Hypex NC122MP repair : blown SMD component L15

[somebodyelse]

Whether it's 10k or 50k won't change the slope. That is an independent parameter.
The output impedance of a 10k pot is lower than 10k so the impedance ratio is better than 4.7:1. If you are worried by the ratio, or need long cables between the preamp and power amp, then get an active preamp not a passive attenuator.

 

[wyup]

Whether it's 10k or 50k won't change the slope. That is an independent parameter.
The output impedance of a 10k pot is lower than 10k so the impedance ratio is better than 4.7:1. If you are worried by the ratio, or need long cables between the preamp and power amp, then get an active preamp not a passive attenuator.

Ok for the slope.
I just want to improve the dac's pot with a passive stepped pre for better sound quality at low volumes, since I have lots of gain on the amp, the source inpedance is low and the amp input is high, making a good candidate for passive in between. I don't need an active preamp, source voltage is higher than max power on the amp, maybe an output buffer in the passive preamp but if I really need it, I prefer to keep the signal pure. I'll be using short cables.

The pot is variable output impedance but I'm using such low volumes that the impedance must be close to the max.
Can a ~5 amp to pre ratio hurt the amp performance?

 

[somebodyelse]

IIRC output impedance of the pot is the parallel combination of the series and shunt parts of the attenuator (assuming the output impedance of the source is small), so lower than either part. Worst case is when they're equal, so 1/4 of the nominal value. For 10k that's the parallel combination of 2 x 5k giving 2k5. That makes your ratio look more respectable.

 

[wyup]

IIRC output impedance of the pot is the parallel combination of the series and shunt parts of the attenuator (assuming the output impedance of the source is small), so lower than either part. Worst case is when they're equal, so 1/4 of the nominal value. For 10k that's the parallel combination of 2 x 5k giving 2k5. That makes your ratio look more respectable.

Do you mean the Aune pot or the relay-shunt ebay attenuator?

 

[somebodyelse]

I forgot it was the relay variety not an ordinary pot. The general point stands though - the output impedance is lower than 10k so the ratio isn't as bad as you think.

 

[wyup]

I wanted to add that I contacted Hypex and they explained to remove two resistors on the board, R76 and R162, to bypass the input stage gain, about 13.5db, keeping its input impedance. This is enough for me. I wish I didn't have to send the amp for service and be able to change some jumper or dip switch.
Bypassing the input buffer altogether is explained in the main datasheet, but that involves removing and changing more resistors and reducing the input impedance to 1.8k, not practical. Then Hypex sent me a doc with a calculation formula for custom gain adjustement of those resistors' values.

 

[Rick Sykora]

I wanted to add that I contacted Hypex and they explained to remove two resistors on the board, R76 and R162, to bypass the input stage gain, about 13.5db, keeping its input impedance. This is enough for me. I wish I didn't have to send the amp for service and be able to change some jumper or dip switch.
Bypassing the input buffer altogether is explained in the main datasheet, but that involves removing and changing more resistors and reducing the input impedance to 1.8k, not practical. Then Hypex sent me a doc with a calculation formula for custom gain adjustment of those resistors' values.

Pleased to hear Hypex was helpful. That has been my my experience as well.

Thanks for sharing their response! It should be helpful for others with comparable issues.

 

[Tangband]

I wanted to add that I contacted Hypex and they explained to remove two resistors on the board, R76 and R162, to bypass the input stage gain, about 13.5db, keeping its input impedance. This is enough for me. I wish I didn't have to send the amp for service and be able to change some jumper or dip switch.
Bypassing the input buffer altogether is explained in the main datasheet, but that involves removing and changing more resistors and reducing the input impedance to 1.8k, not practical. Then Hypex sent me a doc with a calculation formula for custom gain adjustement of those resistors' values.

I have also checked with the Hypex support yesterday and removing those two resistors indeed makes the buffer *1 and the gain will be 13,5 dB less - keeping the 47 kohm input impedance. Much easier than to bypass the whole buffer stage.