ODD TUBES, what to look for?

[TomWaits]

I 'm reading the thread "Favourite Tubes" and after reading all of your posts I want to know what rare but odd tubes often get over looked?

I have access to a few tube surplus sellers and often they only pick out the classic valve part numbers and leave behind tonnes of candy to harvest?

We all know 6L6 , 6550, 300B, KT88, EL84, on and on...

BUT what tubes often go unnoticed that have great value including rectifier types. I want to know so I can become a better shopper and you guys seem to know the odd gems.

I want to compile a shopping list of cheapies and oddities to hunt for and I would love to see your suggestions.

Thanks,

Shawn.

 

[ErikdeBest]

Interesting question! I recommend that you give a look at Geek's forum. There are lots of curves traced for odd tubes, triode strapped pentodes etc. Some really good things among them!!

http://geek.scorpiorising.ca/GeeK_ZonE/index.php

Both Geek and Serpentine are members of diyaudio, Serpentine under the nickname 4E4

 

[Ouroboros]

I'm a fan of the 829B, using two of them (with both sections paralleled) for a push-pull amp. An OPT with a primary Z of about 4.5k seems to work well.

Beware that there's a lot of late-manufacture RCA ones about (with the big red RCA logo) that appear to be quality rejects.

 

[Tweeker]

Sweep tubes.

 

[EC8010]

Beware that there's a lot of late-manufacture RCA ones about (with the big red RCA logo) that appear to be quality rejects.

Odd you should mention that. I bought a batch of fifteen and when I tested them on my valve tester, four were duds for one reason or another.

 

[andyjevans]

One simple answer is loctals. Get some loctal bases and save mucho money. A lot of the standard tubes were made in loctal equivalents. Note that most of those below came in 14*** versions which were 12v tubes, and are even cheaper. None of these seem to move on ebay - they're a goldmine which is constantly overlooked and prices are static. Be sure to widen the locking base with a pair of long nose pliers - just pull the two ears apart. these tubes are meant to lock and some do a good job of coming apart in your hand when you try and prise them out (had two CV391s fall to pieces - thin glass).

Approximately speaking:
7N7 = 6SN7
7AF7=6SN7 round plate
7A4=6J5
7C5= 6V6
KT81=KT61
CV391=807
3A/167=437a
7F7=5751/6SL7
7B4= half a 5751/6SL7
7Z7=EZ80/81

There's even some DHTs
1LE3 mu 14.5
3B7 double triode

 

[Tubes4e4]

Hi Erik,

I recommend that you give a look at Geek's forum. There are lots of curves traced for odd tubes, triode strapped pentodes etc. Some really good things among them!!

http://geek.scorpiorising.ca/GeeK_ZonE/index.php

Both Geek and Serpentine are members of diyaudio, Serpentine under the nickname 4E4

Actually it is "Tubes4e4" 🙂

Here´s the link to the index page of traced curves: Some tracings

Tom

 

[Tubelab_com]

In my case I already have a few zillion oddball tubes. I look for tubes that have physical characteristics (the guts, not the envelope) similar to what I need. If I am looking for an output tube, I look for a tube with a large plate structure. If I need a driver tube, I look for a medium sized triode, or dual triode (or triode wired pentode). If I need an input, or preamp tube, I look for a small triode or dual triode. If I find a tube, that I have a few of, that meets the physical criteria, I look it up. Observation of the electrical characteristics can eliminate a few possible choices. It is likely that I have one or more of the target tubes that are physically damaged. I will often take them apart to look at the internal structure. After doing this for a while you learn what works, and what doesn't. Control grids wound with uneven spacing causes distortion.

If the tube has a chance of success, I hook one up and test it in the indended application. I use a breadboard system with all power supplies adjustable, to find the sweet spot for the particular tube. If one tube shows promise, then I test a bunch of the same tube to see how repeatable they are. If I like the particular tube, I will torture test a few of the ugly ones to find out how reliable they are. It is not uncommon to find (especially power tubes) some samples that self destruct at the spec limit, while a different brand of the same tube will live happily at twice the published ratings.

After you have done this, you will have identified a few tubes that were previously unknown. It is also possible to find a tube that has been previously shunned by the audiophile community that works well under different conditions than the audiophiles use.

In this audio world it would appear that a few people state the "facts" and they are blindly accepted by the majority. This happens a lot on one of the other audio forums. For instance, it is an accepted fact that the 12AU7 is not a good audio tube. This I haven't tested yet, but since I have about 1000 of them, I will. I have also heard the same about the 12AT7. I was searching for a good driver tube, when I discovered a large box full of 12AT7's. I wired one into a simple breadboarded SE amplifier using the common resistor loaded circuit, biased at the usual 2 - 3 mA. I tested the distortion spectra, and listened to the resulting (mono) amp. It would seem that the world was right, a thin sounding amplifier with a large amount of third and fifth harmonics. If I had given up there I could have said that the audio world was right. After a lot of experimentation, I discovered that this tube likes more current (5 to 10 mA) and a CCS load. Under these conditions a 12AT7 works really well.

