A word of caution from my experience.
It has been stated that the screen grid voltage rating can be ignored when the tube is triode wired. Some tube manufacturers publish higher screen voltage ratings when the tube is triode connected.
When the amp is idle, on but no music is playing, or the volume is very low, the output tube doesn't know if it's triode, pentode, or UL connected. It only knows that it is constantly seeing an excessive voltage on the screen. Most tubes can deal with this, but there are going to be some that may do OK for a while, then slowly start to exhibit "bias creep" (the idle current slowly starts to creep upward over time) leading to eventual tube runaway and blown parts.
This is especially bad in SE (always class A) and class A push pull amps, since idle is the worse case dissipation condition. Triode wired sweep tubes are particularly prone to bias creep and eventual runaway. I built a triode wired 6LW6 amp years ago. It worked great, and sounded excellent at 400 volts, but unexpectedly blew a tube after a few months. The second tube blew up after a few more months, so I reduced the plate voltage to about 340 volts and added a fuse in the tube's cathode connection. Over the next year and a half, two more tubes ran away, but blew the fuse. Each tube was no longer stable and could not be used after the initial runaway event. All runaway events were during idle conditions.
The 1625, 807 and 6BG6 ARE 6L6 types, and may be 6L6GA's, 6L6GB's or even 6L6GC's (rare) internally depending on who made them, and when they were made. I have seen 807's and 6BG6's that were 6L6GC's inside and even a few with 7027A guts. Obviously the ones with newer guts can eat much more screen voltage.
It has been stated that the screen grid voltage rating can be ignored when the tube is triode wired. Some tube manufacturers publish higher screen voltage ratings when the tube is triode connected.
When the amp is idle, on but no music is playing, or the volume is very low, the output tube doesn't know if it's triode, pentode, or UL connected. It only knows that it is constantly seeing an excessive voltage on the screen. Most tubes can deal with this, but there are going to be some that may do OK for a while, then slowly start to exhibit "bias creep" (the idle current slowly starts to creep upward over time) leading to eventual tube runaway and blown parts.
This is especially bad in SE (always class A) and class A push pull amps, since idle is the worse case dissipation condition. Triode wired sweep tubes are particularly prone to bias creep and eventual runaway. I built a triode wired 6LW6 amp years ago. It worked great, and sounded excellent at 400 volts, but unexpectedly blew a tube after a few months. The second tube blew up after a few more months, so I reduced the plate voltage to about 340 volts and added a fuse in the tube's cathode connection. Over the next year and a half, two more tubes ran away, but blew the fuse. Each tube was no longer stable and could not be used after the initial runaway event. All runaway events were during idle conditions.
The 1625, 807 and 6BG6 ARE 6L6 types, and may be 6L6GA's, 6L6GB's or even 6L6GC's (rare) internally depending on who made them, and when they were made. I have seen 807's and 6BG6's that were 6L6GC's inside and even a few with 7027A guts. Obviously the ones with newer guts can eat much more screen voltage.
Inna Gadda Da Vida! The neighbors hated that one when I was a teenager with several hundred watts of audio and huge cabinets with 4 X 12 inch speakers in each one!
For years I had a Tubelab SSE hooked up in my work room. It could be switched to the PC output, the turntable, the CD player, or a Samson Mixpad (small mixer). My workbench had a master switch, with a KILL button. In case an experiment got out of hand I could smack that button and everything except the PC went dead.
The SSE was always on whenever the bench was on, and connected by relay to the PC by default whenever the PC was on. So the amp idled a lot. It didn't seem to mind. I swapped in an SPP, and several other amps as they were built, even a 125 WPC version of Pete Millett's big red board. For the last several years in Florida, I had a switchbox so that I could swap up to 4 amps live by turning a knob, but somehow I always gravitated back to my original "industrial strength" SSE made on diamond plate aluminum. Sadly it was dropped during the move out of Florida, and doesn't look too good right now.
