diyAudio (
-   Solid State (
-   -   Intro a circuit of study.. (

turnA_zpoint 30th July 2008 04:19 PM

Intro a circuit of study..
Hi All !

I tried a circuit using OP amp for 1st stage, and Addition integration type DC servo ....

It is recently study.

In case 2Vp-p sin wave 8ohm load, -3dB about 1MHz.

Sorry, it's text is only Japanese ,but I think the circuit is the world commonness...


CBS240 31st July 2008 03:47 AM


Interesting circuit, I like it.:) J-fet's really are great, are they not? One very minor thing, why did you use the higher voltage devices (2SC1845/2SA992) for the amplifying transistors and the lower voltage ones (2SC1815/2SA1015) designed for "linear Hfe" as the cascode? Or are these not the devices you reference in your circuit. Wouldn't you want the more linear devices for the amplifying transistors and the lower Ccob transistor for the cascode? :confused:

As for the circuit, I may have to experiment a bit with that J-fet set-up you got there:)

turnA_zpoint 31st July 2008 02:46 PM

Thanks !
Thank you for your Reply.

J-fet is (thar is the first stage of input) square characteristic ,and
to stabilize to the temperature change than the transistor.

 Moreover, the transistor thinks that the influence is a little even if it uses it by 0 biases of FET in the idling current in my impression while tone quality is changeable.

 The second stage 2SC1845/2SA992 ,I tested
A970/C2240,A1145/C2705,A1015L/C1815,A1191/C2856,A1358/C3421,and(FET)213/J76,K2013/J313(this so large imput cap) is good result.. if it does as the current etc. but,they are the best,2SC1845/2SA992 Finally, it decided it by hearing however.

The cascode 2SC1815/2SA1015 "linear Hfe" is..just you said,
The idea when other people produce it. thinking good obtained one.(but only 2SC1845/2SA992 are useing..) I tried A1145/C2075,think not so difference.(Cascode it's input is Emitter think not so influence)

Though it is a supplementation,
the input buffer LME49710 ,MAX410EPA is good result but it's 5V one..
Though K1529/J200 the Source resistance is less,there is no ploblem in my case.
The 1st stage to which necessary deflection for error margin amplification is reduced with OPamp, that LM4562 ,AD8620
is good result,too.

Constant Power supply

Addition integration type DC servo


lumanauw 31st July 2008 03:02 PM

2 Attachment(s)
Very interesting :D

What is this diagram mean?

turnA_zpoint 31st July 2008 03:27 PM

Fig 1 (left) is nomaly Active LPF.

Fig 2 (right) is Addition integration type DC servo,
This AMP circuit is inverted (and Current feedback) type.
so input signal and output signal Phase is inverted.

Then input signal and output signal is canceled each other,
at the same time by LPF ,only offset DC is effectively extracted.
and feed backed.

Therefore, this amp is FLAT to DC Hz , with only 0.1 uF cap.


lumanauw 31st July 2008 03:47 PM

Is it the 47k-2k2 and 470k-22k?

turnA_zpoint 31st July 2008 04:42 PM

(1) 22K-470K is for ADD DC servo.
(2) 47K-2K2 is for inverted amp gain.

gain= 47K/2K2

then output signal must to be 2K2/47K time (Dividing calculation
for to be same amplitude).
In a high frequency which gain of operational amplifier becomes enough with LPF little difference is no problem,
Low frequency, each other inverted signal is canceling ,and at  the sametime by LPF dividing)

If CAP (0.1uF) is large ,then it is OK this.
But thinking for LPF , 22K-470K.

(If this OPamp is AD8620 fet input, the register is 220K/4700K ,OK)

The data (Though it is not the one of this circuit) when it examined it before is put.

ADD type DC servo AND Normal Act LPF data


jan.didden 31st July 2008 08:31 PM

This is a very sophisticated circuit. Hat off! Did you simulate it or build it?

Jan Didden

turnA_zpoint 1st August 2008 12:25 AM

Hi Jan

This actually built, and I listens to music with this.
This circuit ,can tune oscillate with enforceable of the phase amends only in 1Kohm-2pF of A1145/C2705 at common emitter of the second step, (and under the present situation,if is not them,no problem ).

It makes it only to Coil and Register to make amends to a capacitive load of the output, this actually driving the speaker because of a rectangular wave. Compensating network of the integration type by Cap and Register .

The reason to extend the frequency response to 1MHz,
the frequency response of about 500KHz or more is given, when LME49710, OPA627, AD797, and MAX410, etc. are evaluated to the OP amp in the input buffer, this understands the difference easily and is. This is thought the influence of a county delay.

The photograph is put though only the volume of the input is different.

AMP inner photo

It makes it to the volume made of plastic contacting of the build-to-order manufacturing now. 2CP-2508-S
(The maker is Tokyo KO-ON DENPA. )



turnA_zpoint 1st August 2008 12:31 AM

It is an addition.
The above-mentioned volume is the one having bought it with



All times are GMT. The time now is 03:18 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright ゥ1999-2015 diyAudio

Content Relevant URLs by vBSEO 3.3.2