Hello,
I think that Doede's DDDAC has been fine tuned for quiet some time now.
I remember before buying it there already were some modifications being done by owners present here on this site and Doede took some serious time to '' implement '' these into the smaller board that is mounted on top of the mainboard. These personal mods were mostly done in a way i wouldnt do it so i decided to wait until Doede himself did them like they should be done!
The improvements that can be made to DDDAC are 95% of the time located outside the circuit boards. It will usually be the power supply, output caps or the use of an output transformer.
If you have a one board DDDAC adding extra boards will improve things. Most people will go up to 4 because going further will require a lot from your power supply. You could ask Doede himself if the best investment would be going up in boards or tackling the power supply.
Greetings. Eduard
@BMW850
P.S Using a small transformer to charge ultracaps and starting up a raspberry will probably not work
I think that Doede's DDDAC has been fine tuned for quiet some time now.
I remember before buying it there already were some modifications being done by owners present here on this site and Doede took some serious time to '' implement '' these into the smaller board that is mounted on top of the mainboard. These personal mods were mostly done in a way i wouldnt do it so i decided to wait until Doede himself did them like they should be done!
The improvements that can be made to DDDAC are 95% of the time located outside the circuit boards. It will usually be the power supply, output caps or the use of an output transformer.
If you have a one board DDDAC adding extra boards will improve things. Most people will go up to 4 because going further will require a lot from your power supply. You could ask Doede himself if the best investment would be going up in boards or tackling the power supply.
Greetings. Eduard
@BMW850
P.S Using a small transformer to charge ultracaps and starting up a raspberry will probably not work
Hi Kenneth,
Indeed a comparison that I made because off the compact build from the Ian DAC. But I honestly think that the DDDAC is the better one. At least to my ears.
It sounds more analogue, more detail and more 'peace'.
Also I'm curious to know how other diy'ers experience this.
My DDDAC is connected by I2S to the FifiPi Q7.
Regarrds, Jan
Hello,
AND it looks like Doede has done more research and won't release a new board because he found a tiny improvement. I remember people kind of begging him for a release date of the new boards but he and Guido Tent( the designer of newly added shunt supplies) took some serious how to make things work properly.
I also have the idea that back when i bought the DDDAC there was much more communication going on between the owners of Doede's circuit and Doede himself. It could very well be there is not much more room for improvement unless you are willing to spend double the amount of money so the path of improvement is kind of closed.
Doede, being retired, just enjoys his music and his wine? Just like we should do.
greetings, eduard
AND it looks like Doede has done more research and won't release a new board because he found a tiny improvement. I remember people kind of begging him for a release date of the new boards but he and Guido Tent( the designer of newly added shunt supplies) took some serious how to make things work properly.
I also have the idea that back when i bought the DDDAC there was much more communication going on between the owners of Doede's circuit and Doede himself. It could very well be there is not much more room for improvement unless you are willing to spend double the amount of money so the path of improvement is kind of closed.
Doede, being retired, just enjoys his music and his wine? Just like we should do.
greetings, eduard
Hi Jan,
Thank you for the replay, this make me more calm about keeping DDDAC.
Can i ask which transformers you use on the output.
Best regards Kenneth
Hello,
Doede also has knowledge about analog sound reproduction so i assume he will try to get his digital gear sounds as good as a good analog set up.
Also he seems to be less biased He wasn't so sure about the ultracaps being the post possible power supply for his DDDAC.
He can have a very pleasant life without us buying his circuits so that will be one more reason to accept his ideas.
Greetings, Eduard
Doede also has knowledge about analog sound reproduction so i assume he will try to get his digital gear sounds as good as a good analog set up.
Also he seems to be less biased He wasn't so sure about the ultracaps being the post possible power supply for his DDDAC.
He can have a very pleasant life without us buying his circuits so that will be one more reason to accept his ideas.
Greetings, Eduard
I was tying up some loose ends on the Power thread to make sure my Ian Infrastructure was complete before the Group Buy.
