EF86 NOS tubes and varying plate currents - tested as new

I've been looking up NOS EF86 tubes for my G7 project and found that I know very little about tubes.

How can it be that these folks claim that a rating of 2.3mAh at the oputput gets a rating of good, and an output of 3.0mAh is considered NEW, but...

There are tested NOS Telefunkens with output rating of 5.8mAh, and even as high as 6.5mAh!

How can 3.0 be considered new while there are tubes outputting over 100% more?!

In a G7 circuit, I assume the output will be directly proportional to the maximum rated output of the tube, regardless of the plat voltage. Or am I incorrect in thinking like this and it works more like a voltage regulator where the output is 100% designed by the circuit, and it will not vary between 2 known good tubes with different max outputs?

I am wondering how much better a G7 will sound with a NOS Telefunken EF86 rated at 6.0mAh output current, compared to a NOS Telefunken tube with a tested 3.0mAh max output current.

Should I spend the extra bux for the higher rated tube, even though 3.0mAh is considered good/new? What will the difference be soundwise? I assume you will be driving the transformer harder and thus hear more influence from it, correct?

Can I use 2 different rated tubes (but close) and still get a close match between 2 mics?

I'd buy a bunch of these tubes and test this myself, but Fuk they are expensive!!!

i heard little difference between my NOS telefunken ef86 and a jj electronics ef806 in my G7.i dont remember in what thread(in mic meta) it was but others agreed that the g7 circuitry was very forgiving regardless of the type of tube used.i'd get something cheaper and use the money where it can be better utilized (bang for your buck wize) jmho

could be me, but it's more maketing.
A lot of the testers i've seen have a broad , usually green colored range,
that is considered the "good" range.

now, it probably varies by tester, and the "good" range was
BROAD, as i mentioned.
it's also probably why they usually list the meter make and type.
if you plug it into "x" tester, and it tests in the "good" range,
well, then i guess it's "good".
might be the low end of the "good" range, but i suppose you can't call them a liar............

there also apparently must be manufacturers info
(i've never seen it, but #'s are often mentioned)
that tells what the manufacturer considered as testing as new.
you often see references of "110% of new", and i've seen it often enough
that there must be data out there on the original "specs" as designated by the manufacturer.

buying used/nos tubes is such a crapshoot.
i buy used/tested on occasion, but i try not to pay much so if they are crap
i'm not out much.
(i don't have a tester to "verify" any sellers claims)

The tester that was used on teh ones I was looking at is a Funke W19s. It seems that 3.0mAh is GOOD on this tester, and 6.0mAh is NEW.

I suppose if I can get a pair of tubes rated close to 6 it would be worth it. Sets of two seem to go for about $60-$80 or so, and at tubestore new premium EF86 are about that after shipping.

I know there's probably a lot of good tubes for sale now, but with the Chinese factories pumping them out now QC has certainly had to drop in order to make them at a competitive price.

I's like to get ahold of a couple NOS Telefunken, Mullard or Amperex tubes. I have also heard that these NOS will virtually last forever.

The screen voltage is super important on those tubes.

The tube tester probably puts the screen at some weird voltage, maybe not, but the best way to test is in the circuit.

Noise is the main problem with those tubes.
You find a quiet one and 6 months later, the popcorn starts popping.

So it is the luck of the draw.
If you run the plate ate say, 250 volts B +, try running the =screen way down at 100 volts.

If it sounds good, leave it.
This will keep the popcorn popper away for maybe two years if you are lucky.

The trade off for all this hassle is, in my opinion, a nicer sounding amp then a 12AX7 or 12AY7.

Not that a 12AY7 based old RCA ciruit with cool transformers is going to sound bad, just different.

If you own both a 12AX7 and an EF86 based preamp, then you have the tube scene covered pretty good.

Category 5 wrote:
"The tester that was used on teh ones I was looking at is a Funke W19s. It seems that 3.0mAh is GOOD on this tester, and 6.0mAh is NEW. "

If this is the case, I wouln't call NOS the one that only measure GOOD (3mA).
My understanding is that NOS means NEW Old Stock. A tube that only measure 50% of a new one is either:
-half dead (may be usable but certainly not NEW).. or
-new, and even "NIB", but suffering from vacuum leakage to the point of degrading its characteristics...

We're not talking VF14 or WE 300B here! IMHO a "less than perfect" EF86 is not worth the shipping fees...

That said, for microphones applications, plate current is not the most important parameter. Grid leakage, noise, microphonic sensivity are, and that's not the kind of parameters that common eBay sellers advertize....


