Pabloaguas
Active Member
Posting this carefully because cap threads can go sideways fast.
Up front: any attempt to hijack this into a generic debate about whether capacitor aging is real, whether capacitors can influence sound in circuit, wether capacitor sound changes during burn-in, etc... or whether owners should work on their own gear will be ignored. This thread is simply a report of what I found and what I subjectively heard and measured in my own unit and system.
First: I remain very positive on Schiit and on the Gungnir 2 specifically. I think it is an excellent value, smartly laid out, thoughtfully engineered, and the multiform / DAC8812 implementation is genuinely clever. This is not a manufacturer-bashing post. I’m a repeat Schiit customer and still think they offer a lot for the money, and having owned the Gungnir A2 and now the Gungnir 2, I am absolutely pleased with their house sound. With a 5-year warranty, I would expect them to support legitimate issues.
This is simply one owner’s observations after opening his own unit and experimenting.
Why I Opened It
I had the sense the unit had drifted sonically over time. While inside, I measured several stock UKW electrolytics that appeared tired versus nominal value. Some had fallen materially in capacitance and showed elevated loss behavior (high D / low Q on my meter).
I am not claiming a universal defect or epidemic. I am only reporting what I found in my sample.
There has also been forum chatter over the years regarding some Nichicon audio lines (more often KZ discussions than UKW specifically) aging less gracefully than long-life industrial series in continuous-use equipment.
My suspicion: if some owners perceive a gradual change in sound over time, electrolytic aging in warm always-on service may be one possible contributor.
After finding multiple tired UKWs, I became concerned that any remaining UKWs in critical support positions could become the next weak links, which motivated the later refreshes of UKW.
Mechanical / Install Notes
Internal build quality and PCB layout are very good in the Gungnir. This is not a sloppy or careless product inside.
The challenge is that some replacement capacitors use larger lead diameters than the stock parts. Several plated-through holes in the four-layer PCB are clearly optimized around the original lead size of 0.6mm. My solution for some larger lead diameter replacement caps was side-mounting with very short lead extensions.
It is not beauty-contest work, but it is:
Two-Step Upgrade Process
Important caveat:
Each step changed two things, so sonic observations are attributed to the pair of changes in that step, not to one single component with certainty.
Step 1
Changed:
The eight stock 470 µF / 25 V UKWs in the analog reservoir positions were electrically very large large (overkill) for the duty of driving sixteen AD4898 dual-channel opamps and several of the stock UKW measured around ~300 µF, showing elevated loss. I replaced them with Kaisei 330 µF / 25 V non-polar parts, which fit well physically and are widely regarded as strong performers in local decoupling / reservoir roles.
For the DAC8812 local support positions, Panasonic FR was chosen because it is a modern, long-life, low-impedance series widely used in demanding support/decoupling applications in the audio industry. Yes, it is electrically “faster” than UKW, but in true chip-local support roles that is generally an asset.
Result:
These two upgrades were a large improvement.
Changes heard:
Step 2
Changed:
Black Gate N needs little introduction. It has long been highly regarded for coupling use in audio circuits.
For the precision voltage reference support positions, I intentionally did not choose an ultra-low-ESR part. I chose Würth because it is a modern, long-life industrial electrolytic with moderate impedance / damping behavior closer in concept to the original UKW, rather than something hyper-stiff. Conveniently, Schiit already uses Würth wet electrolytics elsewhere in the DAC, so it was not an alien choice.
Current status:
The BG N are still early in runtime. Even so, they already sound at least as good as the tired UKWs they replaced, likely better.
Current early impressions:
Remaining Plan
I still plan to replace the remaining UKW caps on the USB/ digital input board with Würth long-life parts in the local bulk-support positions.
The reason is simple: once multiple UKWs in the unit measured tired, I no longer trusted the remaining ones to be the best long-term choice.
Final Thoughts
The stock Gungnir 2 is already a strong DAC and a great buy. Nothing here should be read as “needs mods to be good.”
But in my unit, refreshing prematurely aging electrolytics and selectively upgrading these produced a substantial gain.
If others are hearing a gradual softening, reduced bass grip, or loss of sparkle over time, it may be worth considering UKW capacitor aging as one possible factor rather than assuming imagination or system drift. This observed Nichicon UKW aging may seem premature after only one year of use, however, there are 8760 hours in a (non-leap) year, and the UKW are rated for 2000 hours at 85 degrees, so one year in a warm, continuous duty DAC may be their reasonable lifespan.
Up front: any attempt to hijack this into a generic debate about whether capacitor aging is real, whether capacitors can influence sound in circuit, wether capacitor sound changes during burn-in, etc... or whether owners should work on their own gear will be ignored. This thread is simply a report of what I found and what I subjectively heard and measured in my own unit and system.