I forgot how much I wanted for my LiFePO4 MKIII Flagship to be in parallel, but since this is my first serious Series/Parallel project I wanted limited risk so I did enough due diligence to comfortably implement.
So I think that since there is UcPure chatter on tips of how to parallel Ultra Capacitors, I'm going to try to create a diagram similar to my LiFePO4 MKIII project. It will be nice to archive in the Power Thread, so anyone that wants to look it up years from now will have a guide on how to parallel UcPure. It's not a game and there's real world risk, so having experienced members fine tune on how to parallel will be valuable. For my project, it was a sample size of one use case wise. So since there are multiple sample sizes and growing for the Ultra Caps use case, it will be good to document. I'll then archive it in the Power thread when all is said and done.
Beautiful UcPure/Eaton Parallel rigs BTW.
Reminder for anyone with a Compute Module 4 AND UcPure to try out a DFRobot Tray:
It's minimal so it might not be as power hungry. If I had this and a Compute Module 4, I would send it over to Ian to test with UcPure. DFRobot products are available on Mouser too to save on shipping costs.
https://www.dfrobot.com/product-2196.html
There will be kinks to work out so appreciation on tips after I post the UcPure diagram later this week.
I forgot how much I wanted for my LiFePO4 MKIII Flagship to be in parallel, but since this is my first serious Series/Parallel project I wanted limited risk so I did enough due diligence to comfortably implement.
So I think that since there is UcPure chatter on tips of how to parallel Ultra Capacitors, I'm going to try to create a diagram similar to my LiFePO4 MKIII project. It will be nice to archive in the Power Thread, so anyone that wants to look it up years from now will have a guide on how to parallel UcPure. It's not a game and there's real world risk, so having experienced members fine tune on how to parallel will be valuable. For my project, it was a sample size of one use case wise. So since there are multiple sample sizes and growing for the Ultra Caps use case, it will be good to document. I'll then archive it in the Power thread when all is said and done.
Beautiful UcPure/Eaton Parallel rigs BTW.
Reminder for anyone with a Compute Module 4 AND UcPure to try out a DFRobot Tray:
It's minimal so it might not be as power hungry. If I had this and a Compute Module 4, I would send it over to Ian to test with UcPure. DFRobot products are available on Mouser too to save on shipping costs.
https://www.dfrobot.com/product-2196.html
There will be kinks to work out so appreciation on tips after I post the UcPure diagram later this week.
Last edited:
As for my thought process if I was going to procure 2.7V versus 3V both @3000F, I would likely stick with 3V with the power of hindsight.
The 3V is already a sunk cost, so there may be some bias. But I like to standardise for future projects, so I want it all to be 3V and not vary. Plus they are Eatons, they power Heathrow International Airport and they last for 20 years so I'm okay with the extra costs. Plus 3V is battle-tested by Ian and Ian states 3V provides the best SQ.
But you can make a case for 2.7V too:
I initially purchased 3V because you can run 3.3V without Balancers, but with the launch of UcBalancer, UcBalancer is mandatory IMO just for the peace of mind. So since UcBalancer basically only runs in 2.7V, you can easily make a case for just 2.7V @ 3000F as the extra headroom of 3V will likely never be needed.
But for I, it's a Ian standardisation, my own personal standardisation, Ian raves about the SQ of the 3V and the Industrial HQ of Eatons allows me to justify the extra costs.
But if I was building something like 15.5V + 15.5V + 15.5V in parallel, I can lean towards the 2.7V as costs will add up.
There's an inflection point somewhere where I would be swayed to not standardise.
The 3V is already a sunk cost, so there may be some bias. But I like to standardise for future projects, so I want it all to be 3V and not vary. Plus they are Eatons, they power Heathrow International Airport and they last for 20 years so I'm okay with the extra costs. Plus 3V is battle-tested by Ian and Ian states 3V provides the best SQ.
