Tung-Sol 5881 - Hard, Medium, Soft?

[TomCarlos]

Friends...

I went looking for a set of Tung-Sol 5881 and came across the posting on Musician's Friend. I noticed they have 3 versions of this tube: Hard, Med, and Soft. Are these just a matter of breakup or are we talking needing to change bias or something along those lines?

Thanks in advance...

Tom

[Jazz P Bass]

I do not have a clue what they are describing.

My bet is that the tubes, if placed on an output voltage/gm bell curve, would go from low to higher output.
Tubes are like this off the production line.
Which is why 'blindly' biasing any tube at a specific current draw, is plain stupid.

[TomCarlos]

Thanks John...

I went to the Tung Sol web site... and found a response (Click here).

"Soft tubes reach saturation and break-up quicker. These tubes have lower plate current (Ip) and transconductance (Gm) matching numbers. They are preferred by blues guitarists for the break-up and sustain they provide."

"Hard tubes have the highest amount of clean headroom before break-up and distortion. These tubes have high plate current (Ip) and transconductance (Gm) matching numbers. They take longer to reach saturation and are preferred by jazz, country, and bass players. Hard tubes are also used by guitarists who rely primarily on effects pedals to generate distortion."

"Medium tubes fall in between soft tubes and hard tubes. They exhibit good headroom, but will break up and distort when pushed. These tubes have plate current (Ip) and transconductance (Gm) matching numbers in the middle of the range. Classic rock guitarists and players who play a wide variety of styles generally use medium rated tubes. "

Seems like in a fixed bias amp, you would adjust the bias and all three would be about the same. In a cathode biased amp, you could end up with some of these tubes running hot. Is that a fair assumption?

[g1]

So Mr.Pittman is working for Tung-sol now?

[Leo_Gnardo]

Seems like in a fixed bias amp, you would adjust the bias and all three would be about the same. In a cathode biased amp, you could end up with some of these tubes running hot. Is that a fair assumption?

'Zackly right Tom, that's been the game since GT started. Ignore your bias control, try all our tubes and settle on the ones you like. As if we have time and budget to do that. Hm, now where's my little screwdriver...

[Chuck H]

The "sperts" will tell you that biasing with a scope is TONAL DEATH!!! Nonsense. As variable as todays tubes are, how else is anyone supposed to know what's going on when accurate data is rarely provided?!? That isn't to say that the typical proscribed method (max clean volume, bias out the crossover notch) is the best way to go. It's almost always NOT. But you do need to know how much power a tube is capable of and how it relates to the bias point. Most Rusky data sheets are regurgitations of old RCA and Phillips specs that are arbitrary to the actual tubes in the box. You need to wing it, give or take a few watts. A scope is the easiest way to grasp what's the tubes you have are really doing.

[pdf64]

The big problem with crossover notch biasing is that because the max signal transfer function at reasonable dissipation often isn't particularly linear (but can be forced to be more linear by reducing the bias voltage and running crazy hot), it means that the bias setting is dependent on test signal level and the degree of negative feedback.
So the static dissipation may be rather lower on amps with plenty of negative feedback where the bias has been set using a medium a signal level, but rather too high if set using a 'max before clipping' test signal level / amp with low / no negative feedback.

The above idea gelled for me when using nickb's interactive data sheets, which show the transfer function for the operating conditions selected http://bmamps.com/Tech_tds.html

[J M Fahey]

I'd have to buy 6 sets of tubes, 2 pairs each grade, to be able to measure them and try to find a correlation between marketing name and actual characteristics.

Read above as: "it won't happen" .

so just guessing, probably hard/mid/soft just means how high the bias voltage must be set for a given (undisclosed, of course) idle current.

Say, for +450V plate, 400V screen, 30mA they require, say, -52/-50/-49 V respectively , which also will mean that different groups will not be reasonably matched , definitely bad choice for common cathode resistor biasing.

And probably (although it's also a guess,given the lack of real data) the ones with higher emission will be the ones which require more negative bias and will provide highest current at saturation (say, 280mA peak) while lower emission ones will need less negative bias voltage to persuade their low performance cathodes to pass current, any current, from idle conditions to saturation (which might be, say, 200mA peak).

In a way you might select them based on how pre-aged or pre-worn you want your tubes to be from the beginning.

I also think that along time, all "hard" tubes will become "mid" ones, then "soft" ones.

Whoever found a way to sell all tubes, no matter quality, is a Marketing Genius.

To boost, the way the ad is worded, there's a reverse snob appeal attached: metalheads obviously will choose the ones labelled "distort more" ... and will prefer the weakest

Funny thing is that they are being sold on (assumed) tonal differences ... and any NFB in the power amp will iron flat them

They are inventing nothing new, local grocers hawk just arrived , still very green and harsh tasting bananas as "great for frying" or "long life ones, still good in a week or 2" , sell mid ones normally, and at last sell (very) ripe ones as "extra sweet and creamy, great for shakes/smoothies" ... none gets unsold

As a side note, notice they are being sold as "equivalent to 6V6" ... that alone should be a red flag.

Also some users report that 9substituting these for factory 6V6 makes filament fuses blow in some cases".
Another reports "blown on the first day"

So let buyer beware.

[TomCarlos]

Thanks everyone... I appreciate the replies.

