I tried a search and didn't get anything useful, so I'm asking the question here. How close can you mount output tubes? Expressed in either centre to centre or preferably edge of glass to next edge of glass.
I particularly want to know for EL84s, where I have 4mm between them glass-to-glass and 4 in a row.
But feel free to include any tubes you have information for on the thread to make it more generally useful.
I particularly want to know for EL84s, where I have 4mm between them glass-to-glass and 4 in a row.
But feel free to include any tubes you have information for on the thread to make it more generally useful.
4mm sounds awfully close. I believe I've seen 1-1,5x the buld diameter mentioned as a rule of thumb.
1x the build diameter seems to be done in practice. My Leak Stereo 20 has EL84 tubes 38mm between centres, and the tubes are 19mm wide.
What would be the contra-indications of mounting tubes too close? Over-heating or what?
What would be the contra-indications of mounting tubes too close? Over-heating or what?
Last edited:
There are many dual power pentodes with their elements in close proximity in use and they seem to do just fine. E
1950 GEC specified 4 inches (10cm) apart with an unperforated chassis with the hot sides of anodes not facing each other. Over the years with straight tubes, EL34/KT series in some designs I´ve come across only a cm gap with the old rules completely ignored. I preserved the 4 inch rule on all big power tubes regarding hottest parts of anodes being completely away from each other...Clearly adding a perforated chassis with plenty of through-air makes an enormous difference. The same applies for EL84´s when it comes to tube holders but I will check on mine if all vendors apply the same internal anode positioning within the tubes.
Below, my twin channel 500W beast. I followed the 4 inch rule (this calls for a large 4mm ally chassis) as alot of heat is generated. Looking down the avenues, notice the symmetrical placement of tubes, anode wise. Any measure will prolong tube life and any hot components under. It may appear "boring" work with a pillar drill to architectually perforate such a chassis, but with patience is worth it. Note, the thick aluminium chassis capable of supporting two large output transformers weighing 13kg each, the mains sim but using switch mode. Hold-on Total weight 65kg.
Bench Baron
Below, my twin channel 500W beast. I followed the 4 inch rule (this calls for a large 4mm ally chassis) as alot of heat is generated. Looking down the avenues, notice the symmetrical placement of tubes, anode wise. Any measure will prolong tube life and any hot components under. It may appear "boring" work with a pillar drill to architectually perforate such a chassis, but with patience is worth it. Note, the thick aluminium chassis capable of supporting two large output transformers weighing 13kg each, the mains sim but using switch mode. Hold-on Total weight 65kg.
Bench Baron
If mounting the tubes center at least 2 times bulb max diameter is not possible make at least sure to mount the tubes the way the smaller surface area face each other. Measure glass temp with thermopads and, if necessary, derate to keep the glass temperature below the max temp specified on the datasheet.
Last edited:
Found with a Google search for "EL84 distance"
"An EL84 data sheet says the envelope has a diameter of ~22mm; with a 1.7-times-diameter spacing, the centers only need to be ~37.75mm (~1.5") apart, center-to-center. If you lose cooling efficiency due to tight tube spacing, you may need to replace cooling efficiency some other way."
That said, in practice is they are sticking out of the top of the chassis, 20mm is good enough, but if you mount then inside a sealed box (and make an oven) I would go to at least 25mm, 30mm would be nicer.
I mount 6P45S sweep tubes that run at ~40W Pd with 63mm centre to centre spacing, and any other tubes that can be used in my Modular Amplifier 1 designs. I built a phono stage that the tubes were under 100mm apart taking inspiration from old tube computers that had rows upon rows of 6SN7 basically touching each other.
"An EL84 data sheet says the envelope has a diameter of ~22mm; with a 1.7-times-diameter spacing, the centers only need to be ~37.75mm (~1.5") apart, center-to-center. If you lose cooling efficiency due to tight tube spacing, you may need to replace cooling efficiency some other way."
That said, in practice is they are sticking out of the top of the chassis, 20mm is good enough, but if you mount then inside a sealed box (and make an oven) I would go to at least 25mm, 30mm would be nicer.
I mount 6P45S sweep tubes that run at ~40W Pd with 63mm centre to centre spacing, and any other tubes that can be used in my Modular Amplifier 1 designs. I built a phono stage that the tubes were under 100mm apart taking inspiration from old tube computers that had rows upon rows of 6SN7 basically touching each other.
