SSS KT88 schematic and info?

[Luthierwnc]

Great stuff!

It might be a while before I get to this project. I've got four guitars to get out first. That might be for the best since there is a lot more I need to discover about this amp. SSS amps seem to be one-offs with most of the documentation on the 6L6 versions. The driver tube is a mystery. The voltage and resistor variations between 001 and 002 schematics here on TAG are substantial.

Thanks to everyone for their input and good luck with your own projects. sh

[talbany]

Yes, that is the article I was referring to.

I am looking at the plate curves (on KT-88 ) and am unclear about what you are talking about..Could you show me on the curve what you are thinking?

If you pick 550V for plate voltage, and then a load line that stays within plate dissipation limits, say 4 or 5k, the knee in the 0V Vg1 curve looks well positioned at 300V Vg2. If you run the screen voltage up to 550, the whole family of grid curves moves up and to the right, and the load line will cross far below the 0V Vg1 knee. That means a loss in headroom and the screen current will begin to shoot up long before Vg1 is driven to zero. Basically this type of tube seems to be capable of delivering current far beyond its plate dissipation limits when the screen and plate voltages are similar.

The main problem I have with plate curves are they are a
general guideline for a specific operation, mainly Hi fi...The curves on a guitar amp will look much different depending on how you set it up! (Where is the tube biased!!,, are you taking into consideration screen current, how much current are the screens drawing at full power 5% 8%?..etc,etc) which we may not even be operating close to (in our guitar amps)..Example:..When we bias our tubes we normally do not take into account screen current ..

The Svetlana KT88 curves are not a very complete set since they don't show screen current in pentode operation. However they convey a lot of information about the tube, regardless of where you choose to operate it. You have to scale them if you want to use a different screen voltage, but the bias point you choose has no effect, and screen current is very low at idle conditions.

We can take for example the Marshalls loaded with 6550's that ran there screen voltages close to the plates 500V ( beyond there limits) and still survived ..How many 6L's that run 470-480 with screen supplies close and do just fine (with screen dissipation @ only 6 watts)..How many Deluxes run there plates at 430V on 6V and last years..So IMO.. Why would a much more stout tube like a KT-88 (42w Plate/8w screen dissipation) have any problems running screens at 500v?...

6L6 at 480V plate and screen is a much different situation than a 6550 at the same voltages. With a 4k load the 6L6 is inside its plate dissipation limits, and the load line goes right through the knee of the 0V Vg1 curve. That tells me that the screen dissipation will not be a problem. For the 6550, there is no problem with plate dissipation, but since the load line (even 3k4) crosses well below the knee (a little more than half-way up). The screen current is going to take off long before the grid voltage goes to zero, and that will be very hard on the screens.

Martin
Thanks for taking the time!..

Your right 550 is pushing it!..Many of those old Marshalls that ran in the 560 range and the only thing that saved those KT-66's/6550's were they were well built and those Marshall power supply's sagged heavily... IYO can a well built 88 (and accounting for PS screen sag) handle 500v on screens

KT-88 (42w Plate/8w screen dissipation) have any problems running screens at 500v?...

Thanks again!
Tony

[martin manning]

Your right 550 is pushing it!..Many of those old Marshalls that ran in the 560 range and the only thing that saved those KT-66's/6550's were they were well built and those Marshall power supply's sagged heavily... IYO can a well built 88 (and accounting for PS screen sag) handle 500v on screens?

With a lot of screen voltage sag (independent of plate supply voltage), the screen dissipation problem goes away. If you set up a KT88 at something like 530 plate and 500 screen, you would want some large screen resistors to bring the screen voltage down 150-200V when the plate voltage (not B+) is pulled down to 100V or so at high power. The large screen resistor brings the 0V Vg1 curve down and keeps the screen current from rising (back to my comment on page 1). How large? I can't say without a better set of data.

If the plate voltage and load line are such that the plate dissipation limit will be exceeded, then the saving feature is B+ sag, and some Marshalls depended upon this. KT66 running at 500V (and being a 25W Pa max tube) needs a lot of that. If the whole supply sags, screen and plate voltages sag together, so that needs to be factored into the screen voltage sag (or lack thereof).

