• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Question about RLa-a for *LW6 in triode connection?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm thinking of building a 36LW6 PP triode amp but I'm having a hard time figuring out a good RLa-a. I'm thinking about 300V B+ and fixed bias at about 50ma. I use 12AV5GA at this level into 3k RLa-a and get a nice 20W, 30W with still less than 10% THD. Since the 36LW6 has more power though I was thinking of boosting the B+ but worried about the max screen rating. I assume the screen stopper will save the screen from excessive current though. Maybe a B+ of 380V?
 
George (Tubelab) has mentioned that the long term stability (reliability) of the _LW6 tubes is not so good in triode mode, subject to runaway. So keeping the screen V down for triode would seem a good idea. Power wise the _LW6 tubes should be able to handle down to 2200 Ohms P-P if you want the power.

However, figuring the Triode Rp of the tubes as Mu/gm = 10^6 x 3.7/12000 = 308 Ohms, you would want at least 3X (up to 5X) that loading for each tube for reasonable triode distortion, then 4X that for P-P would put you at min. of 3700 Ohm P-P (class A op). Some class AB range in there could lower that. You will probably want to sacrifice some power for reliability, so maybe not boosting the B+.

A variable power supply with current limiting for testing could be a big help.
These PS500XT things are cheap:
 

Attachments

  • Hoefer_PS500XT.jpg
    Hoefer_PS500XT.jpg
    175.2 KB · Views: 320
Last edited:
OOPs, that 3700 Ohm P-P I calc'd above would be for class aB Op already. More class A could lower that Zp-p.

Current limiting power is useful in any case. Even the timer on those Hoefer things can be useful for when the listener falls asleep on the couch auditioning the sound.
 
What about class AB1 but class A limit of about 1W?

I think that would basically fit the usual class aB description.

Could the 6LW6 in triode sub for a 300B I wonder...

a) 36LW6 in triode, 50 V/div horiz., 50 mA/div vert., Mu approx. 3.7 at 125 mA

b) 300B Mu 3.9 at 60 mA

c) $4 tube, for reference, best used as an emulant tube in inverse triode

d) $1 tube, 10 Watt, frame grid, Mu 30 (curve tracer selected)

e) same tube in Crazy/Twin drive (perfect pentode now!)

The _LW6 looks better, but probably just a poor curve tracer got used for the 300B datasheet.
 

Attachments

  • rsz_36lw6_t.jpg
    rsz_36lw6_t.jpg
    54.4 KB · Views: 310
  • 300B.JPG
    300B.JPG
    739.7 KB · Views: 288
  • Perfect_triode.JPG
    Perfect_triode.JPG
    163.5 KB · Views: 294
  • TFH_in_triode.jpg
    TFH_in_triode.jpg
    97.9 KB · Views: 280
  • Crazy_TFH.JPG
    Crazy_TFH.JPG
    172.3 KB · Views: 52
Last edited:
I think those would be -9V grid1 steps for the 36LW6 curves. That fits the Mu 3.7 spec, and 90V was the max step range available back when 10 steps (11 curves) were available on my Tek 576 tracer (my 1st modification).

Later modification (#2) to the Tek 576 allows 14 steps (15 curves) and 168V step range (so 14 steps of 12V)

------------------

The 300B curves look like they might have been combined curves from 2 step ranges. The last two curves (HV curves on RT) look inconsistent with the left 6 curves.
I used to use the same kluge when I needed more than a 90V step range, just wire in a fixed DC V supply in series with the grid1 lead wire and trace again. Been there, done that. The inconsistency on the 300B graph is probably due to capacitance to ground of the floating supply. (possibly some R to ground too)

-------------------

The 6LW6 datasheet is available online as the M2057 (different bottle and base for RF) The actual (and a nice clean copy) of the 6LW6 datasheet is available, along with all the other hard to find tube datasheets (like 6LR6), on DVD here:

TCA Notices & Announcements
 
Last edited:
Ex-Moderator
Joined 2011
Thanks! Lots of good information.

I'm thinking of building a 36LW6 PP triode amp but I'm having a hard time figuring out a good RLa-a. I'm thinking about 300V B+ and fixed bias at about 50ma. I use 12AV5GA at this level into 3k RLa-a and get a nice 20W, 30W with still less than 10% THD. Since the 36LW6 has more power though I was thinking of boosting the B+ but worried about the max screen rating. I assume the screen stopper will save the screen from excessive current though. Maybe a B+ of 380V?
What's the key spec you are shooting for? e.g., max power or best fidelity?
 
Probably a combo of the two. I'm thinking class A triode connection, but without the tube specs it's hard for me to calculate... I'm also not opposed to pentode connection with stabilized screen.

I also would prefer to use a B+ of about 300V but higher current which is why I'm thinking of the *LW6.
 