Gwaggin390 wrote:
"there also apparently must be manufacturers info
(i've never seen it, but #'s are often mentioned)
that tells what the manufacturer considered as testing as new. "

Have a look a this Telefunken datasheet for the EF806S:
http://www.mif.pg.gda.pl/homepages/frank/sheets/128/e/EF806S.pdf

Ia 3.2+/- 0.6mA

Between 2.6 and 3.8 mA is what Telefunken consideered "new"
At the end of the first page there are "End of life" values;
If the plate current drops to 2mA (less than 40%) the tube is officially "dead"...

Axel

> output of 3.0mAh is considered NEW

Where are you getting "mAh"???

That's a battery rating, not a tube thing.

Two "identical" tubes from the same factory WILL vary +/-20% one to the next. The Type Acceptance specifications usually allow +/-40%.

And BTW: if you are just measuring some notion of "maximum current", it is as CJ says VERY dependent on G2 voltage, in most cases NOT near your actual operating point, and will very often rise a bit in the first 100 hours of use as more of the thorium comes to the surface and wets the cathode. It won't decline much until the tube is dead. Anybody who claims to sort New from Light-used by an emission test is fooling himself.

> the output is 100% designed by the circuit

Since tube current can vary a LOT from one to the next, a key part of the designer's job is controlling circuit operation with different tubes stuck in. Basic cathode bias tends to cut variation in half. G2 resistor also stabilizes operating point. These simple steps will put 99% of random same-type not-sick tubes within 10% of the nominal current. Gain changes roughly as square-root of current, so gain changes like 0.5dB from one tube to another.

That's good-enough for most any practical purpose.

(When it is not, like you have 50 telephone repeaters amps in cascade and can't tolerate +/-25dB gain variations, there's more tricks in the designer's bag.)

The Gyraf G7's EF86 is run at about 1.0mA.

As CJ says... the current is unimportant. In most EF86 applications, Hiss and Microphonics are why you have a pile of discarded tubes in the corner. There is no good test for this. MIL-Specs had some hiss and thump tests to weed-out the real stinkers, but for fine audio it is best to have a large lot of maybe-OK tubes to try, than to have just a few overpriced high-Current tubes.

This questions is probably a bit nieve. But are there ways to AOT the components to get a "not so great tube" to behave better in the ck?. This would be hand if one was planning on having a stereo pair. Maybe there are too many variables and dynamism to force tubes to "behave".

> But are there ways to AOT....

AOT???

> the components to get a "not so great tube" to behave better in the ck? ... Maybe there are too many variables and dynamism to force tubes to "behave".

I thought I covered that point. Maybe I did't.

Anyway: the G7 mike amp does not especially need EF86. It was probably picked because most EF86 are, despite rumors, fairly low-hiss and not too microphonic. If you really can't find a clean solid EF86, re-wire the socket for some other tube type.

I have very limited experience with tubes so I understand if the question seems naive to the experts. My greatest experience with tubes is with my Mesa Boogie Mk2 C+ amp. I bought a set of 4 Sovtek preamp tubes when the amp wasn't working too well and after installing them the amp sounded superb (really superb). I was talked into (by Mesa) buying the Mesa branded AX7s thinking they would sound even better, but despite their cost and my really trying to like them they sounded flat, dull, and noth nearly as sweet as the sovteks. Heck, they probably were Sovteks for all I know, but from a different lot, batch, day or whatever.

I also switched the Sovtek 6922 in my NTK mic for a NOS telefunken (new in package) and despite claims that the tube has very little to do with how that mic sounds, the TFK "did" successfully get rid of the belchy, over compressed sounding high end that my mic always had. It is in fact a much better mic now.

I was hoping that a NOS EF86 would help me avoid the "something doesn't sound right" phase with my G78s, and hopefully help me match the mics betrter. After all, G7 uses the tube for gain while I think eth NTK only uses it as an impedance converter. After all, it should be even more profound with regard to the sound right?

A search surely reveals that the G7 circuit is forgiving of lesser tubes, but that a lot of people had serious issues with bad tubes. Since a lot of 4 NOS TFKs is going for about $25 per tube my thinking was that I couldn't really do much better buying from Tube Depot (who gets $40-$100 for good new tubes). I can get 4 tubes and pick the best two.

It seems the sellers from Germany, with the highest feedback (i.e. 100%) are offering sets that rate at around 5.5-6.5 mA plate current so if I decide to get some TFKs that's who I'll buy from.

I assume that the tubes that are lower than 5.5 are used. The German NOS sellers seem to all sell tubes around 6.0mA.

As far as leakage , noise, microphonics, etc. I guess I'll have to hope I get lucky. But wouldn't the same hold true for tubes from Tubestore or Tube Depot?

