Steinberg UR824, inside photos and Modding

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

JuanV

Member
Joined
Jun 21, 2018
Messages
15
Hi, i did not found any internal photos  of UR824, so i made some.



If I calculate right, there are 16pcs NE5532 and 22pcs NJM2068D op-amp. I would like to change NE5532 to OPA2209 (recommended by TI as replacement) and I don't know replacement for NJM, NJM is seem to be very good opamp.

I believe, the input and output section is similar to MR816, schematic in attachments. MR816 has one 8channel DAC, but UR has 4 stereo DACs.
 

Attachments

  • MR816.pdf
    1.5 MB · Views: 31
I see no one interest for that. So I started alone.

I replaced 1,2,3 for AD8599.

A: Channel 3 Path (original): NJM2068 > NE5532 > NJM2068
B: Channel 1 Path: AD8599 > NE5532 > NJM2068
C: Channel 2 Path: AD8599 > AD8599 >AD8599

A Sound: light, under blanket,
B Sound: brighter and cleaner
C Sound: Little more brighter and cleaner

this replacement is OK without any additional modding, no oscillation, noise floor seems to be same.
so i am satisfied with this replacement.

Tested on SPDIF out (so bypassed internal DAC) to better DAC and B&W Matrix 801.
 

Attachments

  • ModA.jpg
    ModA.jpg
    1.3 MB · Views: 33
JuanV said:
A: Channel 3 Path (original): NJM2068 > NE5532 > NJM2068
B: Channel 1 Path: AD8599 > NE5532 > NJM2068
C: Channel 2 Path: AD8599 > AD8599 >AD8599

A Sound: light, under blanket,
B Sound: brighter and cleaner
C Sound: Little more brighter and cleaner
You should share this with Steinberg's R&D dept. I have no doubt they'll be impressed and thankful for your efforts.
 
JuanV said:
A: Channel 3 Path (original): NJM2068 > NE5532 > NJM2068
B: Channel 1 Path: AD8599 > NE5532 > NJM2068
C: Channel 2 Path: AD8599 > AD8599 >AD8599

A Sound: light, under blanket,
B Sound: brighter and cleaner
C Sound: Little more brighter and cleaner

this replacement is OK without any additional modding, no oscillation, noise floor seems to be same.
so i am satisfied with this replacement.
How did you measure noise exactly?

Note that op amps have an optimal source impedance for which they yield the lowest noise which can be computed from equivalent input voltage noise / current noise. For NE5532 this is 5n / 0.7p = ~7k. In AD8597 its 1.07n / 2.3p = ~500 ohms. These numbers are quite different.
 
If you really want to know if your changes are making a difference either way, make an attenuation cable like this:

attn40mic.png


plug it into the mic channel you want to test, put a 1kHz tone into it and adjust the output level of the signal generator util the channel starts to clip. Then back off the level 2dB. Start a new recording and record 10 seconds of the 1kHz tone. Then hit "pause" on the recording and unplug the attenuator cable at the source end (meaning leave the XLR plugged in because you want to keep the 150 ohm term). Then hit record again to record another 10 seconds of silence. Do NOT change anything in between.

So to summarize, the recording is a total of 20 seconds long with 10 seconds of 1kHz representing the peak level that the device can handle and then 10 seconds of silence. Save it as a wav file uncompressed. If you post that file somewhere, I can post a very high resolution spectrum as described here.
 
RuudNL said:
Did you measure the frequency response before and after?
(Because if it sounds 'brighter' now, you should be able to measure this.)
With opamps, the frequency response is dependant on external components (unless the opamps are used out of their specs); I wouldn't expect any measurable difference in that respect. There may be other aspects of performance (TIM, IMD,...)  that differ, but they may not be easily measurable without specific equipment.
Now, I'm very reluctant to consider seriously comments such as "lifting a blanket"; I tend to think that someone who's undertaking the task of swapping unsocketed IC's will be inclined to justify his efforts with some reward.
 
abbey road d enfer said:
You should share this with Steinberg's R&D dept. I have no doubt they'll be impressed and thankful for your efforts.

coffee snorted through a nose never hurt anyone  ;D

[to be honest I learned my hard lessons in silly opamp swapping on this forum. in the end calibrated instruments and PSU design knowledge were involved]
 
Back
Top