First: I remain very positive on Schiit and on the Gungnir 2 specifically. I think it is an excellent value, smartly laid out, thoughtfully engineered, and the multiform / DAC8812 implementation is genuinely clever. This is not a manufacturer-bashing post. I’m a repeat Schiit customer and still think they offer a lot for the money, and having owned the Gungnir A2 and now the Gungnir 2, I am absolutely pleased with their house sound. With a 5-year warranty, I would expect them to support legitimate issues.
This is simply one owner’s observations after opening his own unit and experimenting.
Why I Opened It
I had the sense the unit had drifted sonically over time. While inside, I measured several stock UKW electrolytics that appeared tired versus nominal value. Some had fallen materially in capacitance and showed elevated loss behavior (high D / low Q on my meter).
I am not claiming a universal defect or epidemic. I am only reporting what I found in my sample.
There has also been forum chatter over the years regarding some Nichicon audio lines (more often KZ discussions than UKW specifically) aging less gracefully than long-life industrial series in continuous-use equipment.
My suspicion: if some owners perceive a gradual change in sound over time, electrolytic aging in warm always-on service may be one possible contributor.
After finding multiple tired UKWs, I became concerned that any remaining UKWs in critical support positions could become the next weak links, which motivated the later refreshes of UKW.
Mechanical / Install Notes
Internal build quality and PCB layout are very good in the Gungnir. This is not a sloppy or careless product inside.
The challenge is that some replacement capacitors use larger lead diameters than the stock parts. Several plated-through holes in the four-layer PCB are clearly optimized around the original lead size of 0.6mm. My solution for some larger lead diameter replacement caps was side-mounting with very short lead extensions.
It is not beauty-contest work, but it is:
- mechanically solid
- electrically sensible
- securely anchored
- short lead length maintained
- Lead extensions were kept minimal.
- I hear no evidence of degradation from the small added length.
- Existing local film / small HF bypass parts remain in place.
- Those local parts should continue handling RF / HF shunting duties.
- This was not random spaghetti wiring.
Two-Step Upgrade Process
Important caveat:
Each step changed two things, so sonic observations are attributed to the pair of changes in that step, not to one single component with certainty.
Step 1
Changed:
- Main analog output-stage / op-amp local reservoir caps → Audio Note Kaisei NP
- DAC8812 local supply decoupling caps → Panasonic FR
The eight stock 470 µF / 25 V UKWs in the analog reservoir positions were electrically very large large (overkill) for the duty of driving sixteen AD4898 dual-channel opamps and several of the stock UKW measured around ~300 µF, showing elevated loss. I replaced them with Kaisei 330 µF / 25 V non-polar parts, which fit well physically and are widely regarded as strong performers in local decoupling / reservoir roles.
For the DAC8812 local support positions, Panasonic FR was chosen because it is a modern, long-life, low-impedance series widely used in demanding support/decoupling applications in the audio industry. Yes, it is electrically “faster” than UKW, but in true chip-local support roles that is generally an asset.
Result:
These two upgrades were a large improvement.
Changes heard:
- much larger soundstage
- more stable 3D image placement
- stronger front-to-back layering
- tighter, more controlled bass
- smoother presentation without loss of detail
- more air
- more in-room realism vs flat speaker-bound presentation
- lower haze / grain
Step 2
Changed:
- Output coupling caps → NOS Black Gate N
- Precision reference support decoupling caps (LT1236 reference area) → Würth long-life electrolytics
Black Gate N needs little introduction. It has long been highly regarded for coupling use in audio circuits.
For the precision voltage reference support positions, I intentionally did not choose an ultra-low-ESR part. I chose Würth because it is a modern, long-life industrial electrolytic with moderate impedance / damping behavior closer in concept to the original UKW, rather than something hyper-stiff. Conveniently, Schiit already uses Würth wet electrolytics elsewhere in the DAC, so it was not an alien choice.
Current status:
The BG N are still early in runtime. Even so, they already sound at least as good as the tired UKWs they replaced, likely better.
Current early impressions:
- slightly darker tonal balance
- bass currently somewhat light
- promising spatial qualities
- no ugly or painful “Black Gate burn-in disaster” phase some report
Remaining Plan
I still plan to replace the remaining UKW caps on the USB/ digital input board with Würth long-life parts in the local bulk-support positions.
The reason is simple: once multiple UKWs in the unit measured tired, I no longer trusted the remaining ones to be the best long-term choice.
Final Thoughts
The stock Gungnir 2 is already a strong DAC and a great buy. Nothing here should be read as “needs mods to be good.”
But in my unit, refreshing prematurely aging electrolytics and selectively upgrading these produced a substantial gain.
If others are hearing a gradual softening, reduced bass grip, or loss of sparkle over time, it may be worth considering UKW capacitor aging as one possible factor rather than assuming imagination or system drift. This observed Nichicon UKW aging may seem premature after only one year of use, however, there are 8760 hours in a (non-leap) year, and the UKW are rated for 2000 hours at 85 degrees, so one year in a warm, continuous duty DAC may be their reasonable lifespan.
Last edited:
