I am now starting my final " leftovers budget" build. After collecting "stuff" and building great gear since 30 years ago. I have now amassed leftovers that I will now use for a "final" use all of it build for a headphone amp. Then that's it. My eyes are too weak for these SMT stuff these days.
Here is what I have leftover over the years.
Improved Sulzer Power Supplies removed from some dead CD players.
AD844 removed from Assemblage DAC
AD811 removed from modded POOGED CD player
Stepped resistor Attenuator replaced by a ladder unit I purchased off Ebay
etc. etc.
I am going to put together a composite amp using the same technique in the 744-811 duo using the feed from the 844 before the output section into the 811 buffer which I believe can put out 150ma.
I can put a current limit resistor before or after the feedback network tap.
I see circuits using around 20 ohms after the feedback but I have heard that I could put around 100 ohms before the feedback and retain the low output impedance but retain current limit.
These are to drive HD598s
Which is the better way? Any suggestions
Here is what I have leftover over the years.
Improved Sulzer Power Supplies removed from some dead CD players.
AD844 removed from Assemblage DAC
AD811 removed from modded POOGED CD player
Stepped resistor Attenuator replaced by a ladder unit I purchased off Ebay
etc. etc.
I am going to put together a composite amp using the same technique in the 744-811 duo using the feed from the 844 before the output section into the 811 buffer which I believe can put out 150ma.
I can put a current limit resistor before or after the feedback network tap.
I see circuits using around 20 ohms after the feedback but I have heard that I could put around 100 ohms before the feedback and retain the low output impedance but retain current limit.
These are to drive HD598s
Which is the better way? Any suggestions
Yes there are a couple of concepts mixed together
current limit is different from series output impedance - although the latter can do some of the 1st
current limit is usually used to keep either the chip or the load from burning up under fault conditions
series impedance is usually needed with faster feedback amps to isolate them from the cable parasitic C load near (well a decade below to be safe) their feedback unity gain intercept frequency - can be up to 100 MHz for CFA
a series resistor in the output can do some of both jobs – but is limited by interacting with load impedance, giving possibly audible frequency response changes
the HD598 Z is lumpy
but our hearing sensitivity is low for low frequency bumps:
ABX Amplitude vs. Frequency Matching Criteria
together both of the graphs suggest to me that ~ 1 dB will be below audibility for the 598 low frequency Z bump
so very approximately ~30 Ohm or less series output R should be fine with HD598
and common load isolating R recommendation for the AD811 or similar CFA would be 10 or more Ohms - so 20 Ohms looks like an OK value for a series output R
current limiting requires some idea of what you need to pass in signal and what you want to protect from in fault/overload conditions – and is a function of your supply V too - 20 Ohms series output R may not do the job alone
current limit is different from series output impedance - although the latter can do some of the 1st
current limit is usually used to keep either the chip or the load from burning up under fault conditions
series impedance is usually needed with faster feedback amps to isolate them from the cable parasitic C load near (well a decade below to be safe) their feedback unity gain intercept frequency - can be up to 100 MHz for CFA
a series resistor in the output can do some of both jobs – but is limited by interacting with load impedance, giving possibly audible frequency response changes
the HD598 Z is lumpy
An externally hosted image should be here but it was not working when we last tested it.
but our hearing sensitivity is low for low frequency bumps:
ABX Amplitude vs. Frequency Matching Criteria
together both of the graphs suggest to me that ~ 1 dB will be below audibility for the 598 low frequency Z bump
so very approximately ~30 Ohm or less series output R should be fine with HD598
and common load isolating R recommendation for the AD811 or similar CFA would be 10 or more Ohms - so 20 Ohms looks like an OK value for a series output R
current limiting requires some idea of what you need to pass in signal and what you want to protect from in fault/overload conditions – and is a function of your supply V too - 20 Ohms series output R may not do the job alone
Last edited:
Hey thanks for pointing out the lumpiness of the Z.
Given that is the situation, would it be a good idea to rig up an output impedance selector that can vary the output impedance? The feedback would be taken before the output resistor.
Say 120 ohms, 60 ohms and 20 ohms. The larger values would minimize the variances and fatten the sound with a loss of efficiency and the lower values would make the phones sound thinner but have less losses. With rail supplies of +- 15V I have lots of gain to spare and can simply bump up if necessary.
I have lots of left over toggle and selector switches from which to choose from.
Given that is the situation, would it be a good idea to rig up an output impedance selector that can vary the output impedance? The feedback would be taken before the output resistor.
Say 120 ohms, 60 ohms and 20 ohms. The larger values would minimize the variances and fatten the sound with a loss of efficiency and the lower values would make the phones sound thinner but have less losses. With rail supplies of +- 15V I have lots of gain to spare and can simply bump up if necessary.
I have lots of left over toggle and selector switches from which to choose from.
Last edited:
- Status
- Not open for further replies.