KT88 mounting confusion

[Gaz]

So. I've gone over the Genalex datasheet that say pins 4 and 8 should be on the same vertical line, but that put the plates at a weird angle. After looking at a bunch of photos I'm 99% sure that the pin-outs are slightly skewed on the new production tubes from the originals (at least the JJs). After looking at a tone of pictures, most of the KT88s seems to be alinged like this:

http://hi-end.on9mart.com/products/a...KT88-bk_06.jpg

Now, that would put the plate seams and 'fins' facing each other, which I think is correct. I'd like to believe the HI-FI dudes know what they're doing...

However, it looks like they're facing the other way in this vintage Hiwatt, although they are using modern tubes, which again, I believe have the pins slight skewed from the originals.

Hiwatt Internals

Does anyone know for sure for sure how they should go and which sides the the coolest?

[Steve Conner]

The plate seams are the hottest parts of the plate by far, which is why fins are fitted there. Ideally you want them not to be facing other tubes, so they have a clear path to radiate the heat away. So, the Hiwatt has it right and the hi-fi amp is wrong.

When mounting tubes horizontally, you also want the grid wires in a vertical plane to mitigate sagging, which means that the plate seams should face to left and right, not up and down. This could well conflict with the requirement for heat radiation, unless you mounted the tubes in a vertical row.

I mention horizontal mounting and sagging grid wires because I believe this is what the recommendation in the Genalex datasheet is for. I don't think they were too bothered about orientation when the tubes were used in the usual vertical position.

And yes, modern KT88s can have the plate assembly set at a different angle to the old GECs. There is no standard nowadays, so I prefer to avoid horizontal mounting altogether and leave generous spacing between the tubes. That way they should get reasonable cooling no matter which way the plate seams end up.

[Gaz]

The plate seams are the hottest parts of the plate by far, which is why fins are fitted there. Ideally you want them not to be facing other tubes, so they have a clear path to radiate the heat away. So, the Hiwatt has it right and the hi-fi amp is wrong.

When mounting tubes horizontally, you also want the grid wires in a vertical plane to mitigate sagging, which means that the plate seams should face to left and right, not up and down. This could well conflict with the requirement for heat radiation, unless you mounted the tubes in a vertical row.

I mention horizontal mounting and sagging grid wires because I believe this is what the recommendation in the Genalex datasheet is for. I don't think they were too bothered about orientation when the tubes were used in the usual vertical position.

And yes, modern KT88s can have the plate assembly set at a different angle to the old GECs. There is no standard nowadays, so I prefer to avoid horizontal mounting altogether and leave generous spacing between the tubes. That way they should get reasonable cooling no matter which way the plate seams end up.

Steve,

Thanks a lot. I'm mounting the tubes vertically @ 4" apart. It makes sense to face the seams towards each other as you said. I attached a photo just to triple check. Like this?

And if that's correct, it seems like just about everyone else is doing it wrong...

[J M Fahey]

Why doesn't that surprise me?
Although I wouldn't use the word "wrong", it's too black and white.
I would go for "more or less efficient" or something like that.

[loudthud]

The plate seams are the hottest parts of the plate by far, which is why fins are fitted there.

I don't think this is right. When the tubes are red-plating, the seams are still gray. Where's that picture of the 6550 or KT88 with the hole melted in the glass?

[JoeM]

6550

[diagrammatiks]

with how much space there is between tubes in a guitar amp I don't think the orientation is that much of an issue.

the ones that are mounted "wrong" are probably mounted that way to make the underside wiring easier.

[Gaz]

You know, after looking at the photo, the 'pins 4 and 8 in line' from the Genelex datasheet would put the tube modern JJ tubes I have at just the right angle! You're right, Loudthud - red-plating always seems to start right at the 'crease' in the plate, not right on the seam itself. But you know, close enough for rock. I just didn't want to be a dummy and put them exactly the wrong way, which is what I was about to do.

[kg]

the hottest parts of the plate are going to be orthogonal to the line between the two grid support pins, since that's where the beam of electrons is focused.

[Gaz]

the hottest parts of the plate are going to be orthogonal to the line between the two grid support pins, since that's where the beam of electrons is focused.

Thanks KG. Cool, that's in-line (pun) with everything else I've been reading.

[Enzo]

I think Steve got it many posts ago. The data sheet refers to mounting the tubes horizontally - sticking straight out the back of the amp. The data sheets don;t care at all which way the tubes aim when standing vertically. It isn;t about the tubes somehow interacting with one another, it is about which way gravity tugs on the internals when they are really hot and wanting to sag. Think about it, there is no vertical line between pins 4 and 8 when the tube is standing up.

As to the seam being hot, my impression was always that the seam was the center of the plate target area. The tubes redplate to the side of the seam because the seam itself has all the extra metal heat sinking some of the excess away. The red stripe on either side of the seam is the edges of the beam, essentially. SO basically the center of the target area is ccooled by the seam even though it is getting the most electrons.

[Gaz]

I think Steve got it many posts ago. The data sheet refers to mounting the tubes horizontally - sticking straight out the back of the amp. The data sheets don;t care at all which way the tubes aim when standing vertically. It isn;t about the tubes somehow interacting with one another, it is about which way gravity tugs on the internals when they are really hot and wanting to sag. Think about it, there is no vertical line between pins 4 and 8 when the tube is standing up.

As to the seam being hot, my impression was always that the seam was the center of the plate target area. The tubes redplate to the side of the seam because the seam itself has all the extra metal heat sinking some of the excess away. The red stripe on either side of the seam is the edges of the beam, essentially. SO basically the center of the target area is ccooled by the seam even though it is getting the most electrons.

Enzo, I actually have to disagree with you concerning vertical mounting. You're right about the concerns of 'sagging' when mounting horizontally, but everything I've been reading has been saying that vertical spacing and orientation does matter because of the tubes' asymmetry, having two sides being cooler than the others. Facing the cool sides towards each other prevents one tube from radiating heat to the next.

You're wrong about the Genalex datasheet too, which specifies parameters for both vertical and horizontal mounting. It says that tubes should not be spaced less than 4" apart (center to center), and that pin 4 and 8 should be in line. This puts the hot parts of the tubes facing away from each other like we've been discussing. Horizontal mounting requires the same distance, but put the hot parts facing each other, and I visualize the radiated heat going out and up in that arrangement.

I'm not sure how much this matters in practice, but I like to not 2nd guess that datasheet when I'm working with unfamiliar tubes.

[loudthud]

The KT88 is a beam tetrode so electrons go where the beams focus them. Examining a Chinese made example I have (no branding on the tube, got them from the guy on ebay who sells Groove tube rejects) aligning pins 4 and 8 puts the screen and control grid support posts in a vertical line. The beams however appear to be in four corners of an upright rectangle exposing the plate structure at the seam, which is in the horizontal plane with the cathode, and vertically to the plate which has three holes on each side.
　
From the Genelex December 1974 data sheet page 5 link: http://www.triodeel.com/kt88p5.gif

INSTALLATION

The tube may be mounted either vertically or horizontally.
When tubes are mounted vertically it is recommended that the centers of the
tube sockets are not less then 4in apart and that pins 4 and 8 of each tube are
in line.
When tubes are mounted horizontally it is recommended that the centres of
the tube sockets are not less than 4in apart and that pins 4 and 8 of each tube
are in the same vertical line. One tube should not be mounted directly above
another.
Free air circulation around the tube is desirable.

My earlier comment might have been in reference to other tubes I've seen go red plate.

Old Sunn amps position power tubes on 4.5 inch centers with the keys facing the back of the amp.

[Gaz]

[SIZE=2]The KT88 is a beam tetrode so electrons go where the beams focus them. Examining a Chinese made example I have (no branding on the tube, got them from the guy on ebay who sells Groove tube rejects) aligning pins 4 and 8 puts the screen and control grid support posts in a vertical line. The beams however appear to be in four corners of an upright rectangle exposing the plate structure at the seam, which is in the horizontal plane with the cathode, and vertically to the plate which has three holes on each side.

I'm having trouble visualizing what you're describing Got a picture?

[kg]

http://www.radiomuseum.org/forumdata...way_Spbg57.png