Ian's output stage was really designed to use OPA1622 in the differential summing position. Unfortunately, OPA1622 is more an ultra-low distortion headphone amp type opamp, and besides it produces extra distortion on 8-pin DIP adapters since its ground pin really needs to be grounded for lowest distortion. (You might try running a wire from the ground pin down to the board ground plane by the shortest possible path.)
Regarding Ian's dual ES9038Q2M dac and the std i/v output stage board, RF from the dac board causes audible distortion in (all) the opamps. Ian devised a diy filter to see if that would help, but I guess it wasn't that successful.
I was able to get the output stage board to sound better by putting a ribbon cable between the dac board and the output stage board and folding the ribbon over on itself sideways into a position that produced least brightness. Started to sound much better that way.
OPA1612 can be used in the differential summing position, but resistors are rather low value for minimum distortion (apparently too much focus on low-noise vs low-distortion).
P.S. Do not expect burn-in of the opamps. Maybe some burn-in for your ears getting used to the sound of the opamps.
Regarding Ian's dual ES9038Q2M dac and the std i/v output stage board, RF from the dac board causes audible distortion in (all) the opamps. Ian devised a diy filter to see if that would help, but I guess it wasn't that successful.
I was able to get the output stage board to sound better by putting a ribbon cable between the dac board and the output stage board and folding the ribbon over on itself sideways into a position that produced least brightness. Started to sound much better that way.
OPA1612 can be used in the differential summing position, but resistors are rather low value for minimum distortion (apparently too much focus on low-noise vs low-distortion).
P.S. Do not expect burn-in of the opamps. Maybe some burn-in for your ears getting used to the sound of the opamps.
Last edited:
That's essentially voltage mode if a Sabre dac. Distortion goes up considerably.
I would really question if OPA1612 on dip8s are legitmate from Ali. Certainly doesn't sounds like the sound signature I had.
I would agree though, the bisesik transformers have finally put my DAC next to my benchmarks. T+A Dac 8 DSD and a Phison PD2.
Thanks for the input guys....Simon, Mark and Seb.
Well maybe last night wasnt a good reference....but today was sounding better and my package from Ian came so I added in the fifo ultimate. Got it all rigged up running sync with the MCLK u.fl cable
Sounding nice now. Still with opa1612 in place. I may try the 5532 back in position for comparison.
My 1612 are on dip8 adaptors.....seem legit to me and have come with a recommendation from another user.
Simon thanks for the offer. Maybe in the new year I shall take you up on that . Although the lundahl are maybe a stretch for me.
I added some extra caps on the fifo psu inputs.....no sypercaps by any means ...2200uF Pana FC and a 470uf oscon...just because they were there!
Happy xmas guys.
Well maybe last night wasnt a good reference....but today was sounding better and my package from Ian came so I added in the fifo ultimate. Got it all rigged up running sync with the MCLK u.fl cable
Sounding nice now. Still with opa1612 in place. I may try the 5532 back in position for comparison.
My 1612 are on dip8 adaptors.....seem legit to me and have come with a recommendation from another user.
Simon thanks for the offer. Maybe in the new year I shall take you up on that . Although the lundahl are maybe a stretch for me.
I added some extra caps on the fifo psu inputs.....no sypercaps by any means ...2200uF Pana FC and a 470uf oscon...just because they were there!
Happy xmas guys.
I thought it would be a good time to post about my experience with Ian's stuff, to give back to all the pioneers earlier in the thread that allowed me to get as far as I have with the Dac and help others. I would particularly like to thank Greg for his insight/experimentation and Ian for the original designs.
To set the scene, I like DSD and have been running either a discrete DSD converter with a discrete output stage (T+A Dac8 DSD) and a Phison PD2 (AK4490 - in Bypass). I tend to run Class D amps, as my preference, and recently switched from Ncore's to Iceedge 1200as2.
My current setup is:
Bisesik Output transformers in balanced configuration
Ian's 9038 DAC with three separate power rails
Ian's Fifopi bypassing the on-board LDO's, utilising NDK clock's squashed down by a layer of Sorbothane and the 45 clock running on its own rail
Recieverpi bypassing on-board LDO's on its own rail
JL Sounds USB to I2s Viii both sides powered on their own rail
I am using two power supplies, a Shanti (two 5v rails from 2x 3.3F ultracaps on each rail) and 6 battery rails with 2x325F ultracaps on each rail.
