Lightspeed Attenuator a new passive preamp

[georgehifi]

Originally Posted by BFNY

I would also add, that if that amp puts out 50W into 8ohms with an input of 150mV rms, it has a voltage gain of about 42dB, about 16 dB more than the commonly seen amp gain of 26dB.

By adding 40k in front of the series LDR, you are reducing (i.e. shifting) the sensitivity of the whole attenuator range down 15dB, and you will make it work similar to a "normal" gain amp.

You need to look and measure that the I/O impedances of both channels at differing levels remain the same to each other, if not and they have different values of Z in and out (i/o), this will effect the sound quality of each channel hence may/will effect the stereo imaging. It will simulate the same effect as having large different lengths of interconnects for each L and R channel.
That is why quad matched sets have a consistency between channels, they remain equal for both at all levels of listening, unlike some of the pseudo Lightspeed Attenuators that are comming thick and fast. There is no free lunch.

Cheers George

[georgehifi]

Let me re-phrase the simulated effect. It will be the like having a dual passive pot where the right bank/section is say 50k log and the left bank 10k log, same level comming through for each channel, but you can bet they won't sound the same.

Cheers George

[BFNY]

Originally Posted by udailey

I did suggest that. perhaps a LDR attenuator in front of the lightspeed. It wouldnt require matching if it just divides voltage. So you would just use trimmers. Total added cost would only be around 12-15usd and would fix the problem.

My understanding is you recommend a constant fixed value for the LDR then, set by a trimpot? From your experience, can this be set to say, in the range of 20k to 40k ohms?

Again, comments appreciated as to why this approach would be preferred over, say, a fixed high quality tantalum or other boutique resistor of the same value.

[BFNY]

Originally Posted by georgehifi

You need to look and measure that the I/O impedances of both channels at differing levels remain the same to each other, if not and they have different values of Z in and out (i/o), this will effect the sound quality of each channel hence may/will effect the stereo imaging. It will simulate the same effect as having large different lengths of interconnects for each L and R channel.
That is why quad matched sets have a consistency between channels, they remain equal for both at all levels of listening, unlike some of the pseudo Lightspeed Attenuators that are comming thick and fast. There is no free lunch.

Cheers George

Whilst your post makes complete sense on it own, from my point of view it does nothing to address the original problem - the sound being overwhelmingly loud at minimum volume (knob turned completely CCW) as stated here -

Lightspeed Attenuator a new passive preamp

and very useful numbers given here

Lightspeed Attenuator a new passive preamp

So for a moment, let's assume the chap has a perfectly matched quad of LDR parts. It may not be true, but let's say it is.

If with his 42dB voltage gain Marantz PM66KI amp he gets speaker output loudness of 82dB @ 1M at *minimum* volume setting using his 88dB/1W sensitive speakers, how do we solve his problem to reduce the volume level to a reasonable minimum level?

[georgehifi]

Originally Posted by BFNY

the sound being overwhelmingly loud at minimum volume (knob turned completely CCW) as stated here -?

As stated before here
Lightspeed Attenuator a new passive preamp
And here
Lightspeed Attenuator a new passive preamp

The on resistance is much lower in impedance on the NSL32SR2S usually by a factor of three times. You just need to look back a few pages or use the search this thread only tab.

Cheers George

[BFNY]

Originally Posted by georgehifi

As stated before here
Lightspeed Attenuator a new passive preamp
And here
Lightspeed Attenuator a new passive preamp

The on resistance is much lower in impedance on the NSL32SR2S usually by a factor of three times. You just need to look back a few pages or use the search this thread only tab.

Cheers George

Cheers George
Thanks for your reply. I have read those posts. And the ones around them.
Not sure exactly what is driving your responses, they seem disconnected from the subject at hand.

The original poster has quoted he is measuring a minimum impedance value of 30 ohms.

I really don't think you are here to tell us this value can be reduced by a factor of 3 times.
Even if it is so, it really does not help him, as he needs a lot more reduction. I can post the full analysis from my hand written notes if anyone is interested.

[soongsc]

Originally Posted by BFNY

Cheers George
Thanks for your reply. I have read those posts. And the ones around them.
Not sure exactly what is driving your responses, they seem disconnected from the subject at hand.

The original poster has quoted he is measuring a minimum impedance value of 30 ohms.

I really don't think you are here to tell us this value can be reduced by a factor of 3 times.
Even if it is so, it really does not help him, as he needs a lot more reduction. I can post the full analysis from my hand written notes if anyone is interested.

A gain of 40+db in an amp is really high, normally it's like 26db or so. But it seems if more attenuation is needed, that means less maximum gain is needed. Thus. by adding a resistor or constant resistance LDR setting in series with the signal path, you basically reduced the overall gain. How big a series resistor is necessary depends on how much attenuation is necessary.

[georgehifi]

If he is doing 30ohms with the NS32SR2 then then are now probally blown/damaged as they are typicaly above 100ohms @ 20mA max and cannot get down that low with <20mA, and anything above that will kill/damage them in the end, this could be the reason why he has no low volume as the shunts are now probally damaged or gone open circuit.

Cheers George

[soongsc]

Or, if it's possible to match the LDRs and control them up to 50K Ohm.

[AndrewT]

Originally Posted by georgehifi

they are typically above 100ohms @ 20mA max

I only have 8 of the NS32SR2 and they were supplied as selected devices. So mine are not typical.
None of mine exceed 50r at 15mA. I have not damaged any of them.

What you have stated for "typical NS32SR2" is simply not true.