DUAL cv12 (1967-72) redesign

[Hayk]

The DUAL, German brand, cv12 is a special amplifier of only 2.5w/ channel. The precedent model was 20w×2, and the succeeding, 30wx2, all in similar enclosure.
It was the first of the Audiophile series from this brand. Although the complimentary pair AD161-162 was available, the designer chose a very modest transistors used in portable radios.
Using only FET transistors what best I could find, I redesigned it keeping as near as possible but much higher power.
It delivers 27w@8 and 45w@4 with 0.2%and 0.05% THD. 1w@8, 0.01%.
It has dominant even harmonics in all power range before clipping.
The output transistors can be ecx10n,10p. From Ali possible to buy used original Hitachi ones.
This design has special open loop output impedance. Lower than Vmos or triodes it has 3 ohms impedance. This character gives tight link to the tweeter damping.

 

[solderdude]

Does that mean all completely different designs (in all possible ways) are a re-design of the DUAL ?

 

[Hayk]

Does that mean all completely different designs (in all possible ways) are a re-design of the DUAL ?

How many class AB amplifiers over 20w and less than 0.1% THD do you know made of 4 transistors?

 

[solderdude]

Does that warrant it to be called a re-design of the CV12 ?
It is a totally different amp with 4 entirely different solid state devices so basically a totally different design with different harmonic and IM distortion, different power output etc.

Why not just call it ... a simple low component count power amplifier ?
The only things it has in common with CV12 is 4 active devices, single rail power supply and output capacitor.

 

[Hayk]

If I have called it otherwise, someone of good knowledge in amplifier topology would have mocked at, to be nothing else but 1967 DUAL cv12, with fet transistors instead of germanium.

 

[Vincent Kars]

The CV12 was the first amplifier I bought.
It was called 12 because it delivered 2 x 6 Watt (probably peak)

https://www.hifiengine.com/manual_library/dual/cv-12.shtml

 

[solderdude]

If I have called it otherwise, someone of good knowledge in amplifier topology would have mocked at, to be nothing else but 1967 DUAL cv12, with fet transistors instead of germanium.

Probably not .... because that person would have said your design is fun low component count J-FET + MOSFET amplifier instead of silicon BJT + germanium BJT output devices design.

I don't think anyone would have thought of it being reminiscent of the CV12 output stage design (on account of 4 transistors being used).

 

[Hayk]

Input stages compared.
12k colector load-12k drain load.
100 emitter resistor-100 source resistor.
DC feedback on emitter- DC feedback on source.
Capacitor coupled to VAS- Capacitor coupled to VAS.
2.2M current bias resistor to collector, 100k-1M voltage bias resistor to ground.

 

[Hayk]

VAS comparison.
Grounded emitter-Grounded source.
Current bias from output for Vcc/2 with 500k input load- voltage bias from output fraction for Vcc/2 by 560k input load.
High G-D capacitance of IRF510 acts as compensation.

 

[Hayk]

Output comparison.
The lateral mosfets don't need thermal compensation and source resistors
The bootstraps are similar except I added 1k series resistor with the capacitor to decrease the input impedance of the output stage.

 

[solderdude]

Yep different design, different operation points, different performance, different parts.
Just basically the same as many other similar designs from that era.
So ... has as much to do with the Dual as with many other designs from that era.

 

[Hayk]

BUT the same circuit topology.

Yep different design, different operation points, different performance, different parts.
Just basically the same as many other similar designs from that era.
So ... has as much to do with the Dual as with many other designs from that era.

The ordinary topology of amplifiers above 2w was with a driver, either single or Linn topology as shown bellow, CV30 Grundig ocean boy and Blaupunkt car radio.

 

[solderdude]

Iindeed there are many, many different ways to design an AB amp. Especially when one does not want/ is allowed to copy the design of others.

 

[Hayk]

I copied the cv12 design as it has been tuned for sound quality.
After 70 years of transistor class AB amplifier design, nothing new under the sun, they say.
I am preparing the PCB , maybe ready tomorrow but can't try out before several months.

 

[Hayk]

The CV12 was the first amplifier I bought.
It was called 12 because it delivered 2 x 6 Watt (probably peak)

https://www.hifiengine.com/manual_library/dual/cv-12.shtml

It says 6w music power with 4 ohms.
The trouble with AC180/181 is the power dissipation only 300mw max.
With heatsink version should not be more than 1w.

 

[solderdude]

At peak (music) power output there is almost no voltage across the conducting transistor and as it can switch 1.5A it can also do peaks of 1.2A.
The power supply voltage sags to 15V (at 4W output) so 'music power' could well be just around 6W (10% distortion or so).

The 'tuned for sound quality' is what all manufacturers said in those days and ironically ... still today.

 

[Hayk]

The board is 65x50mm. The input jfet can be k117 in T092 or its SMD version k209.
The gain is 34db (50).

 

[Hayk]

I will try a BJT version of this circuit using high gain power transistors. MJL4281/4302 or D44H1/45h11
The MJLs are about $7.5 a piece but there is the Korean manufactured for $2 a pair from Taobao. I think 25-50w 8-4ohm is possible.

 

[Hayk]

I will show you what this topology can do.
Bellow is 1khz upon 8ohms with zero bias.
Selecting best BJTs, the distortion is 0.1% for 10w@8 ohm but 1khz or 10khz, same same. It reaches 0.5% at 20w.

 

[Hayk]

I replaced the outputs with C5200/A1943. The operating voltage needs to be adjusted 24v for 8 ohms and 20v for 4 ohms, I added a switch to toggle.
Pioneer used this topology but with drivers in its 70's models. I liked the sound of this SA500.
For 8 ohms 1w, the THD is 0.01% to be 0.1% at 25w the double for 4 ohms.