Low frequency loss thru a deck

[Alex SE]

What can be a reason of low frequency loss thru a deck? I was thought that there is some compatibility issue between a deck and sound card (Creative) but the results are exactly the same even with a new card (Scarlett). The worst thing is that I'm getting the same results with two decks of the same model (RS-B965). The other decks (555, 565 and 727) don't have that problem.
Here is a measurement. (green lines)

Once again, it's not recording, it is signal passing through a deck (monitor).

Anny guessing and advice might help, so, thanks in advance.

 

[J!m]

Are they running signal through Dolby chips even when no NR is selected?

 

[Alex SE]

There are two separate inputs, regular and "CD direct" that is bypassing Dolby IC. Same loss with both inputs. A friend told me that there can be some HF filter but I haven't playing with such a mods. Replaced some opamps, added caps between the rails and replaced some switching ICs. Really nothing special. One of them is recaped but it doesnt makes a difference.

 

[Alex SE]

If this helps...

 

[J!m]

Doesn’t help me… I just know corroded Dolby chip leads can do all sorts of strange things.

 

[vince666]

Are they running signal through Dolby chips even when no NR is selected?

no, the B965 has full dolby bypass, so when dolby is switched off it's totally ruled out of the signal path.

this bass roll-off experienced by AlexSE is making both myself and himself going mad for quite some time now, even because I also own the B965 and the same model/version of Focusrite Scarlett soundcard and I am not experiencing that problem at all... which makes AlexSE's issue even more a mystery, to me.

There must be some reason for that... but which one?

 

[Alex SE]

Well, the problem is located and is kind a weird.
Alex A.N.T. advised me to trace a signal and see where exactly chat cut-off occurs. Using oscilloscope and 20Hz and 1kHz signals I should find a problem... and I did it, but using DMM since that I don't have oscilloscope.
Feeding 1.5V to input of a deck, I've found that levels of both 20Hz and 1kHz are just fine all the way to C29 (C30). On the input of both caps I've measured 296mV, but on the output of it 20Hz is 112mV while 1kHz still keeps level of 296mV. (See marked points 1 and 2)
Anyway, there is a mod on IC9 (add a ZVP3066A transistor one 100K resistor and cut one trace). That in order to control voltage to IC9. What is unknown is if that mod applies only if ANT4066 is used or all variants of 4066?!

 

[vince666]

What is unknown is if that mod applies only if ANT4066 is used or all variants of 4066?!

or if it was needed with 74HC4066 and not with ANT4066?

and what about MAX4066?

 

[vince666]

Since i should give a clean and a drop of oil at capstan and replace the belts, I will soon open my B965M, modified directly by A.N.T., and give a good look at what I have in that zone.
But on my deck there are ANT4066 chips... but, if i understood it correctly, MAX4066 would work the same, since the control voltage of this deck isn't too high.

But, as you just know, on my B965M i am not experiencing any roll-off like that... maybe just half dB less at 20Hz, which is just normal.

 

[Alex SE]

... and a problem is finally found! It is MAX4066 that obviously cant handle voltage on IC9 position.
Here is comparing MN4066 vs MAX4066. Observe that MN was there by the factory and MAX is coming from Mouser, so no fake chip.