Hello!
I have in my possession a UR22mkii audio interface. When I bought it many, many moons ago, I did not realise that the output impedance of the headphone amp was measured at 91 ohms!!! (According to Julian Krausse's measurements, which I trust).
I recently took apart my UR22mkii in order to redo the solder joints of the headphone socket. It was becoming a bit wiggly and I decided to do up the joints with some high quality leaded solder. The headphone socket has basically no mechanical support and just kind rests on the chassis! Hilarious! This pretty much guarantees that this solder joints are going to break. When will manufacturers learn that solder is not glue?
Anyways, while I was in here I decided to trace the output section of the headphone amp. I only did it on the main front board and not where the volume and balance controls live above it. (It's connected the front board via a ribbon cable). Attached below is the output buffer present. It's a very basic inverting opamp setup with some capacitors here and there for frequency response shaping. I wasn't able to measure the small caps that were part of this system as I'd have to take them out of the the circuit and it's all SMD stuff we're dealing with here, so I couldn't be bothered.
There is a FET switching system attached to the right output channel that probably shunts the output ground when the jack is unused, probably because they saved 5 extra cents by using an unswitched headphone jack and two transistors vs. a switched headphone jack. I didn't trace it.
To me, the issue seems blatantly obvious. 90 ohms resistance in series with the output? Was the designer for this insane? I'm thinking of just jumpering one those 181R resistors. Should I pay attention to those unmeasured caps and change their values if I do this? I imagine this will mess with the frequency response quite a bit.
Thanks for any help!
I have in my possession a UR22mkii audio interface. When I bought it many, many moons ago, I did not realise that the output impedance of the headphone amp was measured at 91 ohms!!! (According to Julian Krausse's measurements, which I trust).
I recently took apart my UR22mkii in order to redo the solder joints of the headphone socket. It was becoming a bit wiggly and I decided to do up the joints with some high quality leaded solder. The headphone socket has basically no mechanical support and just kind rests on the chassis! Hilarious! This pretty much guarantees that this solder joints are going to break. When will manufacturers learn that solder is not glue?
Anyways, while I was in here I decided to trace the output section of the headphone amp. I only did it on the main front board and not where the volume and balance controls live above it. (It's connected the front board via a ribbon cable). Attached below is the output buffer present. It's a very basic inverting opamp setup with some capacitors here and there for frequency response shaping. I wasn't able to measure the small caps that were part of this system as I'd have to take them out of the the circuit and it's all SMD stuff we're dealing with here, so I couldn't be bothered.
There is a FET switching system attached to the right output channel that probably shunts the output ground when the jack is unused, probably because they saved 5 extra cents by using an unswitched headphone jack and two transistors vs. a switched headphone jack. I didn't trace it.
To me, the issue seems blatantly obvious. 90 ohms resistance in series with the output? Was the designer for this insane? I'm thinking of just jumpering one those 181R resistors. Should I pay attention to those unmeasured caps and change their values if I do this? I imagine this will mess with the frequency response quite a bit.
Thanks for any help!
The NJM4580D has limited output current capability. The resistance is there to limit load current. You would have to change the opamp to reduce output impedance..
Thanks. I have to wonder what the design philosophy behind this output section was... When you're placing 90 ohms of series resistance onto the output of a a headphone amp, you've gotta realise you've done something wrong? Maybe Steinberg needed to dump massive amounts of 4580 chips haha. They would not be chip of choice if I was designing a headphone amp from the ground up.
No, it's a common (and of course questionable) way of creating a simple/cheap "universal" HP output which also can take some abuse (output full short short gives still manageable current, even when on USB bus power). With 90 Ohms, you can have roughly the same power delivered to the load within a range of 30 to 300 Ohms or so, based on nominal impedance. An any bass peak in the impedance is giving us bass boost around resonance for free ;-)
IME one can reduce those series resistors down to 22R or so provided you never ever short/overload the output, which may result in anything from recoverable malfunction to damage.
Excuse my ignorance, how do other headphone amps achieve such protection with much lower outputs? Even other audio interfaces, which are not necessarily designed with headphone performance in mind, frequently get to 10 ohms of output impedance? (And of course some do have proper headphone amps with an output impedance below 2 ohms) It would seem to me that this is a solved problem that steinberg did not decide to implement for whatever reason.
True, a low output impedance output is a solved problem technically, but it adds circuit complexity and cost -- but the designers sure had tight constraints wrt cost, space, complexity, robustness. This is an affordable entry-level audio interface, corners had to be cut and the 90Ohms HP circuit is OK for monitoring duties. See it as an ancillary output, and a pretty much bullet-proof one.
I can accept that alright. It makes sense. This is just me, but the 90 ohms output impedance feels like an unforced error. Even cheap behringer interfaces measure at 10 ish ohms. Apologies if it sounds like I'm arguing with you as if you were the one approving the design on this product - I'm just really hung up on the fact that this is basically the highest output impedance headphone amp on seemingly any modern usb interface!