D1 Output Stage Help needed.

[pro]

Hi averybody.
I made the D1 Output stage (see attached schematic), for my Teac VRDS8 player that uses 2 BB pcm 1702K per channel, and I have two problems that I can't resolve.
First one, the output level is less then 1 V with a 1KHz 0dB sinewave signal, and second problem, the signal out is highly distorted for strong input signals. Averything works fine if I pull down the digital level of the player of 10-15 dB, but tho output level becames really very low.
l have 32V rail, 0V on the source of the input fet and 16V on the drain.
I tryed to change the load resistor to 2K and 4K, and the bials current to 8.5 and 4.5 mA, but nothing is changed.
Could please samebody help me?

[CheffDeGaar]

Hi,

Had a quick look at your schematic, and I don't understand why your first Mosfet (M4) has no drain resistor...
Are your DAC (I output) balanced (One PCM per polarity ?) Where do you connect the PCM1702's outputs ?
I presume the two inputs are at the sources of M3 & M4. Right ? Feeding the I/V converter at the source of M4 won't give you any sound, and M3 will... Just add a drain resistor to M4, and duplicate the follower built around M1...

[pro]

Thanks Cheef.
The current of each dac is feeded in the source of M3 and M4. I want obtain a single ended output (not balanced), so I don't connect M4 drain to samething. Isn't it correct?
If I duplicate the follower of M3 on M4 how I obtain a single ended output?
Ciao.

[till]

Why not build the D1 stage exactly like in the manual, and only pick up the + signal against GND for you single ended output? (at least if you have one inverted and one noninverted DAC output)

If you have only one noniverted DAC output build half an D1 stage. I don´t understand why you need that BC550 stuff at the sources, and why you build something like 3/4 of the D1 stage.

[CheffDeGaar]

Well, I'm not sure how the M3 "converter" would behave without a drain resistor Put it in, and let's try again. And with a low part count, you could have it balanced (for future use ) Just use any output for single ended.

[hjelm]

The input stage is neat. I like the fact that you pull constant current on the input stage. Other than that i agree with cheff there is no use for the left side of the input stage.

As for your problem have you checked the signal at the positive side of C1 with a scope, or DC voltage if you haven't got a scope.

The only thing i can think of is that you are close to one of the rails and clip with high input signal but you cannot be if you have 16v on the drain of the input fet.

[pro]

For Till.
The dac is balanced two pcm1702 per channel. The bc550 are just CCS, so I can also have a lower minus rail (also unregulated).
For Hjelm.
I don't have a scope (jet) but on the source of M1 I have 14V DC.
The problem seems to be a saturation of the dac not of the output stage. The data sheet of the 1702 just says that the output impedence in 1K and the output current +-1.2 mA.
I have completely isolated the output stage of the Teac so wouldn't be any interferences from it.
Aniway, I will try to apply the second follower and leave it unused.
Thanks to everybody
Ciao.

[till]

Thanks pro,

As the Fet should make the voltage across the resistors you substituted with CCS always constant, i can´t see whats the advantage of current sources. I only see more potentioal sources of problems...

Also a simple zehner regulator one Fet, 30 something volt zehner is less parts than those BC550s and easy to test for function without scope.

In case the problem comes from overdriving the DAC in some way (i suspect the voltage at the source is not properly hold constant at 0V because of thouse CCS) the most simple solution is to use resistors here as Nelson did. As you have balanced current output from Dacs, i would definitely build a full balanced D1 stage and for single ended output use only the + output side. In case voltage output is not enough what about higher resistor values, and in case current bias gets to small this way more voltage at the rails? In case voltage output gets to high, resistor between drains as Nelson volume control. For some reasons i do not understand yet i like the DAC with D1 stage i built better without volume control resistor between the + and - side....?

[CheffDeGaar]

Originally posted by till
i can´t see whats the advantage of current sources. I only see more potentioal sources of problems...

Tested both on my PCM63 version, and preferred the CCS... Just a matter of taste... One advantage is the much higher impedance.

(i suspect the voltage at the source is not properly hold constant at 0V because of thouse CCS)

The CCS is not responsible here for voltage drifts. The gate voltage sets the source at 0V for a given current, and the CCS just adapts itself. Given the large voltage between CCS's collector and the emitter, the CCS is pretty insensitive to the real mosfet source voltage.

As you have balanced current output from Dacs, i would definitely build a full balanced D1 stage and for single ended output use only the + output side.

Couldn't agree more

[pro]

The 0V at input sources is quite stable.
I have already made an half D1 stage for the tda1541 and I had not any problems. See I/U stage D1 clone with 2sk389 finetuned...
Anyway I use regulated supply, and the use of 3 bc550 two diodes and a resistor for 3 CCs doesn't seems to be a lot of parts.
Ciao.