Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: RicharD on October 29, 2011, 02:38:53 am
-
DL gave me this set of iron from an old Baldwin organ. It's been sitting around here for a year. Originally I was gonna build another Stereo Stout, but the love affair is over. The Stout is a great circuit but how many of these do I need to build? So I decided to ponder a Hi Fi amp. Well right off the bat it's PP + organ OPT's so I dunno if Hi Fi is even a fair term to use. Furthermore, I'm gonna hafta tube rectify so I'm calling it Mid Fi at best. Even with the tube recto, I'm expecting my voltages to be a tad on the high side since this is old 115V iron. If I've crunched my number right, I'm idling at 260mA which is pushing it for a single 5U4GB. IIRC, the Baldwin chassis had 2 recto tubes. I'm hoping DL will verify this. PSUD2 says I have enough filament to drive 2 recto's. I selected the 6DJ8 based on gain. Looks like I need about 28V to drive the 6GK6 to full power. I'm getting a gain of about 25 through the preamps so 1V in is getting me close to full power. That seems about right to me. I have not bread-boarded this yet. A peeny for thoughts. I may just be spinning my wheels..... again.
-RicharD
-
I'm calling it Mid Fi at best.
Some of my best musical memories are listening to 60'/70's stuff on an old "mid-fi" console set......somehow LedZep, Springsteen & Mott The Hoople don't sound quite "right" on digital. :dontknow: I like the idea.
and, I like the 6GK6's....I've got a bunch of 'em, as well as the 16v heated 16GK6's
For the recto (if you can't go with 2x 5U4's), how about SS (or GZ34) and dumping unwanted volts via a string of 5w zeners on the CT? I've used this & it works really well. You can breadboard a long string & tap along the line anywhere to get the exact voltage you want, then you'll know exactly how many to mount (tucked away in a hidden corner of the chassis) in the final build.
-
I'm pretty certain I can go dual 5U4GB's. I really don't like blowing off 40 or 50 volts raising the center tap. That puts DC back on the PT. Yes it works, but it's not ideal. I'm still thinking about the circuit. I should prolly go ahead abd bread-board it and see what comes out in the wash.
-
You think you have NFB but C1 shorts it out.
The transformers are fine, probably better than most "Hi Fi" gear.
I don't see why-not a 12AX7 in front. More gain, more NFB.
-
I don't see why-not a 12AX7 in front. More gain, more NFB.
richard, use a 6EU7 or better yet, 6AV6, i have a bunch of them - save the 12ax7's for geetar amps.
--DL
-
DoH on the NFB! That changes everything. Now I see a 12AX7 in this thangs future. I'll probably use a 12AU7 for the PI. I didn't get to mess with it yesterday because I cleaned my room instead. You know how you can never find an IEC power cord when you need one? Well that's because there were 7 of them under my desk not in use. There was a total of 28 derelict cables under my des, everything from USB, to wall warts, to RCA's & XLR's. I am a SLOB!
-
Attached is the revised schematic using a 12AX7 for V1 and a 12AU7 as a split load PI. The circuit appears to have just enough gain. I can feed it with my Android telephone and kick it up full w/o too much distortion, based on my naked ear test. As I suspected, B+ is a tad higher than calculated, but current is a little less so so it all balances out to about 10W per tube. I did run with parallel 5U4GB's. The 5V isn't drawing down at all so I feel comfortable that it's gonna be fine. I ran a quick response test and it does roll off at 100Hz but I did the test with the speaker connected and not a dummy load so that could be part of it. All in all it works as expected. The mother-ship didn't land. No need to race out and build this circuit. Mid Fi at best.
-
Shouldn't be down at 100Hz. Is the OT ultra-tiny?
To get tolerable bass THD, a P-P OT usually has enough inductance to pull 20Hz at _low_ level, even if the goal was only 50Hz.
Your NFB isn't. Rfb:R5 22K:100 programs for gain of 220. Amp forward gain is only 50-60. You probably meant for Rfb to be 2.2K, gain near 23, which is about 6db NFB.
V1 V2 are running much richer than necessary.
R10 R11 are half-meg. Cathodyne loads needn't be 1/2 or 1/4 of that, 250K or 100K each; 15K is far into diminishing returns. In fact you will have more max swing with 47K-100K per load than with 15K (however the max swing is already far in excess of requirements).
Cathodyne input is infinite-enough (except stray wiring capacitance from one socket to the other), so V1 load could be much higher to eke out a wee bit more gain; 270K or 470K would be typical at R3.
-
something to ponder - i know you have a crap-load of 6AN8. if you don't i do.
http://www.freeinfosociety.com/electronics/schemview.php?id=715 (http://www.freeinfosociety.com/electronics/schemview.php?id=715)
-
I did fiddle with this circuit some more the other night, making PRR's suggested changes. The little additional gain seems to help. The circuit runs stable w/0o any NFB, but I must be doing something wrong because 16k/100 (160)is about the sweet spot. Any greater feedback and I don't have enough gain. I didn't really have the time to look at the response. I did grab bias voltages. I gotta run out the door right now. I'm gonna mess with it some more this weekend. The 6AN8 seems worth a try.
-
Had a few minutes to play around. 4k7 NFB sure does clean it up but then my Android can only push it up to 1/2 power. I'm thinking drop the 12AX7 in lieu of a pair of 12SJ7's.
-
6AU6 into 12AU7 SLPI. bypass the heck out the 6AU6. run 6AU6 around 150V B+.
--DL
-
Well I have a mother-load of 12SJ7's and I hit the lab before I saw the 6AU6 suggestion. Of course it's too late to run the volume up but my initial results are quite improved using the 12SJ7's. TubeCad says I was making a gain of about 80 with the 12AX7. I used the RCA chart to design my SJ circuit for a gain of about 160. My NFB is 16k/100 or 160. Is that a coincidence or did I stumble onto something? More NFB = less distortion to a point yes? I'm trying to understand the balance between gain and feedback. I think this is really important, or I'm lost in my own little corner of the world. My new circuit has roughly double the forward gain but the same feedback ratio as before but is MUCH less harsh. It also appears I can now make full power using my cell phone. I'll find out in the AM.
-
> my Android can only push it up to 1/2 power.
One half-12AX7 of gain into sensitive power tubes SHOULD be enough gain for a power-amp. I had 150W amps on that topology. Respectable NFB. Input sensitivity OTOO 1.5V.
Perhaps this "Android" is lamer than a conventional hi-fi "line level"?
-
>Perhaps this "Android" is lamer than a conventional hi-fi "line level"?
I've been less than scientific when examining the output of the Android, but I think your statement is true. Running it up full blast it get's pretty nasty. I question it's overall fidelity. I think I'm using a lofi source. Might be worth my time to generate a test tone file, load it into the Droid and measure it's output. Regardless the circuit is performing acceptably with the higher gain front end with a FB ratio of 16k:100. I'm considering revamping the PI using a pair of 6J5's, bringing the tube count up to an even 10. Octal pentode gain stages, octal split load PI's, 9 pin outputs, and a pair of 5U4GB rectifiers. Right now it looks very cool on the breadboard.
-
Octal pentode gain stages, octal split load PI's, 9 pin outputs, and a pair of 5U4GB rectifiers.
woolly bully! - big guys (octals) are pushing the little guys (9pin) around...
--DL