Gustard H20 headphone amp
Dec 5, 2020 at 11:28 PM Post #438 of 646
Using the Ticha 994 discrete opamp, quite pleased it was able to fit.
I have ad the H20 for about 2 years.

The extender socket height is adjusted.
I beveled the top edge of the 994 board to ease insertion.
There are no traces near the edge that faces the heatsink.

994 is an large sonic improvement over the stock opamp, better sound staging and bass slam.
DSCN1346.JPG


DSCN1347.JPG
 
Dec 5, 2020 at 11:39 PM Post #439 of 646
The H20 has untapped potential.
15V power supplies uses the 7815 and 7915 regulators.

These are very old, Motorola introduced these in the late 1970s, there are far better alternatives 40+ years down the road.

I have chosen to replace these with discrete regulators from Sparkos, shown with the golden tape covering

It has taken the H20 to new levels of transparency and smoothness.
More importantly it allowed the H20 to make far better use of the Ticha 994 discrete opamp.

DSCN1356.JPG


-ve 15V Sparkos regulator as it was being mounted.
Fully discrete design
DSCN1348.JPG


Add insulation, essential given the tight tolerances
DSCN1353.JPG


Completed install
DSCN1357.JPG
 
Dec 5, 2020 at 11:45 PM Post #440 of 646
Another mod for the H20 is to replace the primary rectifier diodes with low loss Schottky diodes.
These devices have no switching recovery time and avoids the problem of slow rectifier turn off and recovery.
This causes nasty side effects as the transient current interacts with the transformer inductance causing the output to show ringing instability.

Sonically this makes midrange a lot smoother and removes the high frequency sibilance.

The diodes are IXYS DSS10 chosen for very low leakage, leakage current numbers are a little bit better than the stock 1N5408 unit Gustard used.

DSCN1358.JPG


Some TBD work to better isolate the diodes.
DSCN1359.JPG
 
Last edited:
Dec 6, 2020 at 12:08 AM Post #441 of 646
Been missing my H20 since Covid. Anyone can figure out why the gain switch is not in advantage?
 
Dec 6, 2020 at 12:39 AM Post #442 of 646
The H20 has untapped potential.
15V power supplies uses the 7815 and 7915 regulators.

These are very old, Motorola introduced these in the late 1970s, there are far better alternatives 40+ years down the road.

I have chosen to replace these with discrete regulators from Sparkos, shown with the golden tape covering

It has taken the H20 to new levels of transparency and smoothness.
More importantly it allowed the H20 to make far better use of the Ticha 994 discrete opamp.

DSCN1356.JPG

-ve 15V Sparkos regulator as it was being mounted.
Fully discrete design
DSCN1348.JPG

Add insulation, essential given the tight tolerances
DSCN1353.JPG

Completed install
DSCN1357.JPG
This is outstanding work. If I still had my H20 I would be inquiring about these mods. I'm all for improving an amps performance.
 
Dec 24, 2020 at 6:46 PM Post #444 of 646
The output stage of the H20 is quite beefy, unfortunately the stock H20 has problems driving my Audeze LCD3F.
It lacked top end compared to other amps, overall presentation was muddy and lacked impact.

On a less demanding headphone like the HD800 this was not the case.

H20's reservoir capacitors is one of the bottlenecks, its ripple current capacity is quite low, and it had trouble keeping with the instantaneous demand from the LCD3F

I replaced the stock 3300uF Nichicon KGs with 4700uF TDK Epcos B41231. Ripple current capacity is 4.5x better.

The result was a considerable lift in top end performance on the LCD3F, much better punch and bass slam.
All this without affecting low end performance or causing sibilance issues on the HD800.

17069BFC-C55C-4A3A-902E-867245CAEFA9_1_105_c.jpeg


0.1uF RIFA PHE450 Metallized Polypropylene bypass, replacing the stock WIMA MKS Mylar Caps.
64A85388-E587-4222-8197-F7C8BF106E13_1_105_c.jpeg
 
Last edited:
Dec 24, 2020 at 11:16 PM Post #448 of 646
Wow, @b0bb that is some great work. I can only imagine what the H20 sounds like now. Outstanding job!
 

Users who are viewing this thread

Back
Top