CS4398 DAC output hiss, help needed.

[Val33]

I have this dac board Dual CS4398 CS8416 24Bit/192KHZ DAC and have a problem with hiss when bypassing the opamps with either a cap or a transformer.

I currently have the transformers fitted. They are these: Sowter Type 5069. I have 250r on the primaries, bypassed with 220pf, 680r across the secondaries in line with 1n. I have tried adding 200pf from the primaries to earth. The opamp section is completely disconnected and the +/-15v is not connected.

The hiss is only present when there is an input and is not volume dependant. There is none (or almost none) when using opamps.

The schematic for the board is here: http://valwalker.co.uk/misc/4398X2Bv1.pdf

I am guessing that either I need a filter somewhere or this is power supply noise.

Any ideas please?

Thanks

Val

[johnmuir]

Originally Posted by Val33

I have this dac board Dual CS4398 CS8416 24Bit/192KHZ DAC and have a problem with hiss when bypassing the opamps with either a cap or a transformer.

I currently have the transformers fitted. They are these: Sowter Type 5069. I have 250r on the primaries, bypassed with 220pf, 680r across the secondaries in line with 1n. I have tried adding 200pf from the primaries to earth. The opamp section is completely disconnected and the +/-15v is not connected.

The hiss is only present when there is an input and is not volume dependant. There is none (or almost none) when using opamps.

The schematic for the board is here: http://valwalker.co.uk/misc/4398X2Bv1.pdf

I am guessing that either I need a filter somewhere or this is power supply noise.

Any ideas please?

Thanks

Val

[johnmuir]

Originally Posted by Val33

I have this dac board Dual CS4398 CS8416 24Bit/192KHZ DAC and have a problem with hiss when bypassing the opamps with either a cap or a transformer.

I currently have the transformers fitted. They are these: Sowter Type 5069. I have 250r on the primaries, bypassed with 220pf, 680r across the secondaries in line with 1n. I have tried adding 200pf from the primaries to earth. The opamp section is completely disconnected and the +/-15v is not connected.

The hiss is only present when there is an input and is not volume dependant. There is none (or almost none) when using opamps.

The schematic for the board is here: http://valwalker.co.uk/misc/4398X2Bv1.pdf

I am guessing that either I need a filter somewhere or this is power supply noise.

Any ideas please?

Thanks

Val

If you were connecting directly to the AOUT exits of the CS4398 you probably will have plenty of HF in the leads. Part of the preamp schematic consists of low pass filters. You may look at the application notes from Cirrus Logic, where you will find proposals. Maybe you can change the opamps to Linear Technology types?
When connecting to a CS4344 (Airport Express, single ended) I had excellent results with the proposed lp-filter consisting of only few passive parts.

[JensH]

With a dual configuration using the CS4398 it should be possible to get close to a dynamic range of 123dB (A), perhaps 122dB (A). I don't think the NE5532 is the main problem (when using op-amps at all). But the resistor values in the filter circuit seems to be fairly high. This could perhaps add some thermal noise. I have not simulated this particular circuit, but I have have a circuit with parallel outputs and it works very well, reducing the noise compared to a single output.
If the measurements on the eBay is done correctly it does indicate that the noise level is higher than what can be achieved. -114.3 (A) is not impressive if the noise level should be understood as dynamic range.

[johnmuir]

Originally Posted by johnmuir

If you were connecting directly to the AOUT exits of the CS4398 you probably will have plenty of HF in the leads. Part of the preamp schematic consists of low pass filters. You may look at the application notes from Cirrus Logic, where you will find proposals. Maybe you can change the opamps to Linear Technology types?
When connecting to a CS4344 (Airport Express, single ended) I had excellent results with the proposed lp-filter consisting of only few passive parts.

In the meantime, I looked into the CS papers. You may find there schematic proposals with the same NE Chip in a 2nd order Butterworth filter(single ended). This will disable the - at least to me - strange use of a volume remote using an offset voltage of +- 15 V from the power supply. This will shift the working ballance in the opamp, I think.