I am going to be building two identical tube amps, a variant of the Dynaco ST-70. Each amp will have two channels that need to be somehow combined to power one channel, i.e., one amp will be used on the left speaker and the other for the right speaker.
I think I have two options for wiring each amp to power one channel each:
1. I can parallel an amp's outputs. The speakers are 8 ohms so I would wire each amp channel for 16 ohms. The amp would be operated in mono mode.
I have been told that I could run into output transformer non-linearity issues (not sure what that means). I have noticed that McIntosh tube amps can be operated in this manner. Any thoughts on the merits of doing this? Problems?
2. the other option I have is to vertically bi-amp the speakers. One amp for each channel, with one of the amp's outputs powering the tweeter and the other the woofer. The tweeter input would be via the passive crossover. Given the high impedance of it, I was told to connect a non inductive 8 ohm resistor across the terminals of the tweeter so that it matches the amp. The woofer would be powered directly by the other channel of the same amp. The second amp would similarly power the other speaker. I've seen references to the highs suffering using this approach. Any thoughts?
Which of these would be best? Are there any other alternatives?
Appreciate any thoughts you may have....
I think I have two options for wiring each amp to power one channel each:
1. I can parallel an amp's outputs. The speakers are 8 ohms so I would wire each amp channel for 16 ohms. The amp would be operated in mono mode.
I have been told that I could run into output transformer non-linearity issues (not sure what that means). I have noticed that McIntosh tube amps can be operated in this manner. Any thoughts on the merits of doing this? Problems?
2. the other option I have is to vertically bi-amp the speakers. One amp for each channel, with one of the amp's outputs powering the tweeter and the other the woofer. The tweeter input would be via the passive crossover. Given the high impedance of it, I was told to connect a non inductive 8 ohm resistor across the terminals of the tweeter so that it matches the amp. The woofer would be powered directly by the other channel of the same amp. The second amp would similarly power the other speaker. I've seen references to the highs suffering using this approach. Any thoughts?
Which of these would be best? Are there any other alternatives?
Appreciate any thoughts you may have....
IMO, paralleling the sections of a stereoblock is a prescription for heat and bad sound. 
The vertical biamp technique, where each amp section drives only a portion of the speakers works. However, you get finesse, not "grunt".
The way to get all the power the amps can deliver into the speakers is by bridging. You drive the sections of a stereoblock with 180 degree out of phase signals and connect the speaker cables to the "hot" terminals. 8 Ohm speakers get connected to the 4 Ohm taps.
The vertical biamp technique, where each amp section drives only a portion of the speakers works. However, you get finesse, not "grunt".
The way to get all the power the amps can deliver into the speakers is by bridging. You drive the sections of a stereoblock with 180 degree out of phase signals and connect the speaker cables to the "hot" terminals. 8 Ohm speakers get connected to the 4 Ohm taps.
Merlot said:
I don't see any need for the parallel resistor, nor should the highs suffer. I'm assuming a typical parallel crossover network. If the tweeter/network combination has a relatively high impedance, it will just make the load line for the output stage more shallow, which typically means lower distortion. Try it.
Sheldon
Is it a Piezo tweeter or something? I can't think of any other reason that a parallel resistor would be needed or advisable.
What I would do is disconnect the passive crossover entirely and use an active crossover before the power amps. That would give the most flexibility and dynamics IMHO.
If you do use part of passive crossover you will have to careful about how you tap into it. Also note that this kind of pseudo biamping still burdens the tweeter amp with amplifying the bass (at least the voltage aspect of it).
What I would do is disconnect the passive crossover entirely and use an active crossover before the power amps. That would give the most flexibility and dynamics IMHO.
If you do use part of passive crossover you will have to careful about how you tap into it. Also note that this kind of pseudo biamping still burdens the tweeter amp with amplifying the bass (at least the voltage aspect of it).
Merlot said:
You have a 3rd option. Since you're building, buy half as many output transformers (2, not 4) of lower impedance and just make two monoblocks. EL34's aren't that hard to drive, even if you wanted to use some MKIII/MKIV modern variant driver boards, you would be fine (I know, cause I've built a similar set of monoblocks).
You'll end up saving a significant chunk of money and your result should be more listenable.
A nice try, but the ubiquitous common ground makes the idea a non-starter.
There's more than 1 way to tackle the problem. You could build a bridging adapter that wires each section of an ECC99 or other "super" twin triode as a "concertina" phase splitter. If your line stage has a VERY low O/P impedance you could use a 1:1+1 transformer to split the phases. Sowter in England offers their model 8920, which is quite suitable. A definite benefit of using a trafo to split the phases is the galvanic isolation obtained. Ground loops get broken up.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.