Battery Bias Mod

I just finished putting an AA rechargeable battery on my cathodes on my preamp tubes. I made sure they were fully charged and matched the voltages out of a few sets of rechargeables I had. Of all the mods I've tried and done, to this amp, remove feedback, triode connection, capacitor bypassing. None have come close to the difference of putting a battery for cathode bias. Voltage will stay constant and it's a pure DC source. I'm really surprised how much difference this made. My bias might of been off some, but this is like night and day difference now. Mostly the highs are so much cleaner and linear, soundstaging and imaging just seemed to open up by a wide margin. I put the - of the battery to ground, and the + to the cathode of my 12AX7 tube. I really was skeptical, or I wouldn't hear much difference, but this was like a curtain was drawn back, it's that much better.

I'm not knowledgeable enough to comment on the practicalities of this, but its a fascinating move -- could you post some photos? I'd like to see how you did it. And how long do the batteries last -- I assume not long, since essentially you are running a toaster.

Do you have more info or links on this Dan?

I'm just learning this tube circuit stuff but this has me baffled and yes, like Nat points out, it sure sounds like a toaster to me too.

Nat, The circuit itself keeps the battery charged, or so I've read. It will last as long as the battery will. I used NiMh batteries so they don't get a "memory", it seems my noise floor also dropped. Everything is much quieter, and allowing more music out. The highs were the most improved, but all through the frequency range I can hear much more detail. I didn't have time to take pics, just solder the batteries in, I used a solid core wire to connect the + to the cathode, then - to my ground. It really doesn't get much easier than that as long as you have room inside and your bias point is close to 1.5v. If it's higher you can use photo batteries or more AA in series. You can even lift the battery off ground with a resistor to increase the voltage some.

Bill, I read some at Steve Bench's site, he has distortion figures of both resistor and cap coupled bias vs. battery. The main difference was the odd order harmonics decreased with a battery. I have a feeling my bias point was off somewhat also, although I don't think enough to make such a drastic change. Another advantage is no AC can pass through a battery and it's a fixed bias, voltage, and current so there is no fluctuation. It's also very low impedance and It's the best "free lunch" mod I've ever tried. Morgan Jones is quoted as saying now battery biasing is feasible due to the long life of these newer batteries.

http://members.aol.com/sbench102/fb_cb.html

http://home.earthlink.net/~jeremyepstein/riaa.html

http://members.myactv.net/~je2a3/JELsc-phonopre1.jpg

http://members.myactv.net/~je2a3/simple45.htm (Scroll down to mods)

http://www.tubecad.com/october2000/page18.html

http://www.tnt-audio.com/shows/etf04_e.html (This is a little different twist)

http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=2283&highlight=battery+cathode+bias&r=&session= (Lengthy thread discussion on the subject)

http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=8420&highlight=battery+cathode+bias&r=&session= (Another lengthy discussion on the topic)

Happy reading

Just an update, I know this has been awhile, but the amp is still going strong with the same batteries I originally installed. I've been amazed sometimes at the amount of bass this amp now produces, but the main benefit is how much quieter the backgound noise is. I guess it's true the batteries do stay charged in the circuit, I will have to open it up and measure the voltage here soon and see how close to 1.5v it is.

MC34B Battery Biasing & Capacitor Bypassing

Hello Dantana,

I've just completed the Battery Bias modification and noticed the improvements you mentioned, such as the increased musical detail which is laid bare. You mention in the thread about Capacitor Bypassing as one of your mods for the MC34B. As I was planning to upgrade the caps, I was wondering if you could explain the procedure, and to confirm whether it improves on having the upgraded caps. Incidentally, there are 4 x 0.33uF and 2 x 0.1uF caps. Some more info on these would be gratefully received.

Thanks.

Platinum

Platinum,

What I probably meant about capacitor bypassing is using a smaller high quality cap and bypass the other caps. Using a say .01 cap parrallel across another cap. This is sometimes a more common practice across large electrolytics on the power supply sections, ie 10-20,000uf caps with a .2uf metallized polypropolene cap. I bypassed most the large electrolytics with metallized caps, I even experimented with bypassing the .33uf coupling caps. None of these mods came close to the amount of effect of doing the battery bias mod. Probably more importantly is to just replace the coupling caps with very good, high quality ones. I was even considering getting some .1 uf Teflon caps and trying those. I haven't yet, but did put in some I bought from Parts Express that were very reasonably priced and sounded good. If you''ve done the 12v heater filament wiring mod also, then now it is more a matter of finding that right synergy of pre/driver and output tubes.

Good Luck,
Dan

Battery Bias questions

Hi Dan,

I have already changed over to 12ax7's (GE NOS 5751's), 12au7's (Amperex Bugleboys) and el34's (jjTesla e34l's).... wow!

I am now getting restless and am keen to try the Battery Bias mod.

Could you please clarify how I should perform the battery bias mod for the MC34B?

I have several questions:

1: Do I only need to do this for the 5751's?

2: I have 1.2v 2500mAh NiMH cells. Will these work OK?

3: Do I need four of them (1 cell per 'half' 12ax7 valve)

4: Do I introduce the cells into circuit as in the diagram that is attached?

5: Finally, do you know of a schematic for the amp?

Hope you can help!!

