DCAudioDIY.com

hi, guys, wanted some input as to if I can run the old metal 6L6 tubes in my Onix SP3, I have posted this on AA, and have got some answers, but nothing really definitive. The SP3 is made for 5881, although obviously you can run other 6L6 variants as long as the parameters are not exceeded, and is rated at 38 watts/channel. I measure from ground 340v to the plate, which is close to the max, which is 360v, the bigger problem is the bias measurement to get a sense of the current and thus the plate dissipation, here pin 1 & 8 are tied with a large 33 ohm resistor connected to them and pin 6, which is grounded. A second wire runs from the pin 1/8 to the bias trimpot, a little 1/4 watt trimpot that has around 37 ohms for a drop of 1.15v (this of course varies as you adjust the trimpot, but not a whole lot...), the other end of the trimpot at ground. So it seems as though the bias adjustment is in parallel with the large fixed cathode resistor, rather than having a single large bias pot for adjustment. So, if I measure the voltage drop across the bias resistor, would I get the current flow thru the tube? I am guessing that I am getting only a part because of the biasing arrangement. BTW, the schematic I found online seems to be a different model, I have a MKII version, thanks!

I wouldn't. The metal 6L6s have a lower dissipation rating than the 6L6GT, which is what the 5881 is based on. Also, old metal tubes are much more likely to be gassy than old glass tubes.

Roscoe

_________________
I can explain it to you, but I can’t understand it for you.

1. Measure the resistance between pins 1 and 6 (Ground), with the amp off, of course. This IS the total resistance of the parallel combination.

2. Measure the voltage between pins 1 and 6 (Ground), with the amp idling.

3. Calculate the current.

BTW, 33 - 37 ohms, or the parallel combination of those is way too low to properly bias a 6L6, using standard cathode bias. Typical values would be somewhere around 250 ohms (with one resistor for both tubes) or 500 ohms (if each tube has its own resistor), which is what your description sounds like. Maybe this is an attempt to mix some (magic) combination of fixed bias and cathode bias?

If these cathode resistors are mostly used for measuring the idling current, then there is some negative voltage source supplying additional bias voltage to the grids. If so, then the trimpot in parallel makes current measurement pretty imprecise. Is there an instruction manual that tells how to bias the output tubes?

BTW, what's the attraction to the old metal 6L6? There are lots of good choices in the 6L6GB/6L6WGB/5881 family. The old SOVTEK 5881 was quite good.....I haven't compared it to the new crop of reissues, but I thought they held up well against TUNG-SOL 5881.

thanks, guys, Roscoe, I went thru and checked for gas on my metal 6L6 tubes and culled all of those that did not pass the test. FerdinadII, I measure the resistance between pin 6 and 1/8, it is 33.5 ohms, about what the cathode resistor is, I checked pin 5 and it is -32.32 v, so it would appear that the bias is set via the grid rather than the cathode, or a combination of both, you do use a probe across a trimpot to set it to 1.15v for proper bias, the voltage drop across pin 6 and 1/8 is 1.15 or whatever the test points are saying, so that would give me 35 ma across the plates.

The attraction to using the old metal 6L6 is that I have about 3 quads of them just sitting here as well as that some have noted that they have very good sonic qualities for that family, so I figured I should give them a try. I can also generally get them pretty cheap, around $2-3 each, nobody wants them! Getting newer reissues are close to $100 for a matched quad, and NOS TS 5881 and about $200/quad, when you can find them. 6PS3s run around $50 a quad plus shipping from USSR...

yea that biasing arrangement does seem pretty imprecise, I note that biasing seems to jump around a bit at times...

Well, here's my take on whether it'll melt into a puddle.....

Idle conditions:
Measured Cathode current = 34mA
Estimated Screen current = 4-5mA
Estimated Plate current = 30mA
Max Possible Voltage across Plate = 340V(measured)
Max Plate dissipation = VI = 340V x .030A = 10W

Full Power conditions:
Estimated Max Plate current = 50mA
Max Possible Voltage across Plate = 340V(measured)
Max Plate dissipation = VI = 340V x .050A = 17W

Max Allowable Plate dissipation = 19W

So, IMHO......
You can get away with it, if you don't crank the volume.
You may get away with it, even if you do crank the volume, but it's not a sure thing.

thanks, ferdinandII, could be worth a try as long as it doesn't destroy my amp!

There should probably be a few extra disclaimers in the above post.
My conscience, the lawyer just whispered in my ear... LOL

The bias of -32V seems low(high...). Most 6L6 designs are running low 20s. The measured idle cathode current matches up with the low bias conditions at the grid, so that's reassuring.

Maybe the overall design has a good amount of headroom. I don't know that much about the rest of the design, though. It's still possible that the designer found a way to run the output tubes up to full throttle. This would put the plate currents up to ~65mA, which would bring the plate dissipation up to ~22W; too high.... It's not likely, just because the plate voltage is fairly low, but you never know. VTL used to typically run low idle currents, but ran high voltages to get the power ratings, and I've seen more than one VTL amp with char marks.

The other thing is, with a metal tube, you can't visually monitor the plates, to see if they are turning cherry red, as a result of too much power being dissipated. Do you have a safe way of measuring the current of one or more tubes during actual operation?

yes, that was my concern as well, can't see the plates! Don't really have a way to monitor the current unless I wire in a meter or outputs to hookup a meter...

BTW, I just checked and the metal 6L6 tubes have Pin #1 connected to the shell. On the glass tubes it's NC (Not Connected).

http://frank.pocnet.net/sheets/049/6/6L6.pdf

Now, this means that the shells will be at the same voltage as your cathodes. Not too high, but it's still disconcerting to have part of the circuit floating around on the top of the case. What happens if two shells are shorted together....? Aren't the sockets on this amp recessed below the chassis? If they were sitting up high, top mounted, you might say no problem. But you don't want them touching the chassis, for sure.

Why did the manufacturer tie pins #1 and #8 together? I see no reason to do that, since they could have easily soldered the resistor and trim pot directly to just Pin #8, right? If anything, pin #1 should be grounded. Is there some physical reason why they did this, or did they have a secret hatred for old metal 6L6 tubes....?