6au6 Tube boost

[isilme]

So first, just for an introduction, I'm a pretty inexperienced electronics enthusiast. I've built a couple tube amps and fuzz pedals and I'm decent at following schematics, but until recently I've had no idea why components go where they go. This is my first try at designing my own circuit (although it's mostly ripped off of other people's).

Onto my project. Basically, I wanted to make a Valvecaster, but all I had was a 6au6 tube. Using the Valvecaster schematic (http://www.beavisaudio.com/projects/ValveCaster/MatsuminValveCaster.gif) and a couple 6au6 preamp schematics, I designed my circuit, which you'll find below.



The circuit works. It boosts, and adds some overdrive. I just feel like it might not work as well as it could. There's some hum (not terrible), and really, it's made from so many different circuits thrown together that there's no way I ended up with an optimum design. So I thought by putting this out there for people who understand it much better than I do, I could get some help improving it, as well as provide an alternative to the Valvecaster. I'd appreciate any input.

(Not mentioned on my schematic is a LM317 voltage regulator feeding 6.3V to my heater)

[iccaros]

you could lower the resistance on the anode and tie the control grid on pin  6 to the anode. This makes it act more like a triode.

you could look at Ricks pentode boost or somethings Renegadrian has posted

I placed a 12ax7 infront of mine, gave me clean boost from the 12ax7 and then overdriving the 6au6 gave a nice smooth overdrive, but I fear too much gain, I need to measure but I think I am putting out more than line level, which makes any effect after it fart out. So I am looking at reducing the gain but keep the overdrive

I really don't like distortion through messing with the bias, which yours and the valvecaster does. It makes an un-linear design worse.. But it may be the best that can be done.
I find that its becomes farty and not distortion.

On the other hand running my mini tube version on my thread, with a valve caster after it, well I get good distortion as I do with the 12ax7 and a 6au6, but its at line level. I did add a op amp between the two so that tightens up things..

you can try a buffer before and after, bias it so it gives no distortion, and then you will just have the tube sound for overdrive, I set mine for unity.

you can see my thread here
http://www.diystompboxes.com/smfforum/index.php?topic=91745.0

[PRR]

Not a bad plan. Good first design.

C1 is probably not needed.

Rule of thumb: cathode resistor is typically MUCH smaller than plate resistor. For the usual tubes, for good gain, usually 20 to 100 times smaller.

12AX7 is often run with 100K in plate, 1.5K in cathode. 67:1 ratio. 12AU7 favors more like 100K and 5K, 20:1. 6AU6 as a straight audio amp might be run with 50K plate and 1K in cathode.

That's for good gain. Here you want some dial-back. Increasing cathode resistor will do that. It also makes the tube more linear (less "tone"). And I wonder if you ever have need of the full 100K?? I think you want values below 0.5K to maybe 5K-10K. You can't conveniently set these teeny values on a 100K pot.

I'd get a bin of brown-stripe and red-stripe resistors. Try them all: what it sounds like with 100 ohms and 3.3K ohms in cathode. Find a pot which just about covers the high resistance case yet can be dialed-down to the useful (and more loud/toneful) lower resistances.

Conversely: raise the plate resistor. Too high is not great because R4 R5 and amp load it. However you might try 330K, maybe with R4 snipped for test.

Triode wiring is worth a try. Different. Perhaps not better, but who knows?

What is your plate voltage? Both extremes: R2 at 100K and at zero ohms? Plate voltages "should" be around "half" of B+. On 300V supply, 100V to 200V, but much lower here. True, there are reasons to do otherwise. But if plate is slammed to B+ or down near the cathode voltage, output is very small and dirty. Which is kinda what you want, but you want a zone that is not weak yet not too clean. Plate voltage is a clue.

Low-low-low-voltage operation is tricky. You have Screen grid G2 going to B+. You never do this at "real" B+ voltages in voltage amps (power amps, yes). You get more gain at lower G2 voltages. But 9V is already very low, so.... Still, you might try feeding G2 from your 6V tap, or even do the conventional thing with 220K B+ to G2 0.1uFd (or 5uFd 10V electro) from G2 to ground.

At very low voltages, sneak current in G1 is not negligible. Some experimenters have used much lower values for your R1. However this reduces input impedance which loads your guitar. Others have leaked some positive voltage at G1, perhaps 10Meg from B+ to G1 (now you better have C1!).

