I assume that you do know that a tubes electron action happens on both sides of the cathode, not just one side.
Right, 2 sides are hotter, two are cooler, hence the "90 degrees"
-
Feeling lucky? Check out the AK 24th Anniversary Giveaway thread and join in the fun!
Tube socket clocking
- Thread starter linuxslate
- Start date
Right, 2 sides are hotter, two are cooler, hence the "90 degrees"
These beam tubes concentrate their heat on two side of the tube opposite each other, so 90 degrees from this hot spot, the tubes are cooler. When these tube red plate, that is where it will happen. You don't want to face these hot sides towards each other if you can help it, especially if the tubes can't be spread far enough apart. I learned this from the KT88 datasheet. As someone else noted, you have to pay attention which way they are oriented if they are near the power transformer too.
Last edited:
Nevermind. 
linuxslate
Well-Known Member
I assume that you do know that a tubes electron action happens on both sides of the cathode, not just one side.
That part, I did get. The Illustration on the Wikipedia page for the 6L6 has an illustration showing the 2 beams. Interestingly, the Beam Tetrode page does not include the same drawing.
So for my 7591 amp, I have rotated the power tubes so that the center post keyway is now at the ~1:45 clock position (50 deg CW from straight up). This is close to the 4 - 8 4 - 8 arrangement, and again easy to measure accurately.
It also somewhat prevents the power tubes from "shining" on the pre-amp tubes. I would think it would be more important to keep the power tubes from "shining" on the preamp tubes than on other power tubes, and certainly more important than worrying about shielded output transformers.
JohnT123
Super Member
Thanks, I knew about this but thought there was maybe another issue. If it's only radiated heat I'm good.
I'm laying out an amp using quad 5881s and was planning on having the "hot" portion of each row face each other (it's most convenient for heater wiring and overall layout). However, I do have a good amount of space between the sockets (see attached picture). Should this be enough, or should I re-think the layout?
EDIT: Upon second thought, I think the anodes of the Tung Sol 5881s are somewhat offset from the keyway, so they're naturally staggered a bit. Nice.
EDIT: Upon second thought, I think the anodes of the Tung Sol 5881s are somewhat offset from the keyway, so they're naturally staggered a bit. Nice.
Attachments
Last edited:
AmadeusMozart
Well-Known Member
The datasheet for the KT66 mentions 9cm (3.5") between centres. I prefer to take at least 1.5* the diameter of a tube between them with some ventilation holes in the chassis around them. I do not like power tubes' sockets mounted on a pcb, driver tubes are fine on a pcb. Power tubes on phenolic PCB were very problematic in the old black and white TV's - tracks would lift and solder would get so hot it would melt resulting in bad contacts.
I like to mount the sockets under the chassis and have a tube poking up through a wide opening - gives an ariflow right from the socket upwards. Cosmetically good looking too but not as tidy underneath. Have now moved to putting the sockets directly on the chassis (without sub chassis) and using Pearl tube coolers on the power tubes.
I line up the sockets so the number 1 pins face in the same direction. For me, this can make repairing and tracing connection issues a bit easier, in the future.
linuxslate
Well-Known Member
I looked at that amp (Fisher 680-A) to see how they oriented the tubes. Event though they staggered them to maximize spacing, it looks like they've ignored pretty much everything discussed in this thread. There's like a 1/4 inch between the tubes and the OPT's! -- And that's Fisher! If anyone does not want their Fisher 680-A chassis (or the console it was used in) because Fisher "broke the rules", please send them to me.
Also based on the Fisher 680-A, I have rotated my OPT's 90 degrees. This allows me to move everything back on the chassis by 1/2 inch. This gives me more room around the driver/PI tubes.
I wire the sockets so that the heater pins are towards the side walls of the chassis. The heater wires run around the periphery of the chassis in the channel formed by the top meeting the sides of the chassis. I then purposely run the signal wires so that they don't cross the heater wires if at all possible. That's the primary reason that the heater wires are run first.
This is why I have taken to drawing out all new builds, in something like Microsoft Paint. I have little cut-and-paste pictures of various tube socket types, scaled to about 50 pixels to the inch. I make a big rectangle to scale for the chassis, smaller scaled rectangles for where the transformers will reside, scaled circles where can caps will go, and such- then start positioning tube socket images, until I have them spaced to where I think it looks good. I then try to see where terminal strips will go, making sure to have enough terminal locations for all resistors, caps, and such. This will oftentimes require rotating the tube sockets 90 degrees or 180 degrees, until I get a layout where as few components have to cross over each other, as possible (this makes post- build service SO MUCH easier), and where noise-sensitive input terminals and their associated circuitry, is isolated, as far from noisy stuff (rectifiers, heaters, 120V power circuits, etc) as possible...
Here's one of my diagrams- this one is for the "military transformer" 6L6 integrated amp that I built. It may not make much sense to anyone else but me, but it allows me to reliably build without a lot of assembly errors.
BE AWARE: This was NOT the final configuration of the amp (I didn't redraw the layout diagram- I just updated the schematic as minor things changed)- in fact, it doesn't show all the wiring, at all (yes, there WERE heaters, naturally- though they're not shown here).. Especially given how much work Dave Gillespie and I put in, to improve stability, after the initial build- there's likely a significant number of things that are now changed, even in the things that are shown here. DO NOT try to just build this, and expect it to work without a lot of tweaking!! But, this diagram is still indicative of the process I use...
Regards,
Gordon.
Last edited:
Yeah, it seems like the guidelines people follow regarding tube spacing and ventilation holes were not considered essential back in the mid-20th, because many (if not most) amps I've seen from the 40s-60s seem to violate at least one of them.
There's a huge difference between a manufacturer who builds one or more prototypes for testing - followed by thousands of production units - and an amateur hobbyist who has to make the first one count. Fisher apparently found that this configuration worked well in this particular product, but unless a builder is planning to use the exact same circuit, transformers and chassis, it's risky not to follow the guidelines more closely.
Jack
Jack
Replacement tubes were also cheap enough to be considered true "consumables" in amps, in years past, too. Putting tubes close together, was a cost-saving method, that was considered "good enough" under those conditions.
I prefer to not have shortened lifespan on tubes, now, due to their relative rarity of supply- so I tend to space them a bit farther apart than many vintage designs, whenever possible. Especially on tubes that are run close to their ratings, in terms of dissipation. It's one thing, to have an rugged output tube (like a 6L6 or such) that's only running at 50 % or 60% of max dissipation- they could legitimately probably be put almost touching each other, in pretty much any orientation. But, the same tube- or a more vulnerable tube (such as a 7591)- at 80%+ max dissipation... that needs more space, for cooling.
Regards,
Gordon.
I prefer to not have shortened lifespan on tubes, now, due to their relative rarity of supply- so I tend to space them a bit farther apart than many vintage designs, whenever possible. Especially on tubes that are run close to their ratings, in terms of dissipation. It's one thing, to have an rugged output tube (like a 6L6 or such) that's only running at 50 % or 60% of max dissipation- they could legitimately probably be put almost touching each other, in pretty much any orientation. But, the same tube- or a more vulnerable tube (such as a 7591)- at 80%+ max dissipation... that needs more space, for cooling.
Regards,
Gordon.