[Luthierwnc]

In surfing the net for KT88 examples, I've found most of them are ultra-linear. Without a Dumble or accurate clone to reverse-engineer, I'm inclined to go with the herd. At least, it seems a lot easier than having to use separate windings. Off-the-shelf UL trannies usually have screen taps at 40% like this candidate:

http://edcorusa.com/p/458/cxpp100-ms-4_2k

Here's what I've got so far. It is a collection of TAG schematics with some cutting and pasting on Paint.

[martin manning]

Not too many UL guitar amps, but Marshall's Major is one with a fan base. I'm going to keep thinking about pentode operation and some way to do a power supply with high plate and low screen ala SVT. I picked up a pair of EH KT88's to experiment with (they're cheap), and found the EH data sheet.

[Luthierwnc]

I started fiddling with amps in the late '90's. At the time, the popular mods were blackfacing silverfaced Fenders because CBS made them too clean. The final straw was bumping the volts up and using UL transformers to get 70 watts out of a Super.

If the object is to get big and clean, the choices seem to be UL, a separate power supply or a big, hot resistor. The power transformer I mentioned earlier has two HV windings so that is still in play. I really don't know much about calculating voltage changes based on the tap point. Most store-bought OTs are at 40% on either end of the coil. I do know Dr. Z uses them a lot. Some of the Leak amps start in the low 500 volt range with UL taps and a relatively small screen resistor.

[talbany]

Not too many UL guitar amps, but Marshall's Major is one with a fan base. I'm going to keep thinking about pentode operation and some way to do a power supply with high plate and low screen ala SVT. I picked up a pair of EH KT88's to experiment with (they're cheap), and found the EH data sheet.

Martin
Looking over those curves and doing some more research that you are real careful (extra filtering larger screen resistors lower bias feeders.increase load and plenty of sag etc,)these things you might can get away with it..But these really need to be run UL! like you say..It's really a mystery to me why the advertise these as guitar amp tubes (worse for SE) when they need to be run UL ..It's also a mystery as to why I was able to run these in a Traynor YBA!-A (560V 5k load line)for almost 10 years (hard at times)and the JJ's in there still work today...I also wonder if some of these tube distributors update these curves or if they just pick some old data sheet and say here you go!..They all publish different specs!..IMO I don't care at all for the sound of a guitar amp run ultralinear (or 100v below screens) ..Look at all the silverface UL amps that are real sterile great for Hi Fi and bass guitar but for us guitar players who rely on output tube distortion as a big part of our tone UL operation is not the same ..I hate them..IMO in running more ultralinear I would want to use the CF driver to try and keep the output section as clean as possible (would still sound Hifi-ish) to avoid that nasty kind of breakup you get running UL into saturation (which he might do in a 2XKT88 amp and doubt he would in a 200w Marshall Major type)..Thought this was worth mentioning..BTW..you could also use an OC3 to regulate and drop the screens 100v
http://www.captain-foldback.com/Leslie_ ... cs/257.gif

Anyway thanks again for the insight!!..I still have allot of questions to be answered so let us know what you find out..I'll do the same!!
KT88's are really rather useless in a guitar amplifier (especially SE)
they are like this:
Vp Max/ V Max Ultralinear & triode (if listed)/ Vg2 Max
Wp Max/ Wg2 Max

Sovtek/Electro Harmonix
800v/440v/440v
42w/6.6w

JJ
800/600/600
42w/8w

Teslovak/Shuguang/Penta
800/600/600
50w/8w

Tesla
800/600/600
35w/6w

GEC (older)
800/600
40w/7w

GEC (newer)
600/600
35w/8w

Genelex
800/600/600
35w/6w

Misc Soviet
600/600
35w/5w

SED (older)
800/600/600
42w/6w

SED (newer)
820/610
44w/10w

So ... who is lying?!!

All The Best!

Tony

[Luthierwnc]

This wouldn't be the first time I'd gotten the right answers to the wrong question. Maybe a better start would have been to ask; "If I want to make an SSS light in the 70-80 watt range and still retain most of the unique voicing, what would be my ideal power supply and power amp?"

Cheers, sh

[talbany]

This wouldn't be the first time I'd gotten the right answers to the wrong question. Maybe a better start would have been to ask; "If I want to make an SSS light in the 70-80 watt range and still retain most of the unique voicing, what would be my ideal power supply and power amp?"

