A question about LM317 cathode bias and new guy rambling

[mesherm]

I guess I'm a tube newbie although I have built many many kits starting with a Knight-Kit 12W transistor stereo amp in the 60's.
I have mostly stuck with solid state (except for an old Heathkit FM tuner kit and home brew FM stereo demodulator) so I have to re-learn all the tube stuff I thought I would never need again.
Having burned out on solid state amps but still needing a periodic dose of flux fumes, I bought one of those 11MS8 tube amps from S-5 Electronics. I figured not much to lose for the price they were asking. I assembled it stock but was not pleased with it mounted on a piece of wood so I disassembled it, tossed the wood and mounted it in a Lexan case I built. Well it came out so nice tube fever struck me. I searched the internet and ebay for weeks checking out what was available to feed my new habit. I finally decided on Shannon's ST-35 board as a good next project since I didn't want a complete kit (I wanted to make another bigger amp again enclosed in Lexan). I have the RevD board, Uncle Ned's trannys, and enough of the Mouser parts to warrant dragging my trusty American Beauty 33W iron from the cupboard again.
Enough of the preface, now to my question.
I am intrigued with the LM317 cathode bias thing and am not that much of a tube purist to resist using a few solid state goodies. They didn't have those 3 term regulators back when I studied tubes. I noticed the board has the footprint for mounting 3 term TO-220s, so is it as simple as leaving the bias resistors off and mounting a 35.7 ohm, 1%, 1/2 w resistor in the spot where the 10 ohm would have gone?
BTW, I am kicking myself because at work we used to have two Tinius-Olsen test machines which utilized two small 3 tube servo amps each. They used a 6X4,12AX7,and a 6BQ5 each. I had to keep them running so we replaced the tubes on a regular basis working or not. When the machines were finally decommissioned about 15-20 years ago due to age and wear, I tossed all the used and unused tubes I had in stock to make room in the parts locker. I figured I would never need them again and if I did they were not that expensive. So out in the trash went about 6 sets of unused RCA and Sylvania 6X4s,12AX7s, and 6BQ5s. What an idiot. I have scrounged the equipment graveyard at work and found an old Fanon 100W PA amp with the original tubes and a collection of new and used tubes (mostly 12AX7s,12AU7s) with at least a half dozen 6AV5s. I know there were two vintage tube bench power supplies (0-300vdc, 0-150vdc, 6.3 and 12.6vac outputs) at one time. Looks like I'm already scheming the project after. I was thinking something like an SE amp using paralleled 6AV5s since I have so many. Hmmmm...

[Shannon Parks]

I am intrigued with the LM317 cathode bias thing and am not that much of a tube purist to resist using a few solid state goodies. They didn't have those 3 term regulators back when I studied tubes. I noticed the board has the footprint for mounting 3 term TO-220s, so is it as simple as leaving the bias resistors off and mounting a 35.7 ohm, 1%, 1/2 w resistor in the spot where the 10 ohm would have gone?

The specs of the ST35 are heavily skewed towards Class AB, so I'm not sure what performance we'll get limiting the ST35 to constant current. I would swag that it doesn't work as good as cathode resistors. But I have this unit on my bench right now in order to test Ned's new Fender Deluxe transformer (with UL taps). Hang tight, and I should be able to do this test pretty easily.

Now, I just tested a trioded version doing just this - replacing the four cathode bias circuits with four LM317's with a single set resistor each (37.5ohm). I lowered the B+ by making the unit choke input, using the open bracket 4H Hammond choke. Performance seemed fine. The Deluxe trannies actually bested the Z565 in this configuration. !

So out in the trash went about 6 sets of unused RCA and Sylvania 6X4s,12AX7s, and 6BQ5s. What an idiot.

I'm an idiot for buying a Peavy bass stack in '89 instead of Microsoft stock. I seriously debated the issue. :sad1:

Shannon

[EWBrown]

I've built a couple of the K502s (from AES, same as the S5 amp), and came up with some good upgrades. First, replace the 6 coupling caps with Wima 0.15 uF / 400V caps, then replace the OPTs with Hammond 1609s (or 1608s, either works perfectly). In order to fully tap the Hammond OPT's potential (pun intended), cut the etches for the SGs of the 11BM8s and connect the SGs to the UL taps.

