• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Phase inverter/power section

redplater

Member
2010-11-16 11:22 pm
After building a good number of tube guitar and bass amps, I'm getting ready to start my first audio build. I'm looking at something in the range of 40-50 watts in a push-pull monoblock amps.

My question is, being a guitar amp guy.....the signal from the guitar preamp typically goes into a long-tail phase inverter into the power tubes. Usually, the phase inverter is a 12ax7 or 12at7.

I have yet to see this arrangement in an audio amp. While I have seen the occasional long-tail phase inverter, there's always some other driver stage in front of it. Why?

Any suggestions for schematics that are simple with few driver stages before the phase inverter?

I'm also building a grounded-grid preamp....do I need a driver stage in between this preamp and the phase inverter?

Thanks!!
 
Think in terms of the total gain needed and the 2 VRMS a "standard" CDP yields to achieve the power O/P desired.

The GG preamp has some gain. That might be enough to make "El Cheapo" style circuitry (schematic attached) work with the KT88s or 6L6GCs you will have to use. If a CDP directly drives an "El Cheapo", the 12AT7 LTP can produce enough gain only for "12" W. types and 7591s.
 

Attachments

  • ElCheapo-23jun06-map.gif
    ElCheapo-23jun06-map.gif
    36.3 KB · Views: 264

redplater

Member
2010-11-16 11:22 pm
Sorry, but I'm a lowly builder of guitar amps...I don't know any of these acronyms: VRMS, CDP or O/P. However, I did kinda gather that you don't think I'll get much bang for my buck without putting a little more juice in the front end of the power section.

So, with that in mind...do you have any opinions of the Dynaco Mark III? It looks like a pretty simple build and should not take me too far from my Marshall/Fender power section/comfort zone that I'm used to.
 
VRMS = Volts root mean square.

O/P = output.

CDP = compact disc player.

The "El Cheapo" splitter/driver is highly competent. FWIW, I think that between the gain of the GG preamp and buffering the LTP with source followers you would have a satisfactory setup. Things could get very interesting if you dragged Soviet surplus 6Π3C-E (6p3s-e) tubes slightly into Class "AB2" with Tubelab's PowerDrive. The Russian tube is an affordable, good sounding, close match for the 6L6GC.
 
I have yet to see this arrangement in an audio amp. While I have seen the occasional long-tail phase inverter, there's always some other driver stage in front of it. Why?

Vixen Main Schemo

Have a look at that. The grid drivers make for easy, DC connected, fixed bias adjust for the finals. That gets rid of the sonic penalties of capacitor coupling to the final control grids. Overdrive can't put an additional negative voltage on the grids when a transient turns on the GK parasitic diode. Since you have current sourcing, the 807s can slip a bit into Class AB2 transparently, making for better clip behaviour.

It also gives much better current sourcing to avoid the problem of slew limiting at the higher frequencies.

It presents the 6SL7 LTP a much friendlier load (Hi-Z, Lo-C) than would the 807s.
 

artosalo

Member
2010-02-16 9:00 am
there's always some other driver stage in front ..... Why?/

One amplifier stage is needed before the phase splitter so that the sensitivity on the whole amplifier is sufficient ( 300 mV to 500 mV).

One reason to the need of additional gain is that in HiFi-amplifiers some amount of negative feedback is used (typically 10 dB to 20 dB) and this lowers the sensitivity with same amount.

Any suggestions for schematics that are simple with few driver stages before the phase inverter?

Below is one. This uses 6L6G, 6P3S-E or 807 tubes ( I have tested all with this design). The power output with 450 V anode voltage is about 35 W, but even 50 W is available with some 550 V.

Essential points are to have separate bias adjustments for both output tubes as well as AC-adjustment at the phase splitter. By these adjustments the output tubes can be "matched" to minimum distortion.

[IMGDEAD]http://kuva.termiitti.com/image/14088.gif[/IMGDEAD]


http://kuva.termiitti.com/image/14088.gif
 

artosalo

Member
2010-02-16 9:00 am
Artosalo, Could you post the power supply for this schematic?

Here it comes. There are some additional components, like DC-filament supply for pre-amp, which was not in the amplifier schematic I sent earlier. This power supply was designed for a integrated stereo amplifier, and therefore so much stuff.

The transformer is from Italy: inMADout Transformers



[IMGDEAD]http://kuva.termiitti.com/image/14101.gif[/IMGDEAD]
 

redplater

Member
2010-11-16 11:22 pm
OK Artosalo...I think I have enough know-how to build that amp...except for a couple schematic symbols I don't quite understand. What is the line (and perpendicular line) through the 2nd 33K resistor coming off the plate of the 6N7S? Similarly, what are the symbols in the bias supply?

Thanks!
 

ray_moth

Ex-Moderator
2004-01-27 8:55 am
Jakarta
OK Artosalo...I think I have enough know-how to build that amp...except for a couple schematic symbols I don't quite understand. What is the line (and perpendicular line) through the 2nd 33K resistor coming off the plate of the 6N7S? Similarly, what are the symbols in the bias supply?
Thanks!

Those are the symbols for preset potentiometers (pots). The line with the perpendicular line at the end of it represents the slider. In the case of the 33k pot in the amp, this is not a very good idea, IMHO, because of the high voltage involved. Basically, it enables the plate load of the right-hand triode to be adjusted to overcome the imbalance of the LTP splitter, caused by using a finite resistor as the tail resistor instead of a constant current sink. Normally, this requires the plate load resistor of the right-hand triode to be about 10% higher, so instead of using a 33k pot here, you could probably get an acceptable result by using a 3.3k resistor in series with the existing 33k resistor.
 

artosalo

Member
2010-02-16 9:00 am
Basically, it enables the plate load of the right-hand triode to be adjusted to overcome the imbalance of the LTP splitter, caused by using a finite resistor as the tail resistor instead of a constant current sink.

This is right, but only partly. I have had also other things in my mind when designing this way.
As can be seen from the amplifier schematics, both output tubes have own bias adjustment. By these trimmer pots the quiescent current of both output tubes can be adjusted to precisely same level.

If the output tubes are not well matched - which is not even necessary - the gain of the output tubes are obviously slightly unequal too. This problem can be solved by adjusting the output voltage of the other side of the phase splitter.

In summary, the separate bias and driving voltage adjustments makes possible to build a good performing and -sounding HiFi-amplifier from not-so-good matched output tubes.