After reading a little about tubes and how they are made it's no wonder tubes of the past are reputed to be better sounding and longer lasting than tubes made today. They know most tubes are going to sit in a Fender amp in the corner of some teenagers room seeing less than an hour of use per week. QC has to be wildly different these days.

Thanks for all the replies thus far. I am slowly learning a lot about tubes, why we use them, why we don't and what they do/don't-do well. They are fascinating to say the least! I am really lloking forward to doing some more projects with tubes!

One can only afford QC equivalent to the old days when it's the 'state of the art', and there's a cold war on with lots of defense issues at stake (and the sort of money that rolls with those contracts).

I've never experienced anything I'd consider reliable noise or microphonics testing with preamp tubes. PRR already addressed current concepts and tolerances. CJ is right about the noise issues. Not true of all types; I have 5879's (RCA's earlier similar concept tube which fell from favor; why?) that have been running fine for 8-9 years, no increase in noise or change in current. In fact, the 'american' G7 preamp would have used 5879's in the early versions at very least, and might have switched to EF86's when they came out, or maybe not.

I see a lot of tubes that test lowest on a basic tester, but clearly sound the best, have the lowest noise, etc. Basic testers can only be considered a 'first weed-out' of obvious problems. RCA sold tube equivalents with different part #s that passed higher standards for noise and microphonics. This was much easier to do when they were pouring off the assembly line constantly, and going into every conceivable type of electronics product. Once you get into the early-mid 1970's you can assume the standards were greatly relaxed and might not equal production of the previous decade(s). And it goes downhill until now; the scale isn't there for today's production to compete. It may be possible to do better, but the budget numbers aren't there.

The only real answer is to become a hoarder-of-tubes-you-need, and select from your private stash. PITA; but I've watched some types I'm lucky to have in my stock spiral out of control in price.

I personally avoid items or projects specifying EF86's, as I don't feel it's reasonable to assume a quality tube can be had without a ton of work and cash outlay, given the competition for existing quality supply.

Just found this link about the Funke 19 ability to test tubes...:
http://reviews.ebay.com/Tested-tubes-Doubtful-measurements-w-Funke-W19-tester_W0QQugidZ10000000007120179

I know that eBay links won't last for ever so here's a copy/paste:

" FUNKE W19 etc., a overrated piece of junk?

quite possible and it´s fair to say because it does not make too much sense to check a tube's cathode capabilities under random conditions pre-determined by the tube tester maker - but that's exactly what these testers do! The Funke 16, 18, 19 do not provide any grid bias other than zero (grid & cathode strapped together) or -2V approx. to check for connection between grid & base prongs. To get the resulting plate current back within limits, Funke also reduces plate and / or screen voltages. This definitely is too far from real operation - and it does not allow measuring anything else like mutual conductance, hum, etc. In addition, the valves often do not become hot enough on these testers, so there often tubes appearing good on these testers while inserted in the circuit they should operate in, they develop inter-electrode shorts or show floating grid bias due to residual gas.

Mainly originates of the bad Funke design, the heater voltage is not stabilized and always too low or high, DC voltages are superimposed by AC because of too small charging and filter caps and so on. Also important reason of wrong measurements is Funke has changed the test cards again and again! This is why for the same type of tube tester and the same tube exist DIFFERENT test cards. Especially dual triodes like 12AX7, ECC83 that tests good on the „Funke“ in fact are often not good. On the Funke a good ECC83 starts at 1.9 mA, but in fact you can be sure that this tube is junk. In comparisons with other gm testers my view is I would´nt trust any ECC83 that is below 2,5 mA on the Funke = 65 to 70 % on a mutual conductance tester.

Generally I would not trust any tester that is an Emission-only tester. Emission alone does not tell the full story of a tube's condition. Each tester manufacturer established their own test criteria. To get a meaningful result about the "quality" of a tube it should be tested for Gm, (TRUE) mutual conductance, given in mmho, or dynamic mutual conductance in % (there is a difference and it can be a big one for some and not as critical for others, but this is another technical discussion)"

Thanks for the link! I was thinking about getting a tube tester and now i know what to stay away from. That's why it pays to ask around here. For sure someone's got your back. :grin:

I've been lurking here for a couple years and wish to say "thanks" to all the regulars who help so many with their generosity. I have learned so much from from PRR NYD CJ and so many others. Thank you all for sharing! (I hope some of you see this ...)

Anyway, I actually have something to add here. I have been focused on EF86 (and 5879) based preamps in the form of the ampex 600 and 601 preamps for almost 2 years. I have had about 20 of these come through my hands so far with all sorts of tubes: telefunken, amperex bugle boy, philips, valvo, GE, etc., etc. The 600's and the early 601's used the 5879, but in late 1957 they changed the circuit to the 6267 (EF86) without changing the model number.

