KT-88s: To the Pain

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I presume you've got a copy of the RDH?
Well, 4th edition is online, and I've been working my way through it. That's probably the top of my list for reference document review.
And read everything that the Valvewizard wrote?
Everything pertaining to the output section, and a smattering of the other stuff there. I do plan on hitting everything eventually, maybe sooner rather than later.
As well as Rob Robinette, who I mentioned earlier, there's a bunch of other people who've gone through and analysed various guitar circuits (e.g. AmpBooks)
I think I'm going to pop for the one on output sections. I doubt it will answer most of my questions, but the table of contents promises much of interest.
Then there's probably a dozen or so threads worth reading here and in Tube Audio, like the 6L6 AB2 thread
I got about 30 pages into that before my eyes started to bleed. It seemed to have drifted away from the original concept. I'll probably come back to it.
For bonus points, work your way through tubelab_com's postings. There's several books worth of material there, if only we can find him an editor :D
Well, that's exactly the problem. But I see he has posted below with some excellent suggestions on how to prioritize.
On "weird ways to drive a valve", get every copy of Sound Practices you can find (e.g. here and the associated Joelist postings) as the lunatic fringe of tube HiFi (and I mean that in a most positive way) spent the last quarter century trying and forensically analysing every historical circuit they could find (e.g. this) and making weird stuff up (like the "monkey" and the "free lunch").
I'm not looking for weird for the sake of weird; rather, I'm just trying to make sure I don't leave something on the table as far as clever ways to extract a few more watts. Since you mentioned it, I took a look over there, and...I think I need more RDH before I dive into that end of the pool.
Do take your sense of humour and perspective with you, as some of that crowd were channeling Hunter S. Thompson!
We should get along fine, then.
 
Getting 250 watts from a pair of $15 tubes isn't too hard either, but they were TV sweep tubes with top caps, also driven with mosfets, breaks three of your criteria.
I'm not trying to be difficult, but I do need a bit of self-convincing before I abandon a constraint.

The first generation Ampeg SVT bass amp made 300 to 400 watts from 6 X 6146 transmitting tubes.
For me at least, that's an impressive bit of Ampeg trivia! I wouldn't consider that much of a commercial success, though.

McIntosh (of HiFi amp fame) made the MI-350 amp used in the Grateful Dead's wall of sound, and at large outdoor festivals like Woodstock.
I'm not sure I'd call that a music instrument amp, but given that it was used as one...

All speakers have a mechanical resonance where the electrical impedance rises to a value well above its rated impedance. 30 to 40 ohms is not uncommon on an 8 ohm speaker.
And that perceptive little insight explains so much that has been the source of confusion in several other forums (not half so technical as this one). It puts several quotes from musicians and amp techs in context. It also explains why gear rated for a voltage fails when oestensibly operated at that voltage. KT-88s at 800V? No problem. At 2000+ V? Yeah, that's a problem.

Even if you have tube sockets made of pure Kryptonite, the arc can occur INSIDE the base of the tube, or across the bottom of the base or the socket.
So, since I can't control what's plugged in downstream, I can't prevent large voltage transients from entering my system that way. And since I don't manufacture my own tubes, I can't brute force my way to a bulletproof system.

So, the first thought that pops into my head is, is there any way to provide protection for the system? Something that might shunt off these massive voltage transients to someplace harmless before they get to the tube? This is something I could be talked into using silicon for, if it's even possible.

Those old amps that ran the KT88's on 600 volts or more are mostly dead now, for this reason. So I wouldn't venture much beyond 500 to 550 volts on these tubes, but if you must, use good quality old stock ceramic sockets and thoroughly clean them before use.
I would love to have both B+ measurements and reliability data for currently manufactured amps that are claimed to run at 650 V "or higher"...as well as what sockets they use.

That design will work, but expect to see somewhere over 4,000 volts on the tube sockets and OPT when you "set the controls for the heart of the sun." Peak power will be over 150 watts at about 40% distortion.
And against that, all I have is proper fusing and good NOS ceramic sockets? I'm starting to see the problem.

I can do the same thing on a bit over 450 volts with a 3300 ohm load and AB2.
Does "the same thing" include similar stress on the sockets and OPT, or is there some relief from backing down to 450 V?

The impedance of G1 transitions from near infinite, to less than 1K ohm when the grid goes positive, and drops rapidly as G1 is driven further positive.
I'm groping in the dark here because the manual doesn't have:

Plate characteristics as a function of postive G1 voltage
G1 current requirements as a function of anything
G1 positive voltage operating limits: voltage, current, power

How would I best find these and any other characteristcs I need for design?

