My design L20D IRS2092+IRFI4020H 200W8R

We were earlier discussing the possibility of increasing the 3.3k input impedance
The "low" input impedance (and I put "low" in quotes because 3.3k is plenty high for any reasonably designed pre-amp output) is inherent to the reference design. If you want a higher input Z then you add an input stage to do it . . . it takes at most a couple dollars in parts.

Any reasonably designed "passive pre-amp" should have an output buffer anyway . . . it is not good design to make the performance of any component critically dependent on the load provided by the next device (including cables) in the chain.
 
Thanks Bernie for bringing this to attention. 3.3k impedance explains the rather large 22uF input cap.
I plan to use the amp for a sub. What is the suggestion : leave the electrolytic input cap ? replace by MKP ?
22µF is a lot of energy. Anyone having trouble with noise in the speaker at turn on or off ?
 
The "low" input impedance (and I put "low" in quotes because 3.3k is plenty high for any reasonably designed pre-amp output) is inherent to the reference design. If you want a higher input Z then you add an input stage to do it . . . it takes at most a couple dollars in parts.

Any reasonably designed "passive pre-amp" should have an output buffer anyway . . . it is not good design to make the performance of any component critically dependent on the load provided by the next device (including cables) in the chain.

We'll have to disagree on this. I have a B1 buffer and a passive preamp with LDR volume control and the B1 though good can be heard to impart some coloration. Same with a tube buffer I have. I'm running the passive pre direct into a hifimediy T3 (20k input) with no issues at all. Ditto with my UcD400hg monos. Saying any decent passive pre should have a buffer makes no sense to me.

So the question remains - can the L20d input Z be increased without a major rework ljm?
 
I'm running the passive pre direct into a hifimediy T3 (20k input) with no issues at all.
If your "passive pre-amp" is designed to work properly with a 20k load it will work properly with a 20k load. It will not work properly with a 5k or 10k or 40k load. There is, of course, nothing wrong with "application specific" designs . . . but if you want your "passive pre-amp" to work with a variety of amplifiers it will need an output buffer to provide the proper load impedance to the "passive" stage and a low Z output to the amp.

So the question remains - can the L20d input Z be increased without a major rework ljm?
And the answer remains - no, you cannot significantly change the input impedance of the 2092 (reference circuit) . . . if you need a higher input impedance you must add an input (buffer) stage (which is what other vendors, like classDaudio, do), the 2092 alone cannot provide it. It works fine, however, with anything that has an output impedance no higher than a couple hundred ohms (essentially any decently designed pre-amp from the past 30 years).

Two other thoughts . . . if your pre-amp is AC coupled (has an output blocking cap) or has a servo loop to control output offset and you never hot-plug it you can remove (bypass) the input blocking cap on the L20D board; and do note that the L20D is, like all 2092 "reference" designs, an inverting amp. If you want non-inverted output you'll need an inverting input buffer anyway (or remember to reverse your speaker leads).
 
I need to correct any problem.

L20D input impedance is 100 K.

And the coupling capacitors is in the input impedance, so I don't suggest behind less input capacitance.
 

Attachments

  • ZBEL.jpg
    ZBEL.jpg
    35 KB · Views: 2,057
If your "passive pre-amp" is designed to work properly with a 20k load it will work properly with a 20k load. It will not work properly with a 5k or 10k or 40k load. There is, of course, nothing wrong with "application specific" designs . . . but if you want your "passive pre-amp" to work with a variety of amplifiers it will need an output buffer to provide the proper load impedance to the "passive" stage and a low Z output to the amp.

Hi dewardh

Thanks for your take on the issue. If I'm making passive preamps for the general market, yes it should match with a wide range of amps and a buffer would be called for. But I'm only concerned here with an amp whose impedance matches my great sounding LDR pre. Using a buffer would be adding another stage and a deal breaker. I suspect there are many others like me who prefer high Z amps.

There's yet another way to resolve any pre-power mismatch for me, and that is to build the LDR volume control right next to the LDR inputs. I've done this before, with the volume knob/control circuit in an external box for monoblock amps.

Although I have output caps on my DAC, I wouldn't want to DC couple the L20D as I may in future use it with a source that's DC coupled too.

Appreciate your pointing out the 2092 phase inversion. That can be easily taken care of by inverting the input or, as you say, speaker lead connections inside the amp.

Can ljm kindly confirm if this particular L20D design does a phase inversion at the output?
 
Last edited:
Hi ljm_ljm ,

Your Design is an inverting Amp. There ist no Signal-Difference between Note R7-R8 and GND.
An externally hosted image should be here but it was not working when we last tested it.


HI.
HI. Very nice of you to answer.

I'm going to see this circuit, that you should be calculated is correct.

Reverse phase input virtual ground, so input impedance should be 1 / (1/100 K + 1/3.33 K)

Thank you for your correct.