Matching a preamp to high output impedance source

[ainami]

Hi,

I've read that DAC/CD source with a high output impedance requires a linestage with a high input impedance. We have a Jolida CD player which uses 12AX7's for the audio output stage. This results in a very high output impedance for the CD player (stated at 47k ohms).

Is it actually better to use a tube preamp which can supply an input impedance at something like 100k ohms? Or is it still okay to use a solid state preamp (having something like 25k to 50k ohms)?

What are the negatives of using a lower input impedance here? Can someone shed some light on this?

Thanks,

Aaron

[PRR]

The commonly accepted rule is: inputs higher-Z than outputs.

And "higher" can mean 10 or more times higher.

Like 470Ω outputs and 22K inputs.

Some outputs will work when loaded with less than their output impedance.

But if this is a naked 12AX7 output: 12AX7 with several-volt output really does not like a 25K load. Output level will be down, and distortion numbers won't be impressive.

47K is a mighty high output impedance, especially in today's world where 1/100th of that is common. But maybe loading does not bother this amp much. It could even have been designed for modern loads.

[mcs]

Are you sure the 47k is not the minimum load? 47k sounds very high.

But if 47k is the output impedance, then the minimum load should probably be 100-250k.

Best regards,

Mikkel C. Simonsen

[audiousername]

Firstly, how a 12AX7 found its way into the output stage of a CDP is completely beyond me... (well, not really, but I wouldn't use one there)

47K would not be an unreasonable output impedance, a grounded cathode stage with a 12AX7 loaded by 150K would have an output Z of around 70K||150K = 47.7K (assuming a bypassed cathode resistor). A similar output Z could come from a lower plate load resistor but an unbypassed cathode resistor (of course )

The input impedance of the following stage is in parallel with the load resistance of the 12AX7 (hence reducing its value), resulting in lowered maximum voltage swing and additional distortion. You want the input impedance as high as possible to reduce this effect. Also, the high output Z and small current available will cause problems with high frequencies if driving a capacitive load (like long interconnects).

Commonly, the input impedance of the following stage is something like an order of magnitude (10x) greater than the output impedance of the preceeding stage. So, here it would be around 470K. Lower values will work, but will degrade performance as I (and PRR) have said. And keep those interconnects short!

[mcs]

Originally posted by audiousername
47K would not be an unreasonable output impedance, a grounded cathode stage with a 12AX7 loaded by 150K would have an output Z of around 70K||150K = 47.7K

Yes, I agree 47k is not impossible, but I thought they would have used a cathode or plate follower. Or perhaps a tube more suitable for the purpose

Best regards,

Mikkel C. Simonsen

[ainami]

Thanks all for the information! I'm contacting the vendor to see how they expect the unit to be used.