Hi TVman, no it is not the musical machine. I did build that amp and it sounded really good but something was missing, it did not have the glorious midrange of my se amp so i tried the concertina and then the paraphase splitter and that was it. A push pull amp sounding a lot like a great SE but better, much better. Deep bass is truly amazing. Midrange still not as good as SE but really close, dynamics much better.
The paraphase schematic is in one of Norman Crowhurst papers, the input tube i use is a 6sl7 the splitter tube is a 6sn7.
I could write down the schematic but i do not think is that special. It is just the splitter and the global negative feedback which makes the amp sound great. I use a pot in the splitter which i call the sweet spot.
Besides i do not believe what works for me works for everybody, it is a question of sinergy with the rest of my audio stuff and my ears of course.
The paraphase schematic is in one of Norman Crowhurst papers, the input tube i use is a 6sl7 the splitter tube is a 6sn7.
I could write down the schematic but i do not think is that special. It is just the splitter and the global negative feedback which makes the amp sound great. I use a pot in the splitter which i call the sweet spot.
Besides i do not believe what works for me works for everybody, it is a question of sinergy with the rest of my audio stuff and my ears of course.
bigwill said:
Good for you!
Pentodes rock! I think they get too much bad press
I'm even coming around to global negative feedback, maybe I'm going insane
Yes indeed, pentodes do get too much bad press, but that's because there are a whole lot more examples of poorly designed pentode amps out there than there are triode amps.
The key to getting good results with pents is active screen voltage regulation. I did two designs (PP 807s and PP 6BQ6GTBs) and included hollow state screen regulation. That helps stabilize the screen voltage, and provides a Lo-Z path to AC ground.
Local NFB may or may not be necessary. It all depends on the harmonic content of the distortion. 807s brag some of the lowest THD figures (as low as 1.8%) however there is a lot of h5 and higher (you can see that with the Twin-T test). This leads to a really nasty dissonance that's like fingernails on a black board. 807s definitely need the assistance of local NFB to linear them up in order to sound good. (The developer of the type, O. Schade, recognized this and suggested feeding back 10% of the plate signal to the grid.)
6BQ6GTBs have higher distortion (2.9% measured) but that's almost all h3, with just a trace of h5. Running open loop, these sound significantly better. No local NFB needed, just some gNFB to take off the excessive edginess. For a type without any audio pedigree, and no suggested audio operating points (it's a B&W TV, wide scan CRT, horizontal deflection power amp) it sounds quite good, even running open loop. These can pump out nearly 40W of Class AB1 power without red plating. Those HD VTs are tough.
If you design your front end right, you won't be needing a lot of gNFB. The 807 project runs ~7.0db(v) of local NFB, and 6.0db(v) of gNFB. For the 6BQ6GTB project, there is no local NFB, and adjustable gNFB. Setting the gNFB to max (13db(v)) the sound is definitely tending to sound "solid statey", and ~6.0db(v) sounds much better. It's enough to get the woofers under control and take the edge off.
It's always a good idea to spend some time listening to a new design running open loop before deciding what needs to be done so far as adding NFB is concerned.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.