> There's some hum (not terrible)

Is it in a box? Tube-works out in the open DO catch ALL the hum in the room, annoying on low-level signal. Everything in the audio path should be in closed metal box. Tubes usually get shield-cans, though you can wrap with foil touching chassis for a test. If you do like Triode, there a trick: use G2 as your signal plate. Ground the designated plate. The plate is the outermost element. If grounded it shields all the other internal bits. G2 works fine as "a plate", up to some small amount of power. At 9V you can't possibly pass enough power in G2 to hurt it.

What do you have for B+ filter?

Are you SURE your LM317 is _clean_? Rough ripply 9V into a 6V regulator can give you ripply 6V "DC" which is a worse buzz than pure AC would be. 10 ohms to 2,200uFd would drop 9V down to 6AU6 heater level and be more-sure to be real clean.

When tubes are worked at VERY low B+ voltage, sometimes it helps to bend the specs on the heater. 7V or 8V in the "6.3V" heater will give some more plate current, and there's still many hundreds of hours of heater-life. However high heater voltage may excite more G1 leakage and force current down, you could try 5V heater.

[vendettav]

woah, some really nice info up here. i'll make sure to check where this thread goes

[isilme]

Wow, thanks for all the great advice. After having tried your suggestions (to the best of my understanding) here's my findings.

Reducing my plate resistor to 50k took my plate voltage right down to about half of B+, and after experimenting I decided that 5k was about the highest value I'd need for my cathode resistor.

Then I tried adding some positive voltage (B+ through a 10M like you suggested) at G1, which seemed to reduce my gain slightly, but significantly reduced the hum.

I tried tying G2 into lower voltages through both my 6.3v from the LM317 and the 6v(ish) from my plate, but both increased my hum without having a noticeable difference in gain.

Some other things: B+ filter? Would this just be a big cap from B+ to ground like in a normal tube amp? I don't currently have one because I hadn't seen any on similar circuits.
Also, I tried feeding my heater with higher voltages and got bad hum, and tried feeding it lower voltages and the heater just wouldn't heat.

[gmoon]

Looks good, especially for a first tube design.

++ everything PPR wrote--higher plate resistor will offer more voltage gain (and as mentioned, there's a trade-off). Yeah, you would normally run the screen lower than B+. If it works as-is w/9V, maybe just try a 500K pot as a voltage divider.

What jumps out at me is that it's a cathode-biased design, but without a bypass cap. If you added one in parallel with R2, anything from 1uF to 100uF, there would certainly be more gain (and low end).

A cap together with a pot (25K?) would make the bypass adjustable. With a pot, I'd "go big" for the bypass cap, you can always dial it down. It's a pretty common approach, especially with transistor designs. Works with tubes, too.

[rutabaga bob]

search for 'lonewolf' on the forum

[frequencycentral]

search for 'lonewolf' on the forum

This: http://sites.google.com/site/jonandersonmn/tubebooster

[isilme]

Here's the updated schematic:



gmoon: I tried adding a cathode bypass cap, everything from .001 to 100uf and I could not hear any difference, so I left it out.
I also tried adding a voltage divider to the screen, but the more resistance I added the worse the hum got and the lower the gain got.

At this point, the hum's pretty negligible (about what you'd expect from an unshielded fuzz), and I'm pretty happy with the sound. However, now that I've started soldering, I found a new issue. My gain pot. When set between no resistance and about 500 ohms, it works perfectly. As I increase the resistance though, it picks up a bad hum. The hum doesn't really get worse as I add more resistance. It just starts at at around 600 ohms and stays until I go below that value.  Could this be an issue with my pot? I've never seen anything like it before. Other than that, the gain control works fine. Any ideas?

[PRR]

> B+ filter? ...I don't currently have one because I hadn't seen any on similar circuits.

You keep mentioning hum. Hum is not inevitable. You want to fix this before you going futher.

Unless your power supply is totally hum-free (battery), then you probably need more filtering than many pedals need.

> Would this just be a big cap from B+ to ground like in a normal tube amp?

Cap AND (choke or) _resistor_, just as in "normal tube amps".

Here's a normal tube amp.



Your circuit is sorta-the-same-job as the Champ preamp stage.

