New tube tester reveals I have some crap power tubes :-(

[BmWr75]

What criteria were you using for determining decently matched pairs when testing with your 'old' tube tester ? Ways of determining if tubes are matched seem to vary across the internet - for me I would be looking at bias voltage, and anode current and then Gm.

Looking at test results in post #22 and ignoring the emission% column - which doesn't mean a great deal - I would think that tubes 2 & 3 and 1 & 4 would be a couple of reasonably close matched pairs. This is borne out when looking at the far right column which shows Gm in uMhos. Umohs was in common use on the more up-market US made testers of the 50s and 60s, 3800 uMhos is equivalent to a Gm of 3.8 mA/Volt.

As far as your old testers agreeing with the Amplitrex it's no surprise they don't, the Amplitrex uses Gm and the old testers use uMhos, in addition I'm guessing the old testers use lower than data book voltages during the testing process, which gives lower than data book results for Gm.
There is a direct correlation between the measured Gm (which is an indicator of the 'goodness' of a tube) and the applied voltages, tube tester makers could and did test tubes at lower than data book voltages - while it is a trade-off between performance and instrument cost, it is a perfectly valid way of testing tubes - the downside is that you get results that don't agree with data book values (or modern tube testers) but instead you need the tube tester manual/databook for the tube tester makers correctly calculated values to refer to.

Gm in uMhos is the only number I got from my old testers, so that is all that could be used for tube matching.

See Post #24, picture 1. It shows this new tester outputs Gm in mA/V and also calculates a Gm% based on the measured value compared to the specification for the tube.

 

[ceylon]

I have long been confused by what readings on a tube constitutes a matched pair. Tubes have many different readings. Which ones are most important? I built a pair of monoblocs 40 years ago using 6BG6 tubes as finals. I run them fixed bias with a pot to adjust the grid bias. I set the bias for about 40 ma.at idle, usually winds up about -35 volts with 425 volts on the plates. The cathodes are grounded through a 5 watt, 100 ohm pot. Center to ground, each side to cathode through a break contact 1/4 inch plug. I can plug in an ammeter and adjust the 100 ohm cathode pot to achieve equal current flow through each tube. If this cannot be achieved, the tubes are not "matched". This puts the output transformer at a neutral magnetic state at zero signal. I realize that many other tube differences can and will result in a distorted output to some degree, but which ones are most important, which ones being equal would qualify that pair as "matched"? Paul.

i am not the authority, but, as far as i know, matched pair are two tubes testing within 5-7% on anode current and 7-10% on transconductance.

in the US and canada, power tubes are matched by plate current and signal tubes by transconductance. ive noticed that most of the rest of the world for the most part matches by plate current or both. matched by both is best. just about any legit tube shop does both.

matched by gain aka "gain matched" has been getting more popular as the maxi preamp (newer popular tube tester) measures gain.

 

[BmWr75]

It's an interesting review. I am not sure if this is the specific model in question.

The reviewer didn't like this tester for two reasons. He thought it was slow and didn't like the computer interface software.

Slow is a relative expectation. The tester comes from the factory with a 60 second warm up period for each tube. I've left it that way. The time required to test a tube is less than 60 seconds am guessing......because the warm up time seems much longer to me. Guess I should time how long after warm up it takes to test a tube.

I've not hooked up the tester to a computer. Don't know if I ever will. So the software issue the reviewer has is irrelevant for my current use. Do wish it came with or would connect two a small credit card receipt type printer so results could be printed. For now I just take a photo of the results screen or write the results on a pad of paper.

 

[CDawg]

in short, matched pair are two tubes testing within 5-7% on anode current and 7-10% on transconductance.

in the US and canada, power tubes are matched by plate current and signal tubes by transconductance.the rest of the world for the most part matches by plate current.

matched by gain aka "gain matched" has been getting more popular as the maxi preamp (newer popular tube tester) measures gain.

I aim for 5% on plate current and transconductance for power tube matching. I wouldn't lose sleep up to 10%. I always heard 10% was the industry standard so I just picked 5%.

Someone schooled me recently on why plate current is better than transconductance for small signal tubes and it made a lot of sense to me. I like having as much information as I can with multiple testers and I think that's the best route if possible.

All the measurements from the Maxipreamp II are useful but I love the noise test the most.

 

[CDawg]

The reviewer didn't like this tester for two reasons. He thought it was slow and didn't like the computer interface software.

Slow is a relative expectation. The tester comes from the factory with a 60 second warm up period for each tube. I've left it that way. The time required to test a tube is less than 60 seconds am guessing......because the warm up time seems much longer to me. Guess I should time how long after warm up it takes to test a tube.

I've not hooked up the tester to a computer. Don't know if I ever will. So the software issue the reviewer has is irrelevant for my current use. Do wish it came with or would connect two a small credit card receipt type printer so results could be printed. For now I just take a photo of the results screen or write the results on a pad of paper.

Amplitrex is an amazing tube tester. I have limited use with one but I loved it. Takes about two minutes to warm up and accurately test a tube on a Maxipreamp II or Maximatcher while I write down the results. I don't think the time to test for the Amplitrex is unreasonable considering the information it provides when you compare it to others. No experience with the software myself but I'm going to imagine it's fairly simple once you get accustomed to it.

 

[familyguy]

I'm with ceylon - post #42 - you really should know the bias voltage, anode current and Gm or transconductance to fully match a pair of tubes - however if your old testers only have transconductance - umohs or Gm doesn't matter - provided the same bias voltage is used it should get you reasonably close as the Gm is derived from the anode current and grid signal voltage - Gm = mA of anode current per volt of grid signal. Note this is grid signal not bias voltage.

 

[ceylon]

I aim for 5% on plate current and transconductance for power tube matching. I wouldn't lose sleep up to 10%. I always heard 10% was the industry standard so I just picked 5%.

Someone schooled me recently on why plate current is better than transconductance for small signal tubes and it made a lot of sense to me. I like having as much information as I can with multiple testers and I think that's the best route if possible.

All the measurements from the Maxipreamp II are useful but I love the noise test the most.

man i like that noise test

 

[ceylon]

I'm with ceylon - post #42 - you really should know the bias voltage,

id hope that the better testers test @ spec voltage. i believe some do. i think i see a lot of folks testing signal tubes on a soviet tester measuring plate current @ 250V.

 

[prww]

Does this tester use AC, or DC filament supply for the noise test? If it uses DC, the test results could be different than your application. Paul.

 

[BmWr75]

Does this tester use AC, or DC filament supply for the noise test? If it uses DC, the test results could be different than your application. Paul.

Don't know, not mentioned in the user manual.

 

[BmWr75]

Just timed two 6BQ5 tests. After 1 minute warm up period, the test takes 19 seconds each.

So, the fellow complaining about long test times in the video above appears to have some unreal expectations.

 

[63pontiacgp]

I've had tubes that test so-so on the tester and perform wonderfully in equipment.

Mike

 

[Bill Ferris]

I`ve had my AT-1000 for close to ten yrs. and for my tube testing needs & desires, I really have no complaints in my stand alone use of it.
It has not failed to be able to test any of my friends or my tubes to our satisfaction and confidence, certainly more than my purchased used B&K 707 that I used from 1975 until 1988 when I gave it to my TV technician friend after when I became a Groove Tubes dealer, as I no longer needed to test tubes, just substitute another new tube into whatever audio equipment that I was servicing from my generous quantity start up dealer stock that GT sold me of tested and select matched USA mfd. at that time output tubes.

I also haven`t had any real reason to connect my AT-1000 to one of my laptops to it or felt the need to purchase and use a label printer, so no comments pro or con on that.

 

[Tom Bavis]

In case anyone is wondering, 1 mMho = 1 mA/V - precisely the same. Hickocks do not (can not) measure at a data sheet operating point. But the roll chart settings were usually "fudged" to produce a meter reading close to the rating, though that is not possible for every tube type. The assumption (usually good) is that if a tube is off at one point on its characteristic curve, it'll be off elsewhere. Likewise, an emission tester measures emission at one point, with the assumption that the entire curve drops with use, so a weak tube will measure low no matter what the test operating point is. Peak emission might be the most important number for a power tube, determining how much output power it can deliver. It's almost never specified, though, except for sweep tubes.

I have a pair of 6550s that measure about 1/3 the rated value (on a Hickock) and 4 more that measure close to rated - all deliver 60 watts in a push-pull amp. It's possible that the "weak" ones need more grid drive and would have more distortion, but I haven't measured it. They do bias differently, but I don't even think about grid bias voltage, only plate current.