Great points, but let me add this: 95% of the worlds OTLs cannot survive longer than perhaps 10 minutes in a real full power sinewave test. Plates and currents are as a rule driven well over the maximum ratings, taking into account the high peak/RMS power ratio of audio signals and thermal inertia of relatively large plates used in typical OTL tubes. There are alo sother problems but I won't go in depth here.
Some time ago I've done an extensive analysis of a class AB OTL circuit (a circlotron - but the data applies to practically any PP topology).
Rare tubes are definitely not for use in OTLs unless you want one that wil last a short time or one which is very low power.
Basically, amongst triodes you have the following to chose from:
6S33S - it's biggest problem is that it's heaters dissipate more than it's plate, leading to a lot of heat in a small space. To boot, the plates are massive so the structure holding them mates directly into pins which transfers plate heat into them very efficiently. Finally, they are almost impossible to match, which is needed for parallel operation. The only viable alternative are separate drivers with adjustable static and dynamic balance, fortunately 2-3 tubes in parallel will get you to usable power output so having 3 drivers is not such a huge additional cost to pay for proper load distribution and extended tube life.
6S19P is cheap, and 4-5 in parallel are an equivalent of a 6S33S. The advantage is that heat can be distributed better and the sockets do not suffer as they do in the 6S33S case. Disadvantage: again, balancing. The low price helps, and they seem to be fairly stable with age so selection for balance is possible.
6S45S is about 40% of a 6S33S but uses the same socket which compensates for it's lower price. The higher Rp also does not help.
6AS7, not to be confused with 6080 (see below) - two of them in parallel are roughly equivalent to a 6S33S but with about half of required heater power. The russian equivalent is slightly lower spec than the original but in this case thios is a bonus as it makes paralleling easyer. The 6080 is generally less suited for an OTH precisely because tubes are generally driven well over spec. 6080 does not have the huge G1 cooling fins of the 6AS7 and wes it will go into runaway and go poof. Unfortunately. Unless some means of balancig is provided, and since 2 6AS7/680 are really 4 triodes, and realistically for 20W or so you would use 2 banks of 4 tubes, doing this with separate drivers is usually well out of the question. But the fact remains that out of all commonly available tubes the output power vs heater power vs number of tubes equation is actually the best, followed only by the 6S19S. Cheap octal sockets, easy distribution of heat and fairly low price are a bonus - but prices are going up.
Running these valves with low ratio OT brings the best of both worlds. The protection offered by an output transformer is essential for rare vintage speakers. The ratio makes excellent performance possible at both bass and treble extremes. Power is far more efficiently used and less tubes are needed to achieve it.
A good 6-7watt PP amp can be made with just a single bottle of 6AS7 running at 100mA and 100V across the tube. I would guess it could easily match the performance of most OTL amps.
Shoog
A good 6-7watt PP amp can be made with just a single bottle of 6AS7 running at 100mA and 100V across the tube. I would guess it could easily match the performance of most OTL amps.
Shoog
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.