Skip to content
May 21, 2020 / consort3

The PW 12-12 amplifier (1969)

I was inspired by Mooly from the Diyaudio forum and this link to a Practical Wireless archive:

https://www.americanradiohistory.com/Practical_Wireless_Magazine.htm

I found this old project of mine, the Practical Wireless 12-12 stereo amplifier which was published April-June 1969. I built one and amazingly it worked. As an student apprentice, I remember taking it into work and checking its performance. It had a poor slew rate caused by the slow germanium output transistors but subjectively sounded great. I salvaged some transistors from my failed Sinclair Z30 namely ME0411, ME4101, ME0401 and ME6101, to build it.  I only remembered this when I saw their type numbers written on my copy of the circuit diagram.  These had an unusual outline TO106. Made by Micro electronics of Hong Kong if memory serves. It looks like I was lucky with these devices, scroll down on this link to the Z30 bit:

http://diy.torrens.org/Sinclair/inside/NktRd.php#fitz

The bias on the PW12/12 was supposedly thermally stabilised with a thermistor but no mention was made of putting that on the heatsink. Anyway it did not thermally run away. VR1 was tweaked to give half the supply voltage at the output (before the 1000uF) and VR2 was tweaked to give 50mA in the output transistors.

pw1212

The Sinclair Z30/50 mentioned above added a differential input and a current mirror but took away quiescent current control, hence crossover distortion could occur.

sincamp

The Sinclair became the basis of the Naim design, current sources replacing the current mirror. A Vbe multiplier was added to set the quiecent current and a protection circuit for the output transistors was implemented. The Arcam design had 2 current mirrors and quasi complementary output stage using the ubiquitous 2N3055

Another amp with the virtue of simplicity the Digi-125 from 1989 Australian ETI

dicker

However this has no provision to set the output transistor bias current and no bootstrap to increase the output swing. Only 2 diodes in the original diagram in the link below (page 80). Also optional 1 ohm and 220nF Zobel on output.

https://worldradiohistory.com/AUSTRALIA/ETI-Australia/80s/ETI-1989-05.pdf

Hugh Dean with his Aksa amps addressed the issues and then came up with the Baby Aksa here:

baksa

The amplifier can be improved still further with these Doug Self mods:
1) Input stage emitter degeneration R2 & R3 below
2) Darlington VAS
3) Input stage current mirror loads Tr10 and Tr11

Diagram does not show the triple output stage which is possible.

dougself

I thought the Z30 amplifier was worthy of tweaking as it has an elegant current mirror arrangement. My tweaks to the Z30 include a Vbe multiplier to stop crossover distortion, an offset tweak and inverting the circuit to allow a NPN Vas transistor . Also I have used integrated Darlingtons for the output stage and used a single 36V power supply.

sinz30

Here is the crossover distortion at 10kHz without the Vbe multiplier

sinz30a

The offset needs R7 to be a 19k resistor (a 1k and 18k in series). The offset is caused by the inbalance of the currents in the input differential pair. I came across the Nelson Pass AB100 amplifier while researching this and it is worthy of study, as he uses current mirrors creatively and has found a way to parallel integrated Darlingtons, a technique which has so far eluded me.

Here is the AB100 circuit courtesy Nelson Pass

Prasi AB100 schematic

Review of the commercial amp the AB100 was based on

https://www.stereophile.com/integratedamps/780/index.html

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: