Clon C-3850

[Fukinagashi07]

I replaced the standard NE5532 with an OPA1612 in DIP8 format, but I'm observing increased noise levels on the oscilloscope.
Are there any supporting components that should be adjusted (e.g., decoupling capacitors, feedback network) to optimize performance with the OPA1612 in this circuit?

 

[chermann]

check if you have the same "noise" with NE5532.
as written above --> maybe fake OPA1612.
or if you have bad luck the NE5532 too.

kr
chris

 

[chermann]

decoupling_ V- to GND 100nF film cap or ceramic + 10µF
decoupling_ V+ to GND 100nF film cap or ceramic + 10µF
as close as possile to the opa

here is a decoupling hint at the opa627

 

[Fukinagashi07]

check if you have the same "noise" with NE5532.
as written above --> maybe fake OPA1612.
or if you have bad luck the NE5532 too.

kr
chris

No noise with the NE5532. Tested with two OPA1612 and I have same noise with both so I suppose something must be changed in the circuit...

 

[Kleinhorn]

Hi,
i am using an OPA1612 too. I had made all the changes, following the advices of the user chalky. I have tried different OPA, all are working, but with different sound. LM4562/LME49720, AD8599, OPA627 (double), OPA1642 and more....
The OPA1612 has the best sound to my ears. Absolutely quiet, no hiss, no hum. I bought my OPA1612 on Aliexpress, but not finished on a pcb. 4 OPA1612 are working fine in other little preamps, installed after my DSP.
Perhaps there is a problem with your OPA612. The changes of components you will find in this thread, perhaps you have made already...

Greets
Peter

 

[Radian]

Ok guys what's the consensus on the CRD mod. Is it a sepp up or just different?

 

[Kleinhorn]

I have used 5,1 k resistors instead of the CRD diodes. As Chalky wrote, it is nearly the same. The other changes to get classA and other gain are more important, if i compare with the original board..


Greets
Peter

 

[Radian]

What other changes are we talking about?

 

[chermann]

Set the gain down.
Change nearly all bias currents = Change resistors and/or crd.

 

[Radian]

CRDs came in today so might as well use them.

 

[Kleinhorn]

Gain change see post 3. List of michael beeny

 

[Radian]

I'm only using the discreete part of the Accuphase clone so this might not have the same effect.

 

[chermann]

if you want to modiy then you have to change something as written in the posts above.

 

[chermann]

A new "Ferrari", C3850 Mono.
Values are the same as on my schematic. It was an idea...Sandwich.
If you want you can use two supplys. Ground completly isolated for two channels...
You can use other output transistors for more power giving to a headphone. C5 has a diameter of 10 mm. No space for a bigger non polar capacitor...or i have to make a bigger pcb. This is as small as possible...only 4 crossings...
I have to test...

Hi Peter

you did the modification of seperate power for the opamp on your LM317 modified clone PCB - dont forget this.
🙂

 

[Kleinhorn]

Hi Chris,
i never forget what i have done...🙂
From 5 to C

Greets
Peter

 

[Kleinhorn]

I have designed a new "Mono" PCB for Ad797 with a pot for offset....not ordered, but the older version it is on its way...i would like to built a "sandwich" to use a very little housing for it...
As i wrote...many projects...too many...

Peter

 

[fesoni]

In the last two stages of the circuit, you can replace bc550 with bd139 bc560 with bd140. 4 transistors for the left channel and 4 for the right. You will get a good headphone amplifier. Just like in the original 3850 circuit. Transistors can be installed without radiators - here is class AB. They practically do not heat up.

 

[STOXX]

...of the circuit, you can replace...

Yes. I opted for SS8050 SS8550 but I still have not bench tested. You should lower the emitter resistors if you choose to drive smaller impedances with more powerful transistors. I opted to do so in the final stage only.

I am sorry to all for never getting this thing on the bench. Some day...too much going on.

 

[STOXX]

Check temps before ruling out heatsinks.

 

[STOXX]

Download the LT SPICE file back in the thread and try all such modifications at home on your computer.
This circuit simulates well in LT SPICE