Biasing a tube cricket amp

[Octavian]

So, without really anything to go on, and not having much other than the schematic, I decided to try my hand at building a little 12V tube cricket amp.

I've gotten it all together, and actually get a (very quiet) sound out of a speaker attached to it. I presume I'll get a bit more sound out of it after properly biasing it?

Here's the issue - I set my multimeter to DC volts, attach the negative lead to ground, positive lead at bias point 3 (VCC): 11.14V. Ok, so at points 1 & 2 I want ~5.57VDC, according to the schematic (1/2 of VCC). I put my positive lead at bias point 2: 11.10V. Dial the trimmer one way all the way: 11.14V. Dial it the other all the way: 11.02V. Hrm. Try a different trimmer. Same thing. Try a 1M trimmer - now I can get down to ~10.6V or something like that.

So, what am I doing wrong? Why can't I dial down very far? Is there something different about working with tubes? Never had this problem with trimmers before.

[duck_arse]

can we see a diagram of this "circuit" to which you refer? please?

[Octavian]

Details here: http://www.beavisaudio.com/projects/TubeCricket/

[GibsonGM]

I would hazard a guess that you don't have the trim pots wired up correctly, with one leg to Vcc and the OTHER TWO connected together, to the top of the tube.  You should be able to go to about 1/2 Vcc....Does your heater light up ok?    Check that your cathode resistors do have one end to ground....no ground there, no tweaking the trimpots...

I don't do much low-voltage tube stuff... I don't think you'll have HUGE results from "biasing" the plates between 5V and 12V - you're really changing the plate load resistance, not traditional "biasing" per se....you do change the load line that way, and maybe you'll get some variance in clipping the next stage... but unless you're WAY off, you SHOULD be getting signal if all is connected properly.

An error with the 386 seems the 2nd most likely place to look, after you've satisfied yourself that your tube connections are correct.  Do you have an audio probe?   If you do, you should have a 'reasonable' signal between C6 & C7.   Again after C9.    Then on to the 386....

[PRR]

> Is there something different about working with tubes?

Tubes have heaters.

(Alternate name is "thermionic device", which implies heat.)

If the heaters don't heat-up, you get zero current through the tube, zero volts dropped across the plate resistors.

> bias point 3 (VCC): 11.14V.
> bias point 2: ...trimmer one way all the way: 11.14V. ...other all the way: 11.02V.

The very slight drop you observed is due to meter-loading. (Roughly a 10Meg meter, typical for VTVMs and many DMMs.)

Does the tube *light up*? Some small tubes hide their heaters, also you are 1V shy of nominal heater voltage, so you may have to turn off all the room lights to see the dull-red glow. Alternatively leave the 12V on for an hour, the glass should be warm.

This heater lights-up because you put 12 Volts across pins 4 and 5. This is not called-out on that schematic, but the connections are shown, and are essential.

If you think you wired it correctly, wiggle the tube up a little then use your sharp meter-probes to check voltage AT the tube pins 4 and 5. Should be near 12V on one, zero on the other.

[Octavian]

Thanks PPR - I checked pins 4 and 5: a bit over 11 volts on pin 4, almost zero on pin 5, and there is a dull glow and some warmth (pretty quickly actually). I also checked my trimpot wiring, and it is correct: VCC into 3, 1+2 connected on both.

I do get sound out of it when I connect to a speaker, it's just VERY VERY quiet - like my strumming on the strings is louder than what's coming out of the speaker kind of quiet.

Will try to check with a probe soon - I always wind up working on these things after my wife goes to bed and try not to make too much noise.

Could there be an issue with the adapter? I'm using what claims to be a 12V regulated (and at +12V on the board it is at 12.08). It is tip positive, so I had to reverse the lugs I'm used to (positive to the big lug, ground to two little ones).

[Brisance]

Maybe the vacuum leaked out(yeah yeah, acually air leaked in), and prohibits proper voltage drop?

[PRR]

> Maybe the vacuum leaked out

Yes, possible. The top (maybe side) of the tube should be silver inside. Brown is a small leak (tube may work); white is a total leak (tube will not work).

Is it a 12AU7 ? 12AX7 may not work this way.

Short both cathode resistors. See if you can now trim the plate test-points down from full 12V. (Do NOT short cathode resistors on big high-voltage tubes!! Here it is safe due to low voltage.) Shorted cathode bias tends to distort-- you might have to find a compromise like 500 or 1K.

[Octavian]

So, PRR, your comment made me realize that I had accidentally picked up an ECC81 (12AT7) instead of ECC82 (12AU7). With the 12AU7 in (both are silver topped, no brown, btw), I can trim the voltage down to 8.75 and up to 11.12 on VR1. VR2 goes a little lower. Should I just use a different value trimmer? (Using 100k per the schematic now).

Still new to this - I presume from the schematic that the cathode resistors are tube leads 3 and 8? And I just connect a wire from them to ground to short? (Sorry if that seems obvious - don't want to screw anything up.)

[duck_arse]

cathodes 3 & 8, yes.

[Octavian]

Now that it's late enough I can make a little noise, I tried a different tube a BAM! I still can't really change the voltage at the VRs, but I do get some nice volume with the 12AU7, and pretty surprising volume with the 12AT7 (which, if I'm correct is an ok drop in replacement for higher gain).

[PRR]

The plan sure looks like it was scaled for 12AU7 (or even 12U7, which is not the same tube).

> an ok drop in replacement for higher gain

Higher gain in *normal* circuit. With voltages high enough to get some electrons flowing.

When voltage is very low, the "high gain" tube will be essentially cut-off,practically dead. The "low gain" tube does not squeeze the electron stream so hard, passes some current, gives some output.

An empirical rule-of-thumb says that your supply voltage should be higher than the Mu of the tube. That means 20V for 12AU7, 60V for 12AT7, and 100V for 12AX7. Yes lower voltages sometimes work, but much more fiddling is needed, and again every time you change tubes (even for another of the same number).

[Octavian]

Thanks PRR.

A follow-up: Could I (without any danger) raise the voltage? I used 1/2w resistors throughout, and my caps are all rated for 50v or 100v. What effect(s) would that have?

[PRR]

> raise the voltage?

I have to ask "WHY??"

Higher voltage gives *LESS* distortion. If that is your goal, stop starving tubes, just use TL072 gain stages with 0.01% THD (and no heater demand).

A more useful frill for guitar tone might be a knob to *reduce* plate supply voltage, to scale your distortion to your signal level and taste for bentness. A range 24V to 6V ought to cover most guitar cases.

The LM386 only needs 0.1V of signal to make full output. A happy tube amp "should" be able to do that with just 12V supply. It happens that the tubes we can get are not particularly happy at 12V supply. As you found the AT7 can't hardly wake-up, the AU7 is marginal.

The heaters and LM386 MUST stay near 12V.

The plate supply can go to 500V without endangering the tubes. The trim-pots may be in trouble at a few hundred volts, but will not be needed with "normal" supply voltages, just use 47K or 100K, and make these 2W parts if you go way above 250V. The nFd coupling caps need to be rated for full supply voltage. The cathode coupling caps won't ever see 25V.