The EF86 was reported to have lower noise and lower microphonics than the 5879. Ampex thought it was important enough to make the change in that rude fashion. I almost always install new JJ EF806s's in my rebuilds.

I have only seen a couple bad EF86's. Most are surprisingly consistent especially if you match the brand. However, the JJ EF806s sounds better to me than any of the vintage tubes I've had. I've never had a Telefunken EF806s though. They sell for over $200. The JJ is said to be a very good remake of the Telefunken and it is only about $25.

I never invested in a tube tester because I read similar things as mentioned above. I have a couple cheap ones I inherited from my dad, but I test tubes by ear by swapping tubes around in a stereo unit. Sometimes I'll run distortion or freq response sweeps, but I find the ears are pretty reliable.

I hope this helps and is accurate,
Here's my main preamp page:
http://www.fixthatmix.com/ampex601.php

Thank Jimbo, cool info!
Ampex was a quality company while it lasted.

I was told to buy the Amperex Bugle Boy with Orange writing, as theses had the best chance of being noise free right out of the box.

They are very pricey also.

Nice work Jimbo :grin:

Wicked.

:wink:

Thanks CJ and Kid Squid!
I have never had an orange bugle boy, just white ones. One of the 2 bad tubes I have found was a white bugle boy. The other was a Holland GE 6267. A real bummer because they were both from pairs.

I really like working on these old Ampex electronics. (except for repairing meters) When I first started I messed with Bogen, RCA, Heathkit and others but when I got my first Ampex I was hooked. The build quality makes them worthwhile to restore and fun to work on.

If anyone wants some 5879's?? emrr? I have some I would share. They only go for about $5 on Ebay though, so most people probably have what they want.

> Generally I would not trust any tester that is an Emission-only tester.

A tube-tester saved my butt. I had a smoked amp, and a pair of 7027 tubes. I replaced the toasted iron. I ran an ohmeter around the 7027 pins, no shorts. But then I put the tubes in the tester. When the heater was hot, there was a massive short in one tube. That's why the amp burned!!! And I wudda burned it again if not for the tester. (Maybe I wudda noticed the over-current in time... and maybe I wouldn't.)

And when a tester says a tube is DEAD, you have saved a lot of time in routine Radio-TV Repair. The set is sick. Some part is sick. Lots of parts in there! But bad tube is good bet, and tubes are socketed. You (or a helper) can mindlessly pop tubes through the tester, maybe get lucky. "Good" or "half good" is not real interesting (even a "half good" tube may work fine in a low-power stage). They may really be good or bad for the application. But a "Dead" tube means re-re-check your tester settings (including maker's updates and errata), then get a new tube and try that.

Do NOT emission-test tubes recklessly. It is a strain on the cathode. It is not as bad as testing car batteries by throwing a wrench across the lugs and eyeballing the size of the spark, and most of the cathode thorium will come back after some hours of normal (non-tester) use, but still tube-abuse.

There ARE good tube testers. They cost a fortune. They sold in tiny numbers to a few labs who really had to know tube properties. When they turn-up, some hardcore geek usually snaps them up; I suspect today most are traded-around a small community of geeks.

Beyond that.... test in circuit in actual application, or a very similar application.

For guitarists, I suggest keeping a Fender Champ around. If you pop four 12AX7s in the hole, and get plate voltages of 170V, 166V, 233V, 172V.... guess which one may be "sick". (Or just be tarnished pins.) 12AT7 and 12AU7 too: the voltages will be different, but all of the same type will be very-near each other, odd-balls are sick. Then too, you can listen to a Champ. Excess noise is hiss, and the volume knob tells you if it is section A or B. You can play through it: if three 12AX7s play well and one plays weak, it IS weak. 12AU7 in Champ will always be "weak", but a few trials calibrate your ear to what is "normal weak" for 12AU7, and odd-balls can be cast aside.

Same applies to other audio amplifiers. Build or adapt a mule with test-points and a standard signal source and monitor. Try different tubes. Good ones are all the same; bad ones will stand-out. And maybe you can hear tubes with "extra sparkle".

To add to PRR's post. I often build the gain stage circuit fragments both fet and tube sometimes on a piece of plywood or in a nice box. I also build a power supply. I install the tubes/fets in sockets and test the tube/fet gain stage and check the operating points. Tubes used in condenser microphones are often run at under 1ma plate current and lower voltages. Building the circuit in a test unit is a good way to check tubes for a microphone.

OK, the first thing you must do is finalize your schematic.

This way, you know how many taps and impedances and aux windings for meters and all that good stuff.

How many transformers are we talking about here?

Most people want to rip those out, heaven forbid.