You don't use the antiquated-type power supply, you are not running 300V, and you are using a plate resistor 5 times lower than Leo did. That suggests filter resistors 5 times smaller, and filter caps 5 times larger. Since it's only 9V, and 25V caps are super-cheap, and the Champ is not utterly hum-free, spend the extra penny, up-size the caps. Suggested values shown. Not critical. (If you have a box of 2K2 resistors, use 2K2 where I show 5K and 100.)

Also note that the preamp pentode plate is run "about half" of supply voltage: 130V of 280V. Pentode screen is much lower at 21V.... this sets both gain and input overload. It is possible you "should" run your screen nearer 21V. But this is difficult with 9V power. Something to play with, _after_ you kill your hum.

[gmoon]

gmoon: I tried adding a cathode bypass cap, everything from .001 to 100uf and I could not hear any difference, so I left it out.
I also tried adding a voltage divider to the screen, but the more resistance I added the worse the hum got and the lower the gain got.

Ditto to the thought you need to chase down the hum issues before you take this further. It's very possible that the changes that increase hum are probably adding gain too, but that's masked by the hum.

RE: bypass cap--you might not hear much additional gain with such a small plate resistor (and obviously there won't be a difference if the gain/cathode pot is set to 0 ohms). But most pentode preamps with cathode bias have decent sized bypass caps (look at old Gibsons, Vox AC15 1960, even Dr.Z's Rt 66). You won't get much gain with a small plate resistor, anyway.

RE: the screen voltage--again, debug the hum. It's pretty universal to keep the screen voltage lower than the plate, and certainly lower than B+. Might not burn out your screen @ 9V, but changes will certainly have an effect. And look into adding a bypass cap for the screen as well (Merlin writes that the screen bypass cap has a larger effect on small signal pentode gain than does the cathode bypass cap).

I think you're on to something here, but it needs some refinement.

[Renegadrian]

This is somehow similar...

[blackcorvo]

I got a 6AU6 hooked up to a 9v wallwart I have here (it actually puts out around 12v instead of 9), and this is the circuit I have here:

a 250k pot in place of G2's resistor as the gain control (full gain when g2 is shorted to B+), a 4k7 resistor instead of the 10k pot (still bypassed), 68k plate resistor, 1M g1 resistor, and a 10k resistor instead of the input cap. My output cap is 33uF.

It boosts nicely, but I can't test it too loud to check the tone because its really late here and I don't wanna wake up the neighbours...

just thought of sharing my attempt.

[amptramp]

You do not show your filament (heater) wiring in your schematic.  Are you running the filament from AC or DC?  If you are running from AC, check that the filament leads are positioned well away from anything connected to the control grid.  If you are running from DC, make sure your power supply has adequate filtering because the filament on a 6AU6 is 300 mA.  At the voltages you are using, you can get a lot of capacitance cheap, so don't skimp on bypass capacitors.  In many amplifiers running with AC filaments, the filament winding is biased to about +30 volts to prevent electron emission from the filament itself from adding to the cathode emission.  The effect is small, but with enough gain in the circuit, even a few percent of emission from the heater will give audible hum.

[isilme]

Okay, I felt bad about ditching this project after all the great advice I was given, so after taking a break to work on other projects, I'm back. I took the circuit back to the breadboard and have been experimenting with all your suggestions. I implemented the power supply filter that PRR suggested, which worked great at decreasing the hum from the power supply. I also added a cathode bypass cap (33uf) which added some gain. Here's the updated schematic.


(It doesn't say it on the schematic, but all +9v are from after the power filter.)

Alright. That's the good news. Now for the bad news.

The hum is still there. The power filter removed some of it, but it's still there. I have my tube in a metal shield, but other than that all my electronics are unprotected. If it would help, I could record a clip.

Also, the cathode bypass cap made the 5k gain control useless. I now have to go to about 50k ohms of resistance before there's a noticeable drop in gain. This seems to go against PRR's suggestion as well as everything I've read about pentode preamps. This is still while using a 50k plate resistor.

I also tried putting a voltage divider pot (250k) in front of my screen again. I no longer get any worse hum by adding resistance, but I do lose gain and there don't seem to be any positive aspects to running the screen lower than B+. Is this something I just need to do to protect my tube?

Thanks for reading. Any suggestions would be awesome