about putting 4 Kt88s in a 5150

[MitchK]

I've been running a 5150 off of 2 kt88s tubes for a while now, it sounds great, feedsback a little more than the 6l6s but otherwise sounds good. The thing is though is that I run it at high volumes and I know I'm over heating the tubes in it. One started to runaway on me the other day, plates were glowing red, but I was able to switch it off before it shorted.

Anyways I've been told I'll want to change the screen grid resistors from 100ohm s to 1kohms, but I've also been told that the OT won't be able to handle a full 4 kt88s at higher volumes.

Can someone explain why this is? I'd just like to know more about how the resistance of the tube works in terms of it interacting with the OT in general.

[bob p]

a KT88 is a tube that's designed for much more extreme operating parameters than 6L6 -- the problem isn't necessarily that you're pushing the tube too hard in your application -- its more likely that you're putting the tube into an an environment that is not suitable for the tube.

before doing anything else, grab a copy of the tube data sheets and read them. if you're trying to drop KT88 into an amp that's not designed for them, you'll run into a few problems:

1. improper bias
2. improper spacing of the tube sockets / insufficient heat dissipation
3. excessive demands placed on the PT
4. possibly excessive power to the OT

the KT-88 is rated to produce much higher power and tolerate much higher voltages thana 6L6 when it is installed properly. if your amp is set-up to handle the KT88 properly, the KT88 should have a long, long service life. in all likelihood though, your amp is probably a bad environment for the KT88. even if you fixed the bias, heat dissipation might still be a problem, and your PT/OT may not be abe to live up to the demands that you place on them. i would rethink the plan.

[Enzo]

You might contact Bob at Eurotubes and talk to him about his KT88 infused 5150. Spiffy meters on the front too.

WIll four KT88s even FIT in the amp where the sockets are stock? Or will you have to mount sockets on the chassis and spread them out some?

Remember that tubes are valves - they control the glow of current through themselves. They don't create power that is not already in the circuit. I don't know the impedances and curves etc of the KT88. How much power can the power supply feed through the tubes? How do the impedances match up against the 6L6, and where will the stock speaker load impedance reflect back at the tubes?

[Enzo]

Good points. Adjusting the bias to proper levels might solve your two tube melt-down anyway. I don't think the heater current is a problem in the 5150.

If you plan to remove the power tube socket board then it doesn't matter, but if it stays here is a tip. To change the screen resistors, unsolder them and poke them out through the holes so they fall into the chassis. Fish them out. Now install the new resistors on the solder side facing you. NO NEED to remove the board from the amp.

[bob p]

the KT88 heater thing is something that really surprised me. i've been looking over the 6550 / KT88 spec sheets over the past few days to spec a transformer for a 100W HiFi stereo amp. one thing that really stuck out in my mind was the heater current: 1.6 amps at 6.3 V per tube, vs. 0.9 A for a 6L6. that means a quad of KT88 would draw 6.4 amps of heater current, compared to 3.6 amps for a quad of 6L6. Yikes!

the KT88 also require spacings of 4-inches center to center between the tubes, and orientation of the tubes is important so that the cold parts of the tubes face one another and the hot parts face outward.

the bias problem is significant as well. typical fixed bias operating voltages for a pair of ultralinear KT88 in class AB1 at 600 volts is something like -75 volts. in that configuration, a quad will produce 200 watts. when running full-tilt, its possible for maximally configured KT88 to need over 650mA of secondary current -- something that your 6L6 PT probably can't deliver. then, even if your PT can handle it the supply demands, your 100W OT probably wouldn't like being driven with 200W.

of course, your amp will have a 6L6 power transformer, it will run the KT88 at lower voltages, so the power output won't be so extreme and your OT might be able to handle it. my only point is to say that changing tubes isn't really as easy or as safe as plug and play. if i were you, i would grab the data sheets from frank's electron tube pages and the first thing i would look at would be the bias voltage to stop the runaway problem. then i would crunch some numbers to see if what you're trying to do really makes sense.

on a related note, i have an open thread about KT88 in the tubes section related to bias issues, so if anyone would like to read that one and help me out, i'd appreciate it.

[MitchK]

Well I more mechanical questions I can answer upfront,

I've removed the sockets from the chassis and mounted them on a sort of riser board I made so they now have approx. the correct spacing. they are mounted horizontally now though, the riser board I made connects perpendicular to the actual chassis, I might change this design though, reason being that the back is open on this riser which I'm sure is generating a certain amount of noise in the amp no? The obvious answer to this question is do I hear a difference but its been too long I can't really remember what it sounded like before this.

the other main reason for changing this is that heat is a big issue, the tubes are now sitting too close to the frame of the head and the heat has nowhere to go, but I'm hoping to maybe fix that with a fan.

This is the real reason the tube was running away on my I believe, I was also at the time running a 4ohm load with the selector of course on 4ohms when really I should have had a 8ohm load running the amp at 4ohms on the amp.

I have made an adjustable bias for it and am now running the tubes hotter than the amps fixed bias which is really cold, but still not near their max.

of course, your amp will have a 6L6 power transformer, it will run the KT88 at lower voltages, so the power output won't be so extreme and your OT might be able to handle it. my only point is to say that changing tubes isn't really as easy or as safe as plug and play. if i were you, i would grab the data sheets from frank's electron tube pages and the first thing i would look at would be the bias voltage to stop the runaway problem. then i would crunch some numbers to see if what you're trying to do really makes sense.

