BC550-560 Symmetric with LATERAL Output - Hiraga style

[lineup]

Here a nice little amplifier with 2SK1058-2SJ162 output.
It is very much Hiraga style with symmetric topology.

Amplifier is stable. No oscillation.
And Fourier shows very nice H2 H3 distribution.
Both at 1 Watt and at 7.5 Watt, +20dBu.

And low level THD. Close to HIFI.
THD 0.007% at 1 Watt feels good.

Settting this amplifier up is easy if you have them Lateral MOSFETs.
You only need 1k pot and some BC550C,BC560C.

 

[Salas]

Or 2SC1815, 2SA1015 (GR)?

 

[lineup]

Yes, of course, Salas.

BC550C, BC560C
or
2SC1815, 2SA1015
or
2SC2240, 2SA970
or
2N3904, 3N3906
or
2N5551, 2N5401

That is all normal small signal which are good complementary.
They will all do and amplifier can be adjusted for zero output with potentiometer
no matter which transistors.

 

[Salas]

Other handy Mosfets maybe? What is the clip power level of that 24dB(?) gain amp?

 

[Dr_EM]

Tempted to try with Exicon 10P16 and 10N16 complement output devices, got lots of them. Likely to work properly?

It's impressively low distortion for such a simple design!

 

[lineup]

Salas, you got me notice.
BC550/560 are only 45V, so for U1,U2 it is alright,
but for VAS I need transistors that can take 50V (2*25)
So, U3, U4 could be 2N5551, 2N5401 (or 2240/970)
New diagram below. THD only 0.002%

Other MOSFETs, like IRFP240/9240?
Yes, it should be possible, if I put more current in VAS. BD139,BD140

Clip level?
It passes over 0.050% THD at 26.7 Watt
With IRFXXXX HEXFET we do not get so high as they take more space from Supply.
VGS 4.5V

 

[lineup]

Tempted to try with Exicon 10P16 and 10N16 complement output devices, got lots of them. Likely to work properly?

It's impressively low distortion for such a simple design!

Yes, those Exicon will work very well.
They are lateral MOSFETs and very easy to drive even for 2N5551 and 2N5401

Notice:
In post above I provide a new schematic
as BC550C,BC560 will not work in VAS stage (U3,U4)

 

[Salas]

2SC2240, 2SA970 will take it also, nice. Enough Watts I would say. Is it biased at 100mA Lateral zero tempco?

 

[lineup]

I did bias at around 250mA. (R8=300)
With adjusting, replacing blue resistor R8 it can be any bias.
Even Class A. But this is not needed for this amplifier

 

[Salas]

Still not too hot, that is significant for costs and practicality. VMOS are handy but would need amperes of bias to be stable tempco and low THD unfortunately. There are Renesas for $20 the quad non matched around thankfully. But those got good enough tracking characteristics between sexes I would say. Is there a reason you see to use matched ones?

 

[WuYit]

Hi,
in a Hiraga style output stage the current depends on the much more stable Vbe of the driver. You can use any device with any thermal characteristic.

 

[Dr_EM]

When building this as a real-world amplifier, is it not a good idea to have a coupling capacitor and RF filter at the input?

Also, on Rod Elliots lateral Mosfet design (P101), he adds capacitance to the 2SK1058 such that the gate capacitace matches the inherently higher capacitance of the 2SJ162. Is this not a concern with this type of driver?

Is a Zobel required for stability too on this amplifier? Or any kind of miller caps?

I'll build this amp with 2N5551/2N5401 complement and my Exicon 10P16/10N16 and see how it goes. Just need some aluminium angle for those TO3 devices.

 

[WuYit]

However, this is not a Hiraga style output stage.

 

[lineup]

Revised version

Here is a better version. Uses 2SC2240 and 2SA970.
R15 and R4 are changed from 100 to 120 Ohm.
970/2240 gives more 2nd order harmonics distortion in relation to H3.
At least this is what is shown in SPICE Fourier analysis.

 

[AKSA]

Nice circuit, Lineup. You are really being very creative!

Hugh

 

[MiiB]

hehe.....why not put the zetex to 92 mosfets into this circuit as well...??

 

[tsz]

Like it a lot, Lineup! Have you built it too?

How about adding cascodes to the input transistors?
The base of a cascode for U1 could be connected to the emitter of U2 (perhaps with a diode (LED or zener) between U1/U2 emitter and R1/R6 for more voltage across the input transistors) and vice versa.

 

[sonnya]

Hi Lineup. Looks good.

R14 will limit the slewrate. even if it is CFB you need a lowimpedance that feedback need to look into.

 

[sonnya]

I did bias at around 250mA. (R8=300)
With adjusting, replacing blue resistor R8 it can be any bias.
Even Class A. But this is not needed for this amplifier

Try adjust the bias current and look at signal on centertap of r14.

You would proberly see the crossover distortion as soon as you go below 350mA.

But anyway it looks nice.

 

[lineup]

Nice circuit, Lineup. You are really being very creative!

Hugh

Thanks, Hugh. You are being kind.
I regret mentioning BC550-560 in topic subject.
But eventhough I now use SA970/SC2240 we still can use BC550-560 for input.
But not for the 50V voltage amp stage.