductance and other crucial tests, but yes, it's expensive. For peace of mind it is always good to test the tube in situ in 'real' apparatus after standard testing.

[Mixc]

The idea was just to purchase a cheap, reliable piece of kit which would provide accurate assessment of the valves health. Ultimately I just something that could accurate grade my valves for future projects, then hopefully enabling me to use the best (healthiest) valves for the job. I understand that only so many tests can be performed upon a valve. I intended to then research each test and it's relevancy. I perform diagnostics on people every day to assess health so I figure I might have a fighting chance of understanding the test values and their relevancy to the valves health ( well that's the logic of a clinician)!!!! Hope you can see my head logic.

Some one Pls correct me if im wrong ..

The way i see it..
Each valve gives a specific output for a given input.
If a valve is tested at *?* input it gives an output *?* and can be matched to another valve with similar results.
If a valve operates an input Between *1* and *10* The best way to test at 1 , 2 , 3, etc this gives an output curve.
I think these testers that can do this are expensive.
A power valve (KT88) has a single output.
Whereas a dual triode (ECC83) has 2 outputs.
Dependent on the use or circuit, matching of the valve to itself or to another valve is important.
Some circuits put both left and right channel through one valve and some have a valve for the left and a valve for the right.
My power amp works on push / pull so each half of the wave form is handled by one 6550.

There is a lot to testing and matching valves, so the type of tester required would be down to what tests needed to be performed.

Correctly matched valves are expensive because of the time involved in testing.
I read somewhere that Audio Research have a reject rate of 98% of their purchase valves.

Mick

[Mixc]

Very difficult to do unfortunately...You can find various emissions only types but they only tell part of the story. The AVO CT 160 is very good indeed and will offer mutual conductance and other crucial tests, but yes, it's expensive. For peace of mind it is always good to test the tube in situ in 'real' apparatus after standard testing.

How would you test in situ ?

Measure the output or measure at various points in the circuit?

[Wakefield Turntables]

Some one Pls correct me if im wrong ..

The way i see it..
Each valve gives a specific output for a given input.
If a valve is tested at *?* input it gives an output *?* and can be matched to another valve with similar results.
If a valve operates an input Between *1* and *10* The best way to test at 1 , 2 , 3, etc this gives an output curve.
I think these testers that can do this are expensive.
A power valve (KT88) has a single output.
Whereas a dual triode (ECC83) has 2 outputs.
Dependent on the use or circuit, matching of the valve to itself or to another valve is important.
Some circuits put both left and right channel through one valve and some have a valve for the left and a valve for the right.
My power amp works on push / pull so each half of the wave form is handled by one 6550.

There is a lot to testing and matching valves, so the type of tester required would be down to what tests needed to be performed.

Correctly matched valves are expensive because of the time involved in testing.
I read somewhere that Audio Research have a reject rate of 98% of their purchase valves.

Mick

Interesting post Mick, I'm sure your perfectly correct, just goes there's more to this valve testing thingy than meets the eye.

[Jazid]

Hi Andy,

As a clinician you appreciate the relevance of separate and complementary tests to ascertain the DUT's (PUT's?) condition. The trouble with a 'one size fits all' approach is the shortfall of usable information. For example do you need to know the emmissions of the valve? Then for starters you really need to know what current it conducts at a specified anode and grid voltage, ideally at several different combinations to construct a chart to compare to those that the manufacturer published. You may also need to know the transconductance in mA/V, again at the specified point as a minimum standard and ideally charts to see if it meets spec. Impossible with a good, ok, bad test. The trouble is valves are inherently variable and often one will have slightly high emmissions but conversely exhibit slightly low transconductance or vice versa, and this can be normal and within manufacturer's tolerance. However without knowing the measurements and their respective conditions one cannot judge whether one valve is equivalent to another which may also be within tolerance but have rather different measured specs. When matching valves this becomes a significant issue as one presumes all valves sold new should be within spec, but not necessarily adequately matched for the purposes of the circuit they might end up in - one reason for the introduction of SQ and uprated types, and selection and marking of 'instrumentation grade' valves by manufacturers of lab equipment.

As an example for power valves in manual bias circuits you normally want close matching emmissions at a particular grid voltage. For push pull self bias circuits you may however prefer valves with tightly matched transconductance curves. In pre-amps a double triode may need matching sections or it may not depending on what the two sections are doing. If they are amplifying the same stage in two channels they will need to be matched, if they are amplifying different stages then balance between the equivalent sections of different valves may be more important than between the two sections of each valve.

Similarly it is good to know that the heater:cathode insulation is high in a valve, and in some circuits imperative that it meets valve design spec, but in others it is not so important; If you are using venerable or uncommon valves then it is prudent to use those which are strong but which due to age have less than brilliant insulation in circuits where the heater:cathode voltage is low, rather than discard them just because they won't work well in a circuit where the design h:c voltage is high.

Likewise testing a valve's emmissions with a high voltage pulse will 'work' and this will be how the Orange tester operates I suspect since it seems to use a wallwart SMPS. It may even excite a flashover if the valve is on the edge. But running a valve at design limit voltage and current (or beyond) for a length of time is a more exacting test, and if there is a potential failure mode eg. hotspots on the anode then in such a test this is more likely to become evident. The ability of the valve to sustain emmissions can also be tested - an important consideration if the plate or cathode are slightly poisoned or there is some slight gas to 'burn off' from eg. valves which haven't been used for a long time.

So I'd say if you are looking to play with your thermionic friends in due course then a decent tester is a worthwhile investment. There is a new-ish tester but I cannot recall the name (maybe a bit DIY but not so much as the pi-tracer). It looked good when I nosed around, think a question in DIY section will elicit the answer. There is also the expensive but impressive Amplitrex 1000 as well if you need to get yourself a Christmas pressy, or a properly fettled Mk4 AVO, Neuberger, or a Kalibr L3-3
*edit* or indeed AVO CT160 as Drew mentions, or VCM163...

[da2222]

How would you test in situ ?

Measure the output or measure at various points in the circuit?

No, I simply mean after testing subject the the tube to 'real' conditions ie an amp or intended device...