The thing with the 5150 is that it is a 120W amp even with the 6l6GCs so I'm a little worried, is this going to push the kt88s too hard? I mean kind of going by what enzo said the tube is just a valve so would a 120W amp run any quad of tubes at 120W, can one tube be more efficient than another some how?

Crunching the numbers is the thing I really don't know how to do, and want to know how to really badly. I don't really know how to read the data sheets, especially the graphs. I can pick up the basics like the heater current, which I've always noticed that too its a lot higher on the kt88s. But I don't really know how you're working out what you said about running the plates at 600v makes for a typical bias of -75v and that equals 200W which means the secondary current (does this mean the bias supply?) will need to be about 650ma max. How do you figure that out, how do the equations work?

[bob p]

Crunching the numbers is the thing I really don't know how to do, and want to know how to really badly. I don't really know how to read the data sheets, especially the graphs. I can pick up the basics like the heater current, which I've always noticed that too its a lot higher on the kt88s. But I don't really know how you're working out what you said about running the plates at 600v makes for a typical bias of -75v and that equals 200W which means the secondary current (does this mean the bias supply?) will need to be about 650ma max. How do you figure that out, how do the equations work?

i am no expert by any means, but i will try to help. none of the numbers that i mentioned were derived from / interpreted from the graphs on the data sheets. those numbers i mentioned were written in the text tables of the KT88 data sheet. look at this one:

KT88 data sheet

the numbers i cited were for an ultralinear application. granted that's probably not how your amp is set up, but its still worth looking at the numbers as an example. looking at the numbers can be confusing, because the USA, UK and Eastern European manufacturers all use different engineering abbreviations. in some respects, you have to look at the numbers and work backwards to figure out what in the heck their abbreviations are supposed to mean.

Go to PAge 4 and look at the top table, for a Class AB1 UL application. The first column is the tube characteristics for a 560 VDC plate voltage, which is in the 600V ballpark.The Voltages for anode and grid 2 (screen) are 560 VDC at idle. The idle current is 50mA per tube and increases to 157 at full signal. The grid 1 bias voltage is -75 VDC. A full scale 140 VAC input signal under those conditions will produce 100W output for a pair of KT88.

If you're running 4 tubes, the total currents and power output will double compared to 2 tubes. The table says that under those conditions, you'll need an anode+screen current supply of 157mA per tube. Multiply that by 4 tubes and you need over 600mA of current from your power supply to get 200W from 4 tubes.

keeping things in perspective, those numbers don't apply to your amp. to get a better idea of how your amp is set up, look at the 6L6 tube data sheets and try to find a set of reference data that resembles your amp's configuration. when you know the operating parameters for your amp, you can look at the KT88 data sheet to get an idea how that tube will perform under those conditions. chances are your PSU will never allow the KT88 to be al that they can be, and your KT88 will run in a safely under-rated mode in your amp. but you still need to watch out for heater current, bias, tube spacing, and heat dissipation.

if you're going to completely rework the innards of your amp, it would really be helpful to have a handle on some of the math that's involved. ultimately, you'll waste a lot less time and you'll burn up a lot fewer parts if you approach the problem by having used math to put you in the right ballpark. if you just try blindly hanging parts, you could be in for trouble.

If you don't have any experience with tube math, I'd recommend starting off by reading the AX84 P1 document. It walks you through a lot of the calculations that come from the tube data sheets. then once you have a better handle on what questions you really need to ask, you'll find that there are plenty of people here who will try to help. have fun.

[Ray Ivers]

Mitch,

Just to add to the other good stuff here... what idle current/plate dissipation are you using? I wouldn't go anywhere near 70% using two KT88's with a 2.5K OT; 40% to 52% (like an SVT) would be more like it. IMO an excellent candidate for this loading situation would be the EH KT90.

Also, not all KT88's are created equal - and in this application, it really does matter a lot. Just FWIW, I consider the original Genelex, Tesla/Vrsovice, and National KT88-USA's as achieving full Genelex spec-sheet performance levels, the early Chinese "KT88"s as coming nowhere even close, and the rest of the field (EH, SED, Shuguang, Valve Arts, GT, etc.) spread out in the middle.

Ray

[Amp Kat]

I think Mitch said he respaced the board and no 4 KT-88's will not fit on that power tube board. IMO the OT isn't going to take 4 KT-88's and as Ray said your going to have to crank back on the bias to around 50 to 60 % and even then at high volumes your pushing it. That tranny isn't very stout and it's unbalanced from side to side. When you bias it use the least highest ma side as your bias because it can be off 10 ma's per side. Not trying to be negative here as I love KT-88's and the way they sound but the 5150 is really not the amp for them with the stock transformers. This amp is preamp driven and the output section from the return in is as loud as any 100 watt amp on the market with the stock 6L6/5881 tubes and clean to boot.

[MitchK]

Hmmm I am more and more thinking about just restoring this all back to stock but I think I jsut want to fiddle with it at this point to see what I can do with it so I'm going to keep going for now, but I'm curious:

Does anyone know any sort of ratings for the 5150 Power transformer and Output transformer? I'm curious for example how much amperage the PT can handle, and just exactly what characteristic of the OT will give under teh stress of 4 kt88s, it doesn't seem to get very hot?