Last edited:
The tubes must be mounted so close to each other (and to other components, transformers, etc) so that you can safely grab them when plugging/unplugging. Cold, of course 😡
No it´s not quite the original. I found the original application report by the MOV Valve company 1972 in an old ledger. Amplifier circuits for KT88. You missed the important details see pic. Fig 3 the normal chassis mounting ensures the hottest part of anodes are at right angles which is the orientation I used.Here's the original GEC data for the KT88
Well done.....spec sheet..The only caveat are those old electrical running specs esp with fixed bias using today´s manufactured tubes which aren´t so robust.... and I would derate accordingly.
Some of us had good previous experience over the bad reputation of FEastern cloned KT88´s and others which after a short time created the perfect internal firework display. An early Shu 88 running at 560V # 60mA (only 34W) was guaranteed to display an orange uneven anode. This is a sure sign of shoddy internal alignment. I have some early 6L6´s pushed at the same rating didn´t show any colour on their bibs. I miss ´em.
Bench Baron
Some of us had good previous experience over the bad reputation of FEastern cloned KT88´s and others which after a short time created the perfect internal firework display. An early Shu 88 running at 560V # 60mA (only 34W) was guaranteed to display an orange uneven anode. This is a sure sign of shoddy internal alignment. I have some early 6L6´s pushed at the same rating didn´t show any colour on their bibs. I miss ´em.
Bench Baron
That said, in practice is they are sticking out of the top of the chassis, 20mm is good enough, but if you mount then inside a sealed box (and make an oven) I would go to at least 25mm, 30mm would be nicer.
45mm center-to-center EL84 distance inside a box with copious longitudinal, cca. 5mm wide openings on the bottom and on top is still an oven.
Speaking from practice - the inside of the top cover has big brown spots from heat of the output stage and that brown coloration cannot be washed anymore - so for the last year and a half or so i am using amp without top cover.
Without the top cover the cooling is now what is called "normal", no more oven.
I build my amps with the tubes sticking out of the top along with the transformers, and everything else inside the chassis with no ventilation holes at all. The only brown or black spots I've seen came from resistors blowing from fault current. Since I changed from 2W to 5W resistors for the screens (tied to plate as triodes), even an arcing tube won't take out anything else with it. Just turn it off, change the tube, and turn it back on.
The real trick to not seeing brown spots on the inside of the cover is to use a black chassis 😛
The real trick to not seeing brown spots on the inside of the cover is to use a black chassis 😛
The real trick to not seeing brown spots on the inside of the cover is to use a black chassis
Sure - the real trick for me is to never build an amp in closed chasis regardless of the amount of holes and whether they are on all sides ( which should be mandatory ).
Even seeing closed box preamps without any holes just horrors me - preamps also need holes if inside a closed box.
Cheers, Krca
For beam power tubes, make sure the 2 hot sides of the plate in adjacent tubes aren't facing each other. The KT88 data sheet talking about this, I am now intentional about how I index beam power tubes in any amp.
It's not only the box: where you put the box can make a difference.
My amps sit on open shelves and last forever.
A friend of mine, a décor oriented guy with fancy cabinets, managed to kill a Leak ST60 and an ARC D90 within a few months' time. Both amps are loaners, they belong to me. Subsequently, he also managed to kill a solid-state Accuphase he paid for, but that took him 10+ years. His cabinets have front openings for the gear, but not much else, so the hot air remains trapped.
Also, for my guitar combo amps, I always mount a recycled and under-volted computer fan blasting the output tubes at an angle, and with a escape path. My main amp has some teenage Chinese 6L6GC that are still rocking.
My amps sit on open shelves and last forever.
A friend of mine, a décor oriented guy with fancy cabinets, managed to kill a Leak ST60 and an ARC D90 within a few months' time. Both amps are loaners, they belong to me. Subsequently, he also managed to kill a solid-state Accuphase he paid for, but that took him 10+ years. His cabinets have front openings for the gear, but not much else, so the hot air remains trapped.
Also, for my guitar combo amps, I always mount a recycled and under-volted computer fan blasting the output tubes at an angle, and with a escape path. My main amp has some teenage Chinese 6L6GC that are still rocking.
- Home
- Amplifiers
- Tubes / Valves
- How close can you mount output tubes?