There are two ways to do this, a driver transformer or a follower.
I might end up at that MOSFET source follower yet, but I think I have to work through the exercise of convincing myself that I shouldn't be using a driver transfomer first.

You can search through the "Tubes" forum for terms like "Dual Drive" "Twin Drive" and "Crazy Drive."
Thank you for that tip; I don't think I would have connnected that schematic to those terms otherwise. I am enamored of cycling the screens with the signal as a means of reducing screen dissipation, especially when in that "on-off switch mode" that comes from having mutiple booster pedals and preamp stages dimed.

And thank you in general; this post was just exactly the kind of thing I was looking for: both things that I knew I didn't know and wanted to find out, and things that I didn't know that I needed to know, but needed all the same.
 

PRR

Member
Joined 2003
Paid Member
>> The first generation Ampeg SVT bass amp made 300 to 400 watts from 6 X 6146 transmitting tubes.
> ...Ampeg trivia! I wouldn't consider that much of a commercial success, though.


Ampeg sold quite a few of the 6146 SVT; basically limited by customers who could afford the insane price. The re-vamp with 6550 was THE reference BIG bass amp for decades. Early on it inspired the Fender 300. The "SVT" *brand* lives on 50 years after birth. Maybe not "commercial" like selling 15W starter-amps in mass quantity, but no turkey.
 
i wonder what you mean by "seek to improve the state of the art...."
I would be happy to generate a few more watts of output power without compromising reliability.

guitar amps are all about sonic signatures more than anything...
I"m not chasing any specific sonic signature. It's possible that I might discover a new and somewhat unique signature, but that would be icing on the cake.
 
Don't want a mosfet, then find a suitable tube from your "approved list" with sufficient Gm to make a decent follower. I have successfully used the 5687 tube, but I suppose you could try the EL84.

I’m working on a “big tube amp” concept - starting with defining the iron and then filling in the required architecture. Same as I usually do for transistor amps. Two AS4T430 in parallel and Hammond 1650W output, plus a smaller Hammond power tranny I have on hand for the front end - the most I’m willing to spend on iron - no custom $1000 transformers. Looks like paralleled 6550 or KT88 would be needed if sticking to common tubes. If I were going to run six tubes and wanted to use direct coupled followers in AB2 to get every stinking volt off the B+, it would be necessary to use one for each valve so they can be individually biased, correct? No brainer for mosftes, but if I got a bug up my butt to use EL84’s would they need to be matched? Saves $100 if I can use what’s in my stash.

I’m also considering free wheeling diodes across the output tubes, and a snubber across the primary to reduce spikes and ringing, unless it seriously hurts the sound. Or is that why it’s not done more often?
 
I would be happy to generate a few more watts of output power without compromising reliability.


I"m not chasing any specific sonic signature. It's possible that I might discover a new and somewhat unique signature, but that would be icing on the cake.

if you haven't heard it yet, there is nothing new under the tube sun....
but hey, don't let me stop you, you may hit gold someday...
 
I haven't specified, but if it has a fairly clean sound up until it breaks, a decent distorted sound after it does, and good bass response (and my foray into designing an output transformer didn't stint on the iron to do that), it should be suitable for both.

There's precedent for guitarists using bass amps. A little less in the other direction, but still some examples.

Question for everyone here: I've been thinkinig a lot about Tubelab_com's comments on induced voltage (as well as all the other warnings about voltage I've gotten). There are "high voltage" octal sockets for TV high voltage rectifiers at ~25k volts. I was unable to tell if a KT-88 would fit them; has anyone here tried using those sockets for anything? They wouldn't solve the "arcing inside the tube" problem, but I suspect they would keep down external arcing if used properly.

I've also heard those same high voltage rectifier tubes were used in x-ray machines, and don't know if they still are. If people are still building x-ray machines with tubes, they need sockets, right?
 
I've also heard those same high voltage rectifier tubes were used in x-ray machines, and don't know if they still are. If people are still building x-ray machines with tubes, they need sockets, right?

Yes, and the tubes are held in insulating/cooling oil. Large amounts of lead are included and much of the internal communications is done by fibre optic.

Trust me when I say you don't want to go down this route......
 

PRR

Member
Joined 2003
Paid Member
...There are "high voltage" octal sockets for TV high voltage rectifiers at ~25k volts....