Ex-Moderator
Joined 2011
Your concern about the screen voltage may be overblown, Tubelab has done some tests, and here is what he said:

"In this experiment I wired a pair of 6LW6's into a Tubelab SE amplifier in triode mode. I used an adjustable power supply to allow operation from 0 to 400 volts. The maximum screen grid voltage spec is 275 volts. That turns out not to be a problem. The tube handles 400 volts without cracking a sweat. I have run the bias up to 125 mA per tube without any signs of distress. <snip>"

In triode mode, the screen grid dissipates very little power, so the voltage can well exceed its maximum voltage rating - just don't do that in pentode mode.

Your proposed 50mA bias suggested that you are going for maximum power, as it's biased closer to Class B operation, if that's the case, you might as well go pentode, you will get gobs of power! OTOH, for the best fidelity, you can crank up the bias with the tubes triode-connected for mostly Class A power.
 
Ya I saw that and forgot about it. Since the max cathode current is 400ma, maybe I could run the B+ at 150V with higher current and lower RLa-a? Maybe 200ma idle and RLa-a of 3k? Or is 3k too high at that OP I wonder.

I don't know if my math is right but I'm thinking 150V/0.2A=750RL = 3k RLa-a? Pd of 30W per tube?

Thank you all for your assistance. I have 4 36LW6 tubes on the way, and this is my first design from scratch.

If George is reading this I wonder if he remembers how much drive he had that wasn't enough? The drive I'm planning on using will swing about 140Vp-p and I'm guessing that's plenty?
 
Last edited:
The tube handles 400 volts without cracking a sweat. I have run the bias up to 125 mA per tube

That testing led me to make a 6LW6 triode SE amp. It played nice and sounded great. It just had one minor problem.....it blew up. One of the output tubes would go into red plate runaway after several months of use. This always happened when the amp was left idling for a while. The 6LW6 doesn't seem to like idling in triode for long.

I think George used 1.5k for his test in SE, so 3k isn't far off for PP.

At first the amp ran on 400 volts. The OPT's were 3K with 4 ohm speakers on the 8 ohm tap, so the load was 1.5K. When the first tube blew, I assumed it was a bad tube and installed another one. A few months later it blew. I reduce the B+ to something in the 350 volt range and the amp appeared to be happy for maybe 6 months, then poof another tube blew. The last time I swapped power transformers to get the B+ in the 325 volt range. Several months later I heard a hum and saw one of the output tubes glowing blue. I yanked the power cord.

Since then I dismantled the amp and haven't looked back. That was at 5 to 7 years ago. I still have the tube removed from that amp and it is so gassy that it will light up blue without even powering up the heater, just connect a power supply between cathode and plate and watch it glow. I believe that the last tube did have a minor air leak that has grown worse over time. When I pulled if from the amp the blue glow was minor. After sitting around for several years it goe much worse.

Were these all bad tubes?????? I don't know, but the odds are against it, especially since I have been really abusing another tube from the same batch. How much abuse? Well, I plotted these crude curves with a set of power supplies. Yes, I did hit 940 mA at 650 volts. That's 611 watts for maybe 2 seconds.......if they were wimpy tubes that one should have exploded!

Why the extreme testing? I'm looking at building a 1 kilowatt tube amp (500 WPC). That's going to take some really stout tubes.

I wonder if he remembers how much drive he had that wasn't enough?

The amp was made by hacking a TSE board which runs a 5842 with a CCS load. It could not push the 6LW6 into clipping from a CD player. That's with the 6LW6 running in triode at 400 volts. The grid was in the -100 volt range. That needs 200 volts P-P to hit clipping.

I had decided to save the 'LW6's for a big pentode amp, like the kilowatt monster. If I was to use the LW6's today they would be in a pentode amp with local feedback around the output tubes. Something like a monster sized version of Pete Millett's Engineers Amp, or driven by one of my Universal Driver boards.

I have stumbled on a unique method of getting triode performance from sweep tubes and have it working well with small tubes. After some setbacks with mid sized tubes I think I have them under control. I should get to testing something big soon, but there are too many outside influences to nail down a time frame right now.
 

Attachments

  • LW6_tests.pdf
    86.1 KB · Views: 84
Ex-Moderator
Joined 2011
The 6LW6 doesn't seem to like idling in triode for long.
Well that's too bad, so they appear to be more suitable for Class aB operation in order to keep the idling power as low as possible.

Thanks, man. Care to share the math as to how to figured that out from plate curves? Once again I'll add I've had no formal training in electronics.

Screenshot - 6_16_2018 , 6_58_42 AM.png
It's just the ratio of the change of voltage to the change of current as shown in the chart above.

As far as making a triode 150V 200ma idle in theory, they shouldn't blow up? :p
They probably won't blow up, but you would get very little output power.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.