One of the amps that pushed the SSE out of its usual spot, was a bigger not so simple SE that used 6LW6's and made over 15 WPC in triode. The SSE reclaimed its spot after the 6LW6 amp ate its 4th output tube. I got mad and stripped it for parts!
Just FYI; the respected Sarser/Sprinkle Musician's Amplifier ran a pair of 807s in class A triode mode at 400V.
The 807 screen voltage in class A spec was uprated circa late 1940's to 400V due to the interest in deploying it as an alternative to the KT66 in the Williamson amp.
The 6CM5 sweep tube has been extensively used for PA amps, and for DIY (although not normally triode connected). Consistent with Tubelab's caution, that sweep tube was prone to runaway if idled anywhere near its supposed plate dissipation spec limit by DIYers. Philips successfully sold a lot of 6CM5 based PA amps up to sextet output stages, but operated them at close to class B, and used a 50% B+ for screen.
The 6CM5 sweep tube has been extensively used for PA amps, and for DIY (although not normally triode connected). Consistent with Tubelab's caution, that sweep tube was prone to runaway if idled anywhere near its supposed plate dissipation spec limit by DIYers. Philips successfully sold a lot of 6CM5 based PA amps up to sextet output stages, but operated them at close to class B, and used a 50% B+ for screen.
I am glad Tubelab_com spoke up.
In triode-wired mode, depending on whether a tube is the early or late design-and-construction and model version, some may not work very long when they are set to the maximum plate voltage rating as their quiescent state.
Also, in ultra-linear mode, depending on the percentage of the UL tap, and the plate swing, the plate will be far below the screen voltage. For example, take a 30% tap, with the screen at 405V and the plate at 400V (due to primary DCR drop). If the plate swings down to 150V (250V swing), the screen will only swing down to 330V (75V swing).
Now the plate is 180V below the screen voltage, at a time when the control grid is dictating max current.
Could be a problem.
In triode-wired mode, depending on whether a tube is the early or late design-and-construction and model version, some may not work very long when they are set to the maximum plate voltage rating as their quiescent state.
Also, in ultra-linear mode, depending on the percentage of the UL tap, and the plate swing, the plate will be far below the screen voltage. For example, take a 30% tap, with the screen at 405V and the plate at 400V (due to primary DCR drop). If the plate swings down to 150V (250V swing), the screen will only swing down to 330V (75V swing).
Now the plate is 180V below the screen voltage, at a time when the control grid is dictating max current.
Could be a problem.
Thank you Tube Lab, 6A3summer and everyone
The 1625's I have are 1944 date of manufacture. The 837's look to be the same Era. I believe both are Military grade. I know it's a stretch, however, I am running low current, around 35ma or less. I have to calculate it out. I'm using 2k for the cathode resistor. The high impedance opt mean less current is needed.
I don't know if the current makes a difference, the dissipation of the 837 and 802 may be near the limit...but far from the limit of the 1625. I may increase the screen resistor...to more than 100ohm to ensure limit dissipation and ensure the screen voltage is always below the plate.
The 1625's I have are 1944 date of manufacture. The 837's look to be the same Era. I believe both are Military grade. I know it's a stretch, however, I am running low current, around 35ma or less. I have to calculate it out. I'm using 2k for the cathode resistor. The high impedance opt mean less current is needed.
I don't know if the current makes a difference, the dissipation of the 837 and 802 may be near the limit...but far from the limit of the 1625. I may increase the screen resistor...to more than 100ohm to ensure limit dissipation and ensure the screen voltage is always below the plate.
Last edited:
It looks as though the plate dissipation of the 802 and 837 are beyond their maximum limit.. approx 13.5 watts.....Not too surprised seeing as this amp was built for the 1625. That's what you get for not measuring.
Now the question will be... lower the plate dissipation via higher cathode resistor, or stick with the 1625...hhhmmmm...ideas brewing...
Now the question will be... lower the plate dissipation via higher cathode resistor, or stick with the 1625...hhhmmmm...ideas brewing...