Usually a 25KV TV rectifier has the 25KV on a plate cap, not in the socket.

"Damper diodes" are different. 6AU4 is rated for 4500V on an octal socket, but the pins next to the plate pin are omitted, and normally we'd also omit the socket contacts so as not to leave a stepping-stone for the 4.5KV.
 
There are "high voltage" octal sockets for TV high voltage rectifiers at ~25k volts.

Eljefe, forgive me for asking: Do you by any chance have any certifications and training in handling the kind of terrifyingly lethal voltages found at electrical sub-stations? Do you have the sort of protective gear that electricians working on those voltages use? Do you have the testing regimen in place to ensure the protective gear is working as it should?

I am becoming increasingly concerned that your posts will suddenly cease, and it will be because you have electrocuted yourself. :eek:

High voltages can do a lot of extremely horrible things to a person, and they can do it in ways entirely alien to those of us who don't have the relevant training. For example, dust that has settled on the exterior and interior of a high-voltage circuit breaker can start an arc-flash when the breaker is opened or close, burning off your face - but leaving you alive, though very likely blind and deaf as well as without a face. :eek:

Please, don't do anything to put yourself in extreme danger. It's not worth it. Leave the electric-chair voltages to those with appropriate training, equipment, support personnel, and work environments.

-Gnobuddy
 
I'm not trying to build an amp with 25kV on the B+.

I'm trying to see if there are sockets rated for high voltage that will fit KT-88s, as opposed to plugging the tubes into cheap Chinese sockets and then wondering why I have catastrophic arcing.

And yes, many moons ago, I did work for the power company. "Just keep your hands in your pockets and don't even point at anything."
 
Airflow and insulation

If and when you eventually construct this "monster", take into consideration the airflow around the sockets.
Hot components cause a fast airflow, drawing in dust which partially due to static, adheres to the HV points and surroundings.
A design that show no signs of arcing when new, may fail at a future inopportune moment.
Probably the worst construction would be to drill a ring of holes in the chassis around each valve, as is commonly done.
The amp I currently use, and has been in daily use for 10 years now, uses 1200v HT but has almost no dust because it has a solid chassis isolating the "hot" from the "warm".
 
guitar amps are all about sonic signatures more than anything....more often a sonic signature unlike another... I do need a bit of self-convincing before I abandon a constraint.

Guitar amps ARE all about the "tone." Want a different "tone" be prepared to do something different. In 70 some years of guitar amp evolution, just about every possible way to connect the typical guitar store tubes together has been explored.

I started building my amps in the early 60's as a young kid with no money. Dead radios, HiFi's and TV sets could be found everywhere, often in the neighbors trash. That's where I started, and yes my amps sounded somewhat "different."

McIntosh (of HiFi amp fame) made the MI-350 amp.....I'm not sure I'd call that a music instrument amp

The letters in the model number indicate Musical Instrument use, but it was intended for sound reinforcement (PA).

KT-88s at 800V? No problem. At 2000+ V? Yeah, that's a problem.

The internals of most tubes can handle plate voltages well beyond the specs. It's the bases and sockets that are the problem. When the first "power tube," the 6L6 was introduced in 1939 it was given the same base hookup that had already been used in other octal output tubes. Nobody ever envisioned kilovolt level voltages on the plate. When kilovolt versions were needed, the tube makers simply stuck a cap on the top, and rewired the base. These are the 6BG6 and 807.

is there any way to provide protection for the system?

Several methods to limit the peak voltages have been introduced, and I have tried most of them. Some work well on HiFi amps where operation well into clipping is not the normal mode of operation. Anything that limits the plate voltage will by nature limit the peak voltage sent to the speaker, and thus change the sonic signature of the amp. Limiting voltage peaks will also reduce the apparent "power" of the amp. Depending on what you are trying to accomplish, this still may be a valid choice.

Does "the same thing" include similar stress on the sockets and OPT, or is there some relief from backing down to 450 V?

Reducing the B+ from 600 volts to 450 volts is a reduction of about 3/4. The peak plate voltage in overdriven conditions SHOULD also be reduced by a similar amount. It's however, not that simple.

You can get 100 watts from a pair of KT88's by running them at 600 volts into a 5000 ohm load in AB1. You can also get 100 watts from a pair of KT88's on 450 volts into 3300 ohms in AB2. Will they sound exactly the same? Are the electrical stresses the same? No, in both cases. Will two identical amps sound the same if OPT's with identical specs are swapped? Again, the answer is no.

