OT load matching

[marsbel]

Hello,
I have an OT of unknown origin that I'd like to use in a new bassamp I'm planning to make.
It has a rating of 250VA and I measured the impedances: 2k to 4,8,16 Ohm.
I'd like to use it in a 200W bassamp preferably with 4 x KT88 because I have them already.

Now my questions:
- Is it possible to get 200W with this setup or is it a mismatch? And if so what tubes would suit better?
- What plate / screen voltages do I need? I'm going to have the PT custom wound anyway so everything is possible.


FYI: I'm an experienced EE, so I can handle the difficulties/dangers involved.
I just haven't figured out how to match the loads on PP pentodes. On the Net I couldn't find the right info. Most of what I found was about triodes in high end audio applications.
I also seem to have bought the wrong books. None of them really seem to explain how to do this.
I also like to learn how to translate plate curves to different screen voltages. Haven't figured that out either.

Regards,
Mark

[Cliff Schecht]

The first thing you'll want to do is calculate what your maximum voltage can be. This is simple Ohms law trick and some quick rearranging of the P=V^2/R relationship will give you sqrt(P*R)=V. So with a 250W transformer and a 2k primary, your maximum voltage (ignoring what the tubes can handle) is 707 V! But this may need to be downgraded for the KT-88's because each tube can only dissipate 42W max at the plate. So my quackulations give me about 580V max at the plates if you are looking for 200W. I've only designed SE amps though so I'm not 100% sure these calculations hold true for push-pull amps, I think you can actually get away with a higher voltage because the plates only conduct for half of a cycle which reduces the plate dissipation at a given voltage by a good 30-50%. Let's see what others have to say about all of this..

Do yourself a favor and read through the Valve Wizards pages on guitar tube amp design. I've also got a lot of great old texts in .pdf form that I can send or perhaps upload to the forums as well, although these don't necessarily give the information one is looking for when it comes to designing guitar amps.

[VacuumVoodoo]

KT88 data sheet, top of page 3: http://www.mif.pg.gda.pl/homepages/fran ... k/KT88.pdf gives conditions for 100W power with 2 tubes.

For 200W output power double the number of tubes to 4 and halve Raa to 2.25K, your 2k transformer will be just fine.

There are other parameters that define transformers suitability for bass amp application: mainly primary winding inductance & core size i.e. cross section area of center column. The bigger the better. A good 20+ lbs weight is a reasonable ball park size for a 200W bass application.

[marsbel]

Thanks Cliff and Aleksander.
So a B+ of 560V - 580V, Ia 60mA @ idle will get me in the ballpark.
I've seen the datasheets off course but I was a bit confused because most high powered bass tube amps (SVT, Hiwatt, Trace Elliot V series, Peavey, ...) with either 6550's or KT88's run plate voltages of 660v - 700V or even more but with lower Ia of +/-30mA @ idle.
Anyway, I'd like to do the design calculations before ordering the PT. So I guess I'll start with the datasheet values and use the method explained on Valve Wizard's site and see if it matches and then go on from there.
I'll post results afterwards.

The amp will either be built on a Bassman 100 chassis and have an Ampeg baxandall preamp in one channel and the Regular Fender 3 band preamp in the other and switcheable.
Or it'll be Orange AD-200B inspired in terms of looks and controls.
I haven't made my mind up on that.

Regards,
MArk

[Andy Le Blanc]

I did 200w some years back, I used a Hiwatt style inverter the feed back
was critical. If your going for a fixed bias and intend to individual bias/balance
each tube it can be a procedure. Three gain stage before the inverter worked
very well with bass. There's a couple marshal bass amps and the old peavey
vta series had interesting use of the gain structure, the peavey used T filter
on local feedback loops in the pre to affect tone control.

Its more a matter of getting a stable and equal plate dissipation for the set of
tube you end up using. The higher voltage set up will give you the most watts
in the end but the tone will be more analytical and even a soft clip won't be pleasant.

[fperron_kt88]

FWIW, Marshall ran the Major with a ultra-linear quad of KT88. 620-640Vdc on the plates, 1900 Ohms primary impedance.

The head weights about 90 lbs with the cab. The OT itself is roughly 30 lbs.

The ballpark you came-up with seems correct then...

The feedback loop is closed over a larger portion of the amp than one usually sees in guitar amp. The PI is in two sections: a concertina splitter feeding a diff amp.

Many folks objected to clipping this power amp (voltage rating of the OT), but mine's got a modern (circa 2001) replacement and was used flat-out for at least an hour on a bench and dummy load without issue. The original OT did burn a long time ago...

[Firestorm]

The amp will either be built on a Bassman 100 chassis and have an Ampeg baxandall preamp in one channel and the Regular Fender 3 band preamp in the other and switcheable.

Interesting. I've built that amp. Used a Fender 135 as the donor amp and kept the OT so it had the 2K primary. Also kept the PT, so between the two it can't quite get to 200 watts. One channel was James/Baxandall for that SVT mid-rangey tone; the other was 5F6A (same as the early Marshall bass amps) to make the floor shake. With those preamps the channels are bridgeable.

Consider using an ultralinear OT; works nice for bass, minimizes impedance match issues, and gooses the output power.

[marsbel]

It would be easiest to use the Fender chassis. But it was designed for 6L6GC which are smaller than KT88's; I have doubts about the tubes being too close together.
Also the transformer layout isn't what I'd like. They're too close together.
Both problems can be solved, but that would mean metalwork and welding. And then it's only a small step towards building a custom chassis.
I'd have to testfit the tubes to the chassis and see if the distance between them is still reasonable.
What would be the minimum?

MArk

[Firestorm]

It's primarily a heat issue. In my build, I used a DC fan blowing on the tubes. It was actually a computer fan so its speed is controlled by heat: when it gets hotter inside the chassis, the fan speed increases.

Since I used the 135 model (ultralinear) as the foundation, there's no choke. so there was plenty of room to mount a filament transformer to power the KTs. I don't know how you envision placing the trannies.