5687 to 6L6 SE ?

[rhcclark]

Hi
im thinking of building an se stereo amp with what I have on hand which is 5687s, 6w6s and 5932s (which are these really cool sylvania military 6L6s I believe) also on hand are a pair of edcor 2.5K SE OTs with a 40% tap and another pair of the same except 5k. I was thinking building something like this

http://tinpan.fortunecity.com/saints/668/primer/6l6-se.gif

but can I drive a 6L6 with half a 5687? Or maybe one paralleled?

any suggestions?

thanks

 

[Eli Duttman]

Use the 5 KOhm trafos. 6.6 KOhms, end to end, is commonly used with PP 6L6s. IMO, 2.5 KOhms is too low.

The mu of a 5687 section is too small, but cascaded sections will be fine. The 5687 has plenty of drive capability. So, even triode wiring the "finals" is not a problem.

BTW, triode strapping the 6L6s allows a relatively simple build. Pentode mode requires regulated g2 B+ and loop NFB for good results. UL mode doesn't need regulated g2 B+, but loop NFB of some kind is necessary. Look into Bandersnatch's "exolinear" topology for pentode and UL modes.

Don't forget the B+ decoupling filter. It's mandatory in a 3 stage amp.

 

[dsavitsk]

I was denied access to your schematic for some reason.

But, I'm in the middle of almost the same amp for someone. If you use a CCS on the driver tube, you will eek out a bit more gain. I found that with the 6w6, the 5687 had enough gain so long as your source has a highish output. It actually sounded quite good as a driver and pushed my 91dB speakers with no problem. I have ended up with a 5963 or a 12AV7 for drivers.

Here is the schematic I drew up:

The CCS is a 10m45 set at about 11mA, and by using the ccs and led for bias (there are actually two LED's in series), popping in new tubes for the driver is pretty easy.

 

[rhcclark]

here is that schematic again
http://tinpan.fortunecity.com/saints/668/primer/6l6-se.gif
it is the JC Morrision 6L6GC Ultralinear SE amp
Thanks for the responses. As my speakers (Rega R1) are only 90db efficient i dont think i will get enough power out of a triode connected 6w6. Though the little T-AMP drives them fine. I dont really want to build a 3 stage amp, though I will if I use the 5687s. I may just have to buy some ef86s

 

[Eli Duttman]

I dont really want to build a 3 stage amp, though I will if I use the 5687s. I may just have to buy some ef86s

Reports from the field regarding the recently introduced EH EF86 are favorable. New Sensor strikes again.

A 3 stage circuit using a 5687 and a 6L6 "final" has some appeal. DC couple the cascaded 5687 common cathode sections. Use "exolinear" topology and connect the 2nd gain block's load resistor to the UL tap on the O/P trafo, instead of the B+ rail. Pentode wire the 6L6 and regulate g2 B+. Take full advantage of the 5 K primary O/P trafo and use combination bias (with an unbypassed cathode resistor) on the "final". You'll get plenty of power, decent damping factor, and good linearity. Also, gain should be adequate to forego a line stage, when a CDP is the source. Use a volume control pot. as the 1st gain block's grid leak resistor, instant "integrated" amp.