An OPT is an exercise in tradeoffs and compromises. Each will have different stray capacitance, leakage inductance, and other unwanted characteristics. These GREATLY INFLUENCE the sound and induced voltages. The peak voltage seen on the output tubes are MOST AFFECTED by the speakers, since their resonant peaks are all different.

I'm groping in the dark here because the manual doesn't have....Plate characteristics as a function of postive G1 voltage

The KT88 and its US brother the 6550, were never intended for positive grid operation, so there is no data in the manuals. However the monster class B amp you found obviously drives the grids positive, since a statement claiming " the output stage grids draw some 30 mA at full output" appears in the text.

Here some experimentation is needed, and as I have always said, if you are looking to squeeze every last watt out of a tube, you will probably blow some stuff up while finding the recipe.

There are "high voltage" octal sockets for TV high voltage rectifiers at ~25k volts.....Usually a 25KV TV rectifier has the 25KV on a plate cap, not in the socket

The rectifier has a large sawtooth waveform on its plate, and DC on its DH heater. The rectifier socket was specially designed for the HV rectifier tube (usually a 3A3), and will not fit the diameter of any useful audio output tube, and most would not take the heat.

The HV REGULATOR tube (usually a 6BK4) has HV DC on its plate and all other elements run on low voltages.....these are the tubes that amateur X-ray guys use.

handling the kind of terrifyingly lethal voltages found

The 25 KV supply in old TV sets had enough stored energy to blow you across the room, but was not usually lethal. The resulting muscle spasm from being zapped by it could result in a deep gash requiring 30 some stitches to fix and a scar that I still have 50 years later. The B+ of an old TV, or worse a 100 watt tube amp is FAR more likely to be lethal!!!!!

I'm trying to see if there are sockets rated for high voltage that will fit KT-88s

I found a box of NOS ceramic sockets in a military surplus store maybe 35 years ago. Those are what I use.

As with ANY project of this magnitude, it is wise to first figure out what your main objectives are. There is a difference between "loudest amp" and "most powerful amp" due to each amp's sonics, and how they interplay with the speaker. Sometimes a really powerful amp doesn't sound as loud and a slightly less powerful that distorts in a way where much of the distorted energy is in a frequency range where the speaker is more efficient, or the ear is more sensitive. A third amp may sound better cranked, because it's not as harsh in overload.

A speaker cabinet will get louder with increasing power up to a point. Beyond that, more power can't push the cone any further in or out, it just makes more distortion. There is a point where the speaker / amp combination just sounds best, and more power doesn't help.......And I haven't even mentioned the guitar, type of music, or playing style yet.

Figure out what you want, then figure out how to get there. There are often two or more paths, and some are easier than others.

I have had the necessary parts to build a 500 WPC stereo tube amp for 10 years. I don't really need one, and therefore haven't started on it yet. I, however have set down my design goals, and have several paths to get there. Most of them have been crossed off the list, and it's getting down to a couple of choices, so build time may not be too far off.

In ANY design, the probability of success goes down as the power or voltage level goes up. Experience in building big amps that are reliable plays a big part in the success factor, and does prep and design choices. The more restrictions put on the design, the lower the probability of success are.

My main issues are, I want a kilowatt of output power, it must be able to eat my guitar playing at full crank, and the biggie......I must be able to carry it up the basement stairs, and I'm 66 years old. It will use TV tubes with plate caps and mosfets to drive them, because both are well known measures to success at this power level.
 

Attachments

  • P2330991_x.jpg
    P2330991_x.jpg
    411.1 KB · Views: 155

PRR

Member
Joined 2003
Paid Member
....the manual doesn't have:
Plate characteristics as a function of postive G1 voltage
G1 current requirements as a function of anything
G1 positive voltage operating limits: voltage, current, power...

KT88/6550 are big 6L6.

6L6 has positive grid curves.

Rough guess is that '88/6550 grid will pass twice the grid and cathode current of a 6L6 at the same point.

6L6 does not sport "limits" on grid abuse, but does have an AB2 suggested condition. Assume it is about as extreme as the developers thought was wise, and the big tubes will take more but maybe not double.

All these "Receiving Tubes" are mere toys. If you must flog a vacuum, use Transmitter Tubes. (And not the puny 807 or 6146.) HK257 (4E27) will make hundreds of Watts.
http://www.mif.pg.gda.pl/homepages/frank/sheets/114/h/HK257B.pdf
http://www.mif.pg.gda.pl/homepages/frank/sheets/088/4/4E27A.pdf
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.