Initially with the Dac, Fifopi (bypassing LDO, clocks from Ian), Raspberrypi and opamp stage (opamps from Ian) all powered by the battery pack the sound wasn't great (4-5/10). Flat presentation, narrow soundstage and an odd mid-range sheen. Moving to sync mode helped a little, maybe a solid 5/10.
Adding in the OPA1612, smoothed out the tone and timbre issues but wasn't as big of a jump as I expected. Still 5/10 for me personally albeit more listenable.
I then added in the NDK clocks, and this was a significant jump to my tastes. Increased depth, the darkness around instruments, more solid centre soundstage, increased dynamics but still a narrow soundstage. They took a while to burn-in, but a solid 6/10 and much more promising now.
At this point I removed the Pi and replaced it with the JL sounds board. This improved the signal quality significantly when comparing the DPLL and high-rate DSD (previously limited to DSD128 - pi limitations) was now able to go into true-sync comfortably up to DSD512. This was a big jump, not sure why and controvertial, however tone/timbre increased significantly regardless of music format. 7/10
There is enough writeups on Bisesik's transformers, but implementing them in and I finally felt this was becoming a high-end dac. 7.5/10
The most dramatic impact for me and I recommend everyone to give it a try on atleast one rail. Was to follow Ian's instructions on a hybrid ultracap supply. This shotgunned the dac beyond anything I have personally had, and I never expected these mobile chips to be competitive with my references. This has finally increased the width and height of the soundstage (a real concern until then), provided even more resolution and transformed the overall sound into something very natural and organic. 9/10
I like it so much that I am buying spare back-up components in the GB.
To set the scene, I like DSD and have been running either a discrete DSD converter with a discrete output stage (T+A Dac8 DSD) and a Phison PD2 (AK4490 - in Bypass). I tend to run Class D amps, as my preference, and recently switched from Ncore's to Iceedge 1200as2.
My current setup is:
Bisesik Output transformers in balanced configuration
Ian's 9038 DAC with three separate power rails
Ian's Fifopi bypassing the on-board LDO's, utilising NDK clock's squashed down by a layer of Sorbothane and the 45 clock running on its own rail
Recieverpi bypassing on-board LDO's on its own rail
JL Sounds USB to I2s Viii both sides powered on their own rail
I am using two power supplies, a Shanti (two 5v rails from 2x 3.3F ultracaps on each rail) and 6 battery rails with 2x325F ultracaps on each rail.
Initially with the Dac, Fifopi (bypassing LDO, clocks from Ian), Raspberrypi and opamp stage (opamps from Ian) all powered by the battery pack the sound wasn't great (4-5/10). Flat presentation, narrow soundstage and an odd mid-range sheen. Moving to sync mode helped a little, maybe a solid 5/10.
Adding in the OPA1612, smoothed out the tone and timbre issues but wasn't as big of a jump as I expected. Still 5/10 for me personally albeit more listenable.
I then added in the NDK clocks, and this was a significant jump to my tastes. Increased depth, the darkness around instruments, more solid centre soundstage, increased dynamics but still a narrow soundstage. They took a while to burn-in, but a solid 6/10 and much more promising now.
At this point I removed the Pi and replaced it with the JL sounds board. This improved the signal quality significantly when comparing the DPLL and high-rate DSD (previously limited to DSD128 - pi limitations) was now able to go into true-sync comfortably up to DSD512. This was a big jump, not sure why and controvertial, however tone/timbre increased significantly regardless of music format. 7/10
There is enough writeups on Bisesik's transformers, but implementing them in and I finally felt this was becoming a high-end dac. 7.5/10
The most dramatic impact for me and I recommend everyone to give it a try on atleast one rail. Was to follow Ian's instructions on a hybrid ultracap supply. This shotgunned the dac beyond anything I have personally had, and I never expected these mobile chips to be competitive with my references. This has finally increased the width and height of the soundstage (a real concern until then), provided even more resolution and transformed the overall sound into something very natural and organic. 9/10
I like it so much that I am buying spare back-up components in the GB.
Sure thing and no worries.
Develop ultra capacitor power supply and LiFePO4 battery power supply
Merry Christmas
Develop ultra capacitor power supply and LiFePO4 battery power supply
Merry Christmas