I used an amuminum chassis, instead of the wood block, for better grounding and mechanical strength.

With the Hammonds, the brown and blue are reversed from the original UTK OPT else the amp turns into a power oscillator :oops:

The original OPTs primary are about 9600 ohm CT to 8 ohms - the 1609 is 10K CT, the 1608 is 8K CT. Next, the power supply caps need a bit more
capacity, I paralleled a 470 uF / 400V cap directly actoss the 220 uF 250 V
cap. Lastly, I installed an ICL (CL-90) inrush current limiter in series with the primary / fuse / power switch, to allow for a softer start-up.

This may be a good candidate for Ned's soon to be released low cost 6500 ohm UL OPTs (I'll try once they become available).

The resulting "work of art" can be seen here:

http://www.diytube.com/jalbum/ed/slides ... ter%20dive)%20a.html

/ed B in NH

[mesherm]

Hindsight being 20/20 I wish now I had done more web research before I built my K502. It would be easy now to swap the caps but since I built the lexan enclosure a little small there is not much room now to put OPTs in that are much bigger. Mine is mounted on a 1/2 thick lexan base and is enclosed with 1/4 inch lexan all around. I drilled holes in the top for the tubes and used standoffs to raise the board so that the tubes just barely protrude. A small muffin fan in back blows air into the case and when operating a nice gentle warm breeze exhales from the space around the tubes. I had heard about the tubes "lighting up" at power on but mine doesn't seem to have a problem. I think i will wait awhile and see what develops in the OPT market before replacing stock ones.
I also decided to forgo the LM317 cathode bias scheme on my ST35 and use the resistors. I plan to use other coupling caps than the orange drops though. Still waiting on some back ordered parts from Mouser. Thanks for the upgrade info on the K502.

[Shannon Parks]

But I have this unit on my bench right now in order to test Ned's new Fender Deluxe transformer (with UL taps). Hang tight, and I should be able to do this test pretty easily.

Nothing is easy it seems! I remembered that I put the 100 ohm 'triode' resistors under the PCB, requiring me to remove the whole thing to do this test. Alas, I will one day, but testing the black Fenders are my top priority. I've already done some testing with them, and they look promising as a budget route.

Shannon

[EWBrown]

Those new UL trannies from Ned are real tempty... I have a couple of the
non-UL ones, for a cheap axe amp project.

I see that AES is 10% off everything until 9/30.

/ed B in NH

[Shannon Parks]

Those new UL trannies from Ned are real tempty... I have a couple of the non-UL ones, for a cheap axe amp project.

Ed,

Seriously, in just straight head-to-head versus the Z565 in triode mode, the non-UL Fender looks to be an excellent performer. Sure, we're just talking 3W a channel, but the numbers looked great in comparison.

The UL model is a bit trickier, and I'm trying to test it properly. The numbers won't nearly compete with the Z565 at higher outputs, but for the low cost, it could make a very pleasing amp.

Shannon

[EWBrown]

I was thinking of using either 6AQ5 / 6005s or 12AB5s, with the Fender ULs, (I have lOTS of each type) and a 12AT7 or 12AV7 for the driver / phase splitter, essentially using a circuit similar to that in the K502, which is pretty generic. THe 12AB5s are a "poor man's" 6973, and can be found for cheap (I just landed 15 for a buck apiece, including shipping, on E-bay. THe 6005s were freebie at the NNETG meet back in june. Now all I need is a good low cost power trannie with either 6 or 12 VAC filament, and 180-200 VAC @ 150-200 mA secondaries, (looking for 225-250 VDC B+) and it'll be a done deed.
I suppose a 100VA isolation and separate filament trannie would work, using a voltage doubler on the 120 VAC secondary, but that entails extra iron and less available real estate on the chassis.


/ed B in NH