by dcgillespie » Sat Jan 29, 2011 6:13 am
Again, this gets into a question we were dealing with earlier. With regards to bass guitar, is the issue one of too much gain over driving the power amplifier section, or too much distortion being generated in the power amp section when a given level of power is reached? Swapping the preamp tube for a 12AU7 provides lower gain, meaning that for a given level of power, the gain control has to be turned up higher. That may give you the impression it has "cleaned things up", but the real question is, is the amp now playing louder with less distortion because of the tube change? I would doubt that it is.
Since the intention of the amplifier is to play bass through it, that places tremendous demands on the output stage. Looking back over your effort throughout this thread, one of the things you have done versus the original design is to use a non-standard OPT. While I have no doubt the transformer you used is a good transformer for it's designed purpose, it's purpose was not for what you are intending to use it for.
No doubt (this is meant in a positive way), you looked at the schematic of the Scott, thought hey, it uses PP 6L6s with fixed bias, so its OPT will make a great OPT for the Ampeg circuit. True, it will "work", but very likely allow for a limited power output and power bandwidth of the original Ampeg design. Since bass guitar requires much more power capability from the output stage, that is why the distortion issue is showing up most with its use.
The Scott amplifier uses push-pull 6L6s true, but they are still actually operated in cathode bias mode, at reduced B+ voltages versus those of the Ampeg design, and more specifically, at a greatly reduced screen voltage relative to the plate voltage, versus the Ampeg design. All of this works to raise the required load impedance the tubes need to operate into, and reduce the power output as well -- all as versus the Ampeg design. The Scott amplifier's OPT is likely about a 6000-6500 ohm impedance primary, where as the Ampeg requires a load closer to 3800-4000 ohms. The Scott was likely a 25 watt amplifier, where as the Ampeg is likely a 50 watt amplifier, all based on the schematics you have provided. Therefore, using the Scott transformer in the Ampeg design will very likely reduce power output, and increase distortion significantly over that of the original transformer specifications. This can be confirmed by measuring the turns ratio and then calculating the intended primary impedance of the Scott transformer. The procedure can easily be found on the web.
If the concerns above are (likely) found to be true, one thing you can try is to change the reflected load back to the output tubes. This is easy enough to do, by simply operating the rated load of your speaker on an output tap that represents double the speaker impedance. That is, if you have an 8 ohm cabinet, connect it to the 16 ohm leads on the OPT. This will have the effect of limiting the effective power bandwidth of the transformer, but also halving the reflected impedance back to the output tubes, and therefore (likely) producing an operating impedance that is more appropriate for the Ampeg operating conditions. Since the Scott was a high-fidelity design to begin with, but you are now using its transformer in a guitar based application, some loss of bandwidth in the transformer can be tolerated, such that the transformer might operate much better in this application, when connected as suggested.
None of this may be what you wanted to hear, but again, I hope it helps!
Dave