There are some tubes that just won't work for HiFi use, no matter how long you play with them. There are others that have a wide variation from tube to tube. The wide variation can be OK for a one off amplifier, but will kill you if you plan to build a few amps.

I have learned to avoid tubes designed for "gain controlled RF amplifier" or "AGC controlled IF amplifier" use. These are often called remote cutoff pentodes or variable Mu triodes. These usually have too much distortion for the signal levels found in most audio gear. They may be useful in phono stages.

Every few months Antique Electronics Supply sends out a flyer that lists a page of "bargain tubes". As soon as I get it I look up all of the under $2 tubes to see if any are useful. I found 3 in the latest flyer, and ordered 80 tubes. As if I don't have enough tubes already.

 

[EC8010]

I get it I look up all of the under $2 tubes to see if any are useful. I found 3 in the latest flyer, and ordered 80 tubes. As if I don't have enough tubes already.

Ah, but in percentage terms you've scarcely increased your valve stock at all, so that's alright.

 

[kstagger]

EF40 = rimlock socket version of the EF86
ECC40 = rimlock socket dual triode
EL38 - nice power pentode with a plate cap
1633 - 25V filament version of the 6SN7 - used for aircraft radios?
1631 - 12.6V filament version of the 6L6
6C6 - plate cap pentode - RCA's version of the 310?

5X4G = same as 5U4G, different pinout
5Z4/GZ30 - good for low current power supplies

and the usual 12SN7, 12J5, 12SL7, 12SJ7, 12SX7 tubes 😎

 

[aerius]

Loctals were already mentioned, but another cheap trick is to look for the same tube in a different voltage, a 12SN7 or 12SL7 instead of the 6 volt ones for instance. There's also the 6A3, and the 6AU7 instead of 12AU7.

 

[Wavebourn]

Compactrons.
Tubes and sockets are available on ePay.

 

[Miles Prower]

In this audio world it would appear that a few people state the "facts" and they are blindly accepted by the majority. This happens a lot on one of the other audio forums. For instance, it is an accepted fact that the 12AU7 is not a good audio tube. This I haven't tested yet, but since I have about 1000 of them, I will. I have also heard the same about the 12AT7. I was searching for a good driver tube, when I discovered a large box full of 12AT7's. I wired one into a simple breadboarded SE amplifier using the common resistor loaded circuit, biased at the usual 2 - 3 mA. I tested the distortion spectra, and listened to the resulting (mono) amp. It would seem that the world was right, a thin sounding amplifier with a large amount of third and fifth harmonics. If I had given up there I could have said that the audio world was right. After a lot of experimentation, I discovered that this tube likes more current (5 to 10 mA) and a CCS load. Under these conditions a 12AT7 works really well.

I've found it helpful to take a good LQQK at the specs. The suggested plate currents for the 12AT7 are a good deal higher than what a lot of folks seem to be running with. That's where I'd certainly start looking for a good loadline. Of course, the main problem here will be less Av since the Rp's will have to be smaller with reasonable Vpp's. Still, you can always use an active load if you need high gains, or make up for it with a preceeding stage.

As for 12AU7s, these appear to work best as SRPP's and Mu Stages, or possibly as LTPs with a tail CCS.

 

[TomWaits]

Expanded shopping list

Great feedback. I have a much bigger shopping list now. 🙂

Thanks,

Shawn.

 

[wa2ise]

There exist tubes like 8BQ5, 11BQ5, 8CG7 and so on. The same as their 6 volt cousins except for the oddball heater voltage. BTW, if you have a 12.6VAC supply avaliable, you can connect an 8V heater to it in series with a SS rectifier diode, no filter cap. This yields about the same amount of heater power that 8V AC would give directly. If all you have is 6.3VAC, run that thru a bridge rectifier and use a filter cap to get 8VDC to run the heater.

 

[Wavebourn]

There exist tubes like 8BQ5, 11BQ5, 8CG7 and so on. The same as their 6 volt cousins except for the oddball heater voltage. BTW, if you have a 12.6VAC supply avaliable, you can connect an 8V heater to it in series with a SS rectifier diode.

If to rectify 6.3V by Shottky diodes you'll get 8V DC. I use such trick to power compactron filaments. 8JZ8 is the example (one triode + power pentode in triode connection to drive power FET follower)

 

[Miles Prower]

BTW, if you have a 12.6VAC supply avaliable, you can connect an 8V heater to it in series with a SS rectifier diode, no filter cap. This yields about the same amount of heater power that 8V AC would give directly.

Mega-bad idea there. That will also yield lots of harmonics that'll add uncorrelated noise to the signal, and really do the sonics no good at all. If you're going to use DC for heating, it needs to be well filtered to get rid of those objectionable harmonics. Otherwise, DC heater power sounds worse than using AC. Better to drop the excess AC with a series resistor.