Andy

Andy,

1. Yes, the driver tubes run about 44v bias.
2. AA cells normally run about 1.4v, you can try C or D if you can get them to fit and they may run closer to 1.5v, plus have a larger reserve and amperage output, but only use NiMH cells.
3. Yes, the bias controls both halves of the tube. I suppose you could split it in halves and do both sides separately, but that's up to you. When you see the schematic then you'll understand they are tied together.
4. Yes, like your drawing, I've attached a jpg on how I've wired it.
5. Yes, I've uploaded it.

Good Luck

Just FYI about batteries...
If you get a standard rechargeable AA battery, it will most likely be 1.12 volts, give or take (it varies a lot!). These days, you might be able to get 1.25+ volts out of professional level Sub-C size cells, which are used for car racing. You might end up spending $10 on the nicest cell available, and charging it for the first time would be difficult without the right equipment. Unfortunately, rechargeable batteries are not going to be close to 1.5 volts any time soon. The solution, possibly not a viable one for you, is to run a plate voltage allowing the bias to be 1.2 volts. Or just run it as it is. I know in guitar amps the bias is run within a wide range, and anything between cutoff and saturation is acceptable, but you generally aim for the middle. Yeah, that's not such an exact art.

I was just thinking about running the bias in parallel between the tubes. This might work if you have grid bias, but if I remember correctly you often cathode bias 12AX7's, and putting the cathodes in parallel, as that would essentially be, would be death, especially in a multi-stage preamp (connecting the input of one tube to the input of the tube that's feeding it= explosion).

I use NiMH batteries and fully charged I think they ran about 1.4v. I think the NiCad will run only about 1.12v or so. NiMH will recharge themselves in circuit while the amp is on. There are so many different topologies you can use, I'd hate to think of every variation.

Recahrgeable batteries have been in commom use for over 20 years now, I am surprised to see there are still a lot of myths regarding their use. For common NiCd and NiMH batteries, as long as charging current is less than 1/10 of the rate capacity (for example, for a 2000mAH battery, charging current less than 200mA), it can be charged indefinitely without ill effects. In a 12AX7 cathode circuit, the current is only a few mA, so there is no danger of overcharging. However, NiMH rechargeable battery has a very high self-discharge rate (typically >1% per day, especially the high capacity variety), there is certain risk of under-charging the battery over a long period of time if the amp is not used frequently, leading to bias drift.

Normally, for a small signal circuit like the 1st stage of an amp, a little drift of bias is no consequence. However, according to schematics in post #9, the preamp stage in this particular amp is directly coupled to the phase splitter/inversion stage, as such, samll bais voltage change in the 1st stage will be magnified by the gain of the said stage and dramatically change the bias of the phase splitter/inversion stage. For this reason, I would recommend the use of low capacity NiCd for this application, because NiCd has much less self-discharge rate than NiMH, and low capacity battery takes less current (and time) to charge.

For the same reason, if you just simply replace cathode resistor and bypass cap with a battery without adjusting anode resistor to get the same anode voltage, there will be bias change to the phase splitter/inversion stage.

Comparing two schematics in post #9, I can't help to notice that the battery biased circuit doesn't have feedback from output transformer secondary like the original. IMHO, removing feedback has a much more significant impact than battery bias itself.

Without feedback, the frequency response will be very different than and not as flat as with feedback. Typically, without feedback, FR rolls off at both ends. If you have a preamp or receiver with mid-range tone control, it is like turning up the midrange or "presence" knob. How much midrange increase depends on the circuit topology and components used.

The second effect of no feedback is the increase of output impedance. This is equivalent to putting a small resistor in serial of the speaker. This serial resistor devides the output voltage and reduces the power to the speaker. Obviously, how much impact it makes depends on the impedance of the speaker. If speaker impedance is independent of frequency, then this extra output impedance has no impact at all on tone balance. However, all speakers have frequency denpendent impedance curve, typically higher at high frequency and lower at low frequency (some speakers have dramatically different curves, but that is a different story), excluding the resonant peaks and crossover regions. The higher the speaker impedance, the less of this voltage divider effect. As a result, this extra output impedance has less attenuation on treble than on bass. This is equivalent to slightly turning up treble control and turning down bass control at the same time, leading to an overall brighter tone balance.

IMHO, the combined effect of frequency response and output impedance change because of removing feedback is the cause of the difference DanTana experienced.

One more point is that rechargeable battery in cathode circuit does pass AC current freely. However, since these batteries have very low internal impedance (typically less than 1 ohm when fully charged), a few mAs of AC current only generate a few mVs of voltage, about 1/1000 of the AC voltage when a 820 ohm resistor is used. As such, there is no need for bypass cap and there is no negative feedback in cathode circuit, leading to much higher gain.

Hi Dan,

Thanks for that - brilliant!

My cells are both running at 1.4v fully charged, so when I next get a minute I'll give the mod a go and let everyone know how I get on!

Cheers,

Andy

ultralc,

Very observant. Yes the second schematic is with no feedback which is basically how my amp is now wired. It doesn't show the triode connection, this circuit is used with CircuitMaker 2000 software which gives me realtime analysis of any circuit changes I may want to try. It is just a screen shot of the schematic, you may notice the 8 ohm resistor on the output transformer in order to make sure the circuit isn't open.

Not passing AC is another benefit of using a battery, at least in my opinion.

Another way of biasing I've read about is using a LED instead of a battery. I've never tried it but it sure seems easy enough to do.