To me just knowing that you could coax music out of that pile of stuff is magic.
1 Like
I just had the transformers on the analog board replaced and wondered if anyone else who has done this noticed a substantial drop in volume from the exchange? I would say a good 25% drop in output as I have had to turn my preamplifier up as well as the DirectStream.
The install itself went fine as the technician explained what what he was doing as he did everything.
Sounds abnormal.
Turbo
Odd. I had my DAC modded by Wayne this weekend and I donât hear a significant change in volume levels.
Might you accidentally now have the 20dB attenuator engaged? There would be an âLâ next to the small volume numbers on the display. Off of the top of my head I donât remember which button on the remote changes it, perhaps something like âLevelâ?
1 Like
I just checked and the âLâ was not on by the volume level digital number. Turning that on really dropped the volume. The sound is good just less volume that I need to increase the volume at the preamplifier and / or Direct Stream but the largest increase has to be at the preamplifier to make up for the drop in the DirectStream output. I am running balanced from the DirectStream to the preamplifier.
The transformer mod shouldnât change the volume, but a few other people have asked the same question. Perhaps they will chime in.
Itâs also the case that cleaner sound doesnât sound quite as loud: you are missing the distortion which adds energy which raises the volume (a little.) Put another way, cleaner sound encourages you to turn the volume up because it doesnât tire the ears as much or as fast.
4 Likes
Is anyone helping (of course for $$) to install new transformers for folks like me, who are incapable to soldering and modding without burning the house? 
1 Like
Wayne @wctaudio did mine and many others. Excellent work reasonable price. You can send the whole unit or just the analog board if heâs still doing them.
1 Like
Maybe the chart above is saying the same thing but did go out and look at the specifications for the two transformers.
The two things I noticed was the lower DC Resistance which is now 116 Ohms compared to the previous DC Resistance of 150 Ohms.
The other was the Maximum Input Voltage. The new transformer is only 7.5Vrms while the previous one,'s maximum was 20Vrms.
Would these two specification changes cause any changes in the output?
Marked as products compatible with edcor 4400 at the company stand in Munich
Itâs good to check the specs but no, those differences arenât significant for the way the DS uses them. The inputs to the transformers already have current sharing resistors so the difference in DC Ohms doesnât matter. The input and output of the transformer can accommodate the output of 1.414VRMS of the DS. The DS was designed for both of these transformers and the XS4400 is the transformer used in the DS Mk II.
I did talked with the company and they indicated only a slight reduction in volume due to core windings but not like I experienced. I asked about a possibility of a defective transformer and they said that isnât likely with a volume reduction.
Hey Turbo,
Thanks for all the great information regarding mods.
I was interested in doing the Vocm mod and had some questions.
In a previous post, you mentioned using a 6.3v cap. for the Vocm mod.
From this post, you recommend 10-16v range.
Is there a reason to not use a voltage lower than 10?
I also have poor understanding of the circuit layout and do not know where the Vocm output is located. Does the negative lead of the capacitor go on the location marked with the red arrow in the picture below?
1 Like
That would be correct.
You could probably use a 6.3v there if you wanted; donât think that would be a problem.
Which transformer specifically?
Ok, I had to re-do the filter (oopsy). Now its working as intended and sounding absolutely marvelous. Can hear absolutely every minute detail, new ones even in very familiar recordings. Sabrina Malheiros âEquilibriaâ album for example- chock full of complex textures, sounds, rhythms woven together masterfully IMO by real musicians. All the individual sounds are maintained distinct, even tiny tiny sounds, no matter what else is going on. It doesnât jumble together in complex passages. It never sounds distorted on vocals (or anything else). The sound is freed from the speakers, as mentioned before. To me that says a LOT- itâs maintaining ultra low level detail, which is not easy to do. Can hear that I2S input sounds better than SPDIF⌠Etc.
Still just a prototype filter, a first iteration at 3rd order CLC. And only 1 OA/channel at the moment. But its working great IMO, in terms of filtering (very good), step response (looks ~ perfect) and 20KHz BW is ~ workable for the moment (-0.8dB). Weâll work on reducing that for the next iteration. OTOH somehow itâs proven/shown near flat to 20Hz The transformer, I suspectâŚ
Full speed ahead. Will we have to wind our own air core inductors ultimately to get the best result? Weâll see 
I donât know how much a common ferrite wirewound inductor affects the audio band performance; Iâm concerned about possible core losses and hysteresis effects. We donât want anything messing with our precious signal. Donât know though, only 1 way to find out ultimately I guessâŚ
352K8 441Hz square wave response (single ended):
Zoom 100x of leading edge of same:
1KHz (176K FS) amplitude basis:
20KHz amplitude for Freq response comparison- down ~ -0.8dB:
20Hz amplitude for Freq response comparison, ~ flat to 20Hz (golf clap):
3 Likes
Awesome stuff T! So you have a CLC after the FPGA buffers and thats it? You are bypassing the transformer completely? Whats the -3dB on the filter? Whats the group delay look like? Curious minds want to know.
1 Like
Thanks DG.
And well, no, the FPGA is not suitable to drive the passive filters and output trans. We have to have amplitude and power to drive them properly. The ludicrous speed (/Space Balls) OA have been re-purposed as digital bit buffers, doing a fine job of that now IMO. What comes out is higher amplitude but still fully digital DSD bit stream. This then drives the filters, where the D/A conversion takes place. The output transformer then creates the center tapped balanced output stage,
IMO we could also explore using capacitors instead. And likely will.
Even better would likely be to provide a bi-polar PSU to the OA; then we would not need to block any DC on the output and no transformer or caps would be needed at all⌠Sounds really enticing to me⌠But not so easy to implement here. HmmâŚ
Requested specs below. Good enough, IMO. Like that group delay is ~ flat through the audio band? I figured youâd say âyesâ Iâm not sure how audible it is though ultimately, but I always strive for it. This is for the next iteration filter BTW (in progress).
Edit- I guess you meant the OA buffers in your original statement.
Turbo
1 Like