Cheers, sh

SSS used both 6L's and 6550's...6550's ran more UL (hi Fi sounding) than 6L's.. Your not going to retain much of the unique voicing because the originals used 4 tubes, your using 2 tubes!.. I have not built a 2 tube version just a 4X6L version..It would make perfect sense if you were doing a 2 tube 80w version the best way would be to use the same tube Dumble used 6550's (either way you have to drop the screens)..Nobody really knows how Dumble did that so were are trying to help give some ideas as to how to go about it.. If your going with the 6550's/KT-88 I would definitely go with the CF driver in your SSS Lite!!

Tony

[ToneMerc]

This wouldn't be the first time I'd gotten the right answers to the wrong question. Maybe a better start would have been to ask; "If I want to make an SSS light in the 70-80 watt range and still retain most of the unique voicing, what would be my ideal power supply and power amp?"

Cheers, sh

I did a small special w mid pot, CF-PI and pair of KT88's at around 490V. I did not want the bias off the HT winding so I had the PT wound with a dedicated high voltage, high current bias tap and I also added a 2W pot for the CF balance. One day I will totally finish this project.


TM

[Luthierwnc]

Thanks all,

It will be a while before I could get to this project so I think I'll approach it then with a fresh pair of eyes. Most of my amps use 6L6s so the KT88 idea seemed (still seems) like a challenge. When I got started I didn't know only the 6L6 versions had been documented.

Most of the KT88 schematics I've found so far are UL. A significant number use a secondary power supply and some just treat them like ordinary tubes with much higher screen voltages but maybe not stratospheric plate volts. In the meantime, I'll keep my ear to the rail for KT88 discoveries here on TAG.

Cheers and good luck with your projects, sh

[Max]

AFAIK ODS 150W #121 is a version of a 150W SSS with an OD-circuit and a tremolo-circuit instead of the filter-circuit. And if you take a close look at the pictures of #121 posted here http://www.maverick-music.com/vintage-a ... todd-sharp you'll probably find a picture of the back showing the idle current of each of the two pairs of 6550A, and pictures of the PT and OT, and a list of all the tubes used in the circuit of this amp.

Cheers,

Max

[martin manning]

Good stuff, Max, thanks! I wonder if the OT drives the reverb springs inductively ;^)

Tony, thanks for the summary of KT88 ratings. Data sheets are sort if funny in that different manufactures often show exactly the same numbers, just in different formats. I've heard that was per some agreement that they were all really aiming for one standard spec for each particular tube type. Maximum ratings is where they seem to differ a bit.

Looking at JCM800 with 6550 running ~465V plate and screen, this is the situation I would avoid. It's like running the wrong Zpri by a factor of two. The tubes survived (1k screen resistors helped), but I don't think those amps sounded all that good cranked up, not as good as EL34's anyway.

[Luthierwnc]

Tony and Martin,

Yeah, when you think about it, people with US-sold Marshalls back in the day couldn't wait to put in EL34s. My sketchy understanding was that the US distributor(s) didn't want the service headaches from EL34s so they used 6550s sacrificing tone for reliability. That could be totally bogus but I do remember being told by a couple of good players that it was worth losing warranty coverage to trick them out right.

I've got another thread going about a standalone preamp and received some expert comments that the loss of preamp sag would hurt the power amp tone. I hadn't considered that before but it makes sense and also bolsters the case that when it comes to pure guitar amps, some power tubes are horses and some are mules. sh

[talbany]

Maximum ratings is where they seem to differ a bit.

Funny with those curves you posted the screen voltage ratings are useless!! and a SE amp with a flatter load line.

I couldn't agree more on the tone of those 6550 Marshalls!!..Ass would be an understatement..And IMO breakup even worse run UL!..I doub't seriously the Marshall Major crowd (Page,Blackmore) never really ran those things into output distortion..Blackmore push the front end with a Akai pre and Page played fairly clean (check out song Remains the same)

BUT>> The Heavy metal guy's love them!!..Metalica paved the way for that tone in the early 90's using those amps and now with the BIG 6550 Sunn Amps making a comeback in that genure of music run UL (Talk about ugly!!) with that crowd the Nastier the better..There is just no way to define tone..I shake my head build it and Shut up

Tony