[g1]

Are you planning on using those for a 6L6GC application? They are spec'd to original 5881 specs, and may not be suitable for many 6L6 type amps.
Some of the more observant tube vendors are rightly warning their customers:
"There are some important things to know about the reissue Tungsol 5881 tube before you decide to use it. The Tungsol 5881 is a 23 watt tube and is rated for a maximum plate voltage of 400Vdc." (www.thetubestore.com - Tung-Sol 5881 Audio Tubes).
So it needs to be biased as a 23W tube.

As far as the bias method discussion, I agree with Chuck, looking at the scope is invaluable. I use the scope and idle current for biasing, not one or the other. But I think even scope alone is more effective than setting a grid voltage like the "good ole days" .
Yes, using hard rules for crossover notch biasing can give errors, but you learn to compensate for different models, circuits, etc. if you use it along with idle current measurements.
For fun, I do crossover notch first, then check current. It usually comes out pretty close.

[TomCarlos]

Thx Greg...

I need a set of 5881s for my Traynor YCV80- that is the stock tube. I borrowed a set of Sovteks from a friend when I rebuilt that amp. Since I will keep the amp and start using it, I need to return a set of tubes to my bud. So we went looking online for tubes and at is how we spotted the different versions of the Tung-Sol 5881.

I learned alot on this thread... Good input from everyone.

[g1]

Yes, the Sovtek 5881 and the tung-sol 5881 are quite different. The sovtek 5881 is like a 6L6, and can be used as such. They were often labelled 5881/6L6. The tung-sol is like a vintage 5881 with lower specs to match, and is not really a sub for 6L6.
In my mind it's sovtek's fault for turning the 5881 designation into something it is not. Now we're predisposed to think that a 6L6 and a 5881 are the same. But there are a lot of 6L6 circuits you would not subject a vintage 5881 to, and the new tung-sols are like that.
The YCV40 runs them right around 400V. So they may be ok, but bias them as a 23W tube and expect a little less power out of them.

[pdf64]

'The YCV40 runs them right around 400V. So they may be ok, but bias them as a 23W tube and expect a little less power out of them'
I can't see why, given that the operating conditions are appropriate, 5881 / 6L6WGB would put out less power than 6L6GC?
As all the characteristics, apart from the limiting values, are the same.

I have the same confusion regarding JJs EL844 cf EL84 (ie why should it lower the audio output power) though I haven't checked that the charts are the same.

[J M Fahey]

I can't see why, given that the operating conditions are appropriate, 5881 / 6L6WGB would put out less power than 6L6GC?

As all the characteristics, apart from the limiting values, are the same.

There's the rub, you can not be sure those are actual characteristics.

I suspiciously see datasheets published which **exactly** match some original RCA/Sylvania/GE/Siemens/Telefunken/whatever datasheet, which is at least improbable ; making me think that they are just plain "drawn"or cut-and-paste.

Also because Russians in the 40's often copied "Occidental" tubes but they kept widely using them much more than us, and developed their own .

Which sometimes are close enough to some of "our" types that vendors and distributors label them so (paint is cheap) but that doesn't make them any closer.

Besides 6L6 proper , I seem to remember that they also had a tube type of their own which was more of a glorified 6V6, to which it was an upgrade, (and which I suspect is the case here) ; the task is easier because they use a large, 6L6 sized bottle and of course, same pinout.

Besides printing "5881" on them (a type which I guess URSS never actually made) they also may print some suitable datasheets.
FWIW they are admitting something that on their pages

In the case of "EL84" something similar happened, but in this case theirs was a ruggedized Mil type.

I have the same confusion regarding JJs EL844 cf EL84 (ie why should it lower the audio output power) though I haven't checked that the charts are the same.

I suspect something similar here, both on the actual tube and "published" charts.

Only actual tracing of both, samples of at least 10 each , 50 or 100 preferred to iron out tube to tube variations will prove or disprove.

FWIW a colleague of mine who is series producing tube amps got so fed up with it, that he runs in his shop an automated test station.

I visit him and at any time of day or night (meaning 24/7) there is a rack with some 20 or 30 (don't remember exactly but next time will sneak some picture) octals, may be 6L6 or EL34 cooking, a lot of tiny relays are heard softly clicking in the background while a dedicated PC varies plate voltage from. say, 50 to 500V in 5V steps, and repeats again varying grid voltage from cutoff (whatever's proper) to 0V , all the time calculating plate dissipation so as not to redplate them.

A brute force approach but which lets him safely ship a tube to a customer in a faraway province , who will plug it instead of the old one without need to rebias.

So far that very anal approach has worked very well and now he's the largest tube amp maker in Argentina.

[g1]

I can't see why, given that the operating conditions are appropriate, 5881 / 6L6WGB would put out less power than 6L6GC?
As all the characteristics, apart from the limiting values, are the same.

First, this is not the 5881/6L6WGB which is my main point! That was a small bottle 6L6 for military applications was it not? And the russian version we are familiar with is actually some russian number tube that is close to 6L6 and will withstand 6L6 applications.
But the tung-sol 5881 Tom linked in his first post is a reproduction of the vintage lower spec 5881 tube.

But I think your concern is whether they will actually deliver less power when used in regular 6L6 applications (or EL844 delivering less power in EL84 applications).
So are you saying they will just work the same way and over-dissipate which would just bring on red-plating sooner?
I haven't looked at that with any kind of measurement but I think it would depend on the actual circuit. In lower power amps I guess it should not be noticeable?
Is all the talk about EL844's reducing output power just chatter? I had a pair of EL844's but didn't use them as the bias was so far out of whack I promised myself never to touch them again. But now I'm curious to do this experiment!