So, does keeping a lower dissipation allow for having a higher screen voltage?
I have been told by manufacturers saying that screen voltage can be higher as long as plate dissipation is kept within rated specs.
That rule makes sense to me.
This may be why some have success with triode strapped higher screen voltage... by keeping dissipation, cathode current, and other parameters well below limiting values.
I can totally, and have, seen thermal runaway at higher cathode currents with plate dissipation close to it's limits.
When I installed the 802's to my amp... It sounded good...but then I saw runaway start to happen. Sure enough, dissipation was too high, along with the screen voltage... I caught it in time... thank you amp meter mounted on the front.
One thing that may cause thermal runaway is when the Control Grid resistance is too high.
I was using a 6550 in a circuit that was originally for an EL34. I adjusted the fixed bias for the same current as the EL34. But the grid resistor was 270k Ohms, not the 6550's requirement of 50k Ohms maximum.
The sound got distorted, and so I looked at the amp, and the 6550 was red-plating.
So I turned the amp off. I checked the current, voltage, and power ratings of the 6550, and the amp was well within all those numbers.
The only rating that was exceeded was the maximum Control Grid resistor. The spec is there for a reason.
I was using a 6550 in a circuit that was originally for an EL34. I adjusted the fixed bias for the same current as the EL34. But the grid resistor was 270k Ohms, not the 6550's requirement of 50k Ohms maximum.
The sound got distorted, and so I looked at the amp, and the 6550 was red-plating.
So I turned the amp off. I checked the current, voltage, and power ratings of the 6550, and the amp was well within all those numbers.
The only rating that was exceeded was the maximum Control Grid resistor. The spec is there for a reason.
"FunAmplifier "with 837 tube
Hello there!
Ihave been reading with interest about your thoughts on 837 and 1625. I have a few 837 tubes and would really love to make acceptable sound with them...without blowing them up. may i ask if you have a circuit for them and opt details. plate and screen voltages. also can i drive this tube with 12ax7 or do i need to use a 6sl7. a single channel single ended would enough to start with.
.i am not particular about pentode triode or UL modes...just decent vintage sounding setup. the trannies can be wound to specs if they are available.
Thanks in advance!!
Amit
Hello there!
Ihave been reading with interest about your thoughts on 837 and 1625. I have a few 837 tubes and would really love to make acceptable sound with them...without blowing them up. may i ask if you have a circuit for them and opt details. plate and screen voltages. also can i drive this tube with 12ax7 or do i need to use a 6sl7. a single channel single ended would enough to start with.
.i am not particular about pentode triode or UL modes...just decent vintage sounding setup. the trannies can be wound to specs if they are available.Thanks in advance!!
Amit
Hi Amit,
It is nice to know someone out there is listening with interest... thank you.
The following amp I mentioned, Misty, did well with the 837. You could use a 12ax7 or 2 6f5's with the heaters in series... simply use different value cathode resistors for the 837 to keep the dissipation within limits.
Misty- SE 1625 Triode Strapped- Newest Build
It is nice to know someone out there is listening with interest... thank you.
The following amp I mentioned, Misty, did well with the 837. You could use a 12ax7 or 2 6f5's with the heaters in series... simply use different value cathode resistors for the 837 to keep the dissipation within limits.
Misty- SE 1625 Triode Strapped- Newest Build
Hi again!!
Thanks for sharing the details of your build. Looks fine and interesting, Im also of the idea to put the trannies below the chassis top plate. does that cause a problem with EMI to other passives of the tube circuit?? just a question i will post more quiries soon. btw do you have sound test video that i can see somewhere?
Best,
Amit
Thanks for sharing the details of your build. Looks fine and interesting, Im also of the idea to put the trannies below the chassis top plate. does that cause a problem with EMI to other passives of the tube circuit?? just a question
i will post more quiries soon. btw do you have sound test video that i can see somewhere?Best,
Amit
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.