But you can make a case for 2.7V too:
I initially purchased 3V because you can run 3.3V without Balancers, but with the launch of UcBalancer, UcBalancer is mandatory IMO just for the peace of mind. So since UcBalancer basically only runs in 2.7V, you can easily make a case for just 2.7V @ 3000F as the extra headroom of 3V will likely never be needed.
But for I, it's a Ian standardisation, my own personal standardisation, Ian raves about the SQ of the 3V and the Industrial HQ of Eatons allows me to justify the extra costs.
But if I was building something like 15.5V + 15.5V + 15.5V in parallel, I can lean towards the 2.7V as costs will add up.
There's an inflection point somewhere where I would be swayed to not standardise.
Last edited:
@iancanada A kind request/suggestion. Any chance you would consider selling your FifoPiQ7 with SC-Pure clocks for those that are interested in purchasing that combo? It helps 1) To save some $$ for folks who want the SC-Pure. Today if I want the SC-Pure clocks with FifopiQ7 the only option is to buy the FifoPiQ7 and throw away the ACU clocks that come with it, right? 2) Less landfill waste
60 ours on the SC clocks now, and the sharp upper midrange is almost gone. I like the way this is going. It is already a nice step up compared to the Accu 338 i have been using quite a long time. Everything from the bass and all the way up is clearer with more details and placement of every instruments. The soundstage Is wider and deeper than before.
The SC clocks is powered 24/7 by UcPure mkIII with the big Eaton supercaps 2.7v 3000F. I have a separate psu powered on only for the clocks when the rest of my system is down. I don’t like to have my hifi powered all the time, and always turns it off before going to sleep.
For you guys waiting for the clocks I will say that you have a nice upgrade coming your way.
The SC clocks is powered 24/7 by UcPure mkIII with the big Eaton supercaps 2.7v 3000F. I have a separate psu powered on only for the clocks when the rest of my system is down. I don’t like to have my hifi powered all the time, and always turns it off before going to sleep.
For you guys waiting for the clocks I will say that you have a nice upgrade coming your way.
Hello,
After the clocks it will time to invest let us say 1000 Euro/$ in improving the listening space.
Of course drinking coffee from beans you just grinded yourself will amplify your listening experience.
Getting rid of your bookshelf speakers can be done next year.
Greetings, eduard
After the clocks it will time to invest let us say 1000 Euro/$ in improving the listening space.
Of course drinking coffee from beans you just grinded yourself will amplify your listening experience.
Getting rid of your bookshelf speakers can be done next year.
Greetings, eduard
Hm, fascinating stuff @smooth dancer. At mine 100 hour mark for SC-Pure, my Accus 318s were clearly better in every conceivable way. So much that I refused to listen and made another round of comparison only now at around 170 hours.
On the technical side like soundstage, layering, separation or details,, they're now pretty much on par, really can't tell them apart. Except for tonality, where I clearly hear added midrange (whole spectrum) about 2-4 dBs and lowered bass, but not much, just on the verge of perception (1-2 dBs). Just based on my tonality preference, I'd still pick Accus now, but it doesn't mean that SC-Pure is worse at anything. And based on my experience, some elements can improve even after 200 hours, therefore no final judgment on my side yet.
So if you prefer SC-Pure so early in their break-in process, that could implicate only two things (or their combination):
1. Your Accus were given sub-optimal conditions to operate and/or perhaps are fakes (btw I never found any official document/link/prospect that Accusilicon was ever producing "338" clock range, you did?)
2. I'm giving suboptimal conditions to SC-Pure to operate. So thinking hard, where could they be hiding?
Regarding point 1 and I'm curious about all of you reading here: Have you shortened the pins of Accus clocks, so that they rest fully on the boards socket? At least for FifoPiMa, I had to do it, otherwise the clocks would be hanging in the air above the socket, which would undoubtedly affect their performance to a rather large degree.
Regarding 2: Even though I put utmost care to minimize the noise coming in and out of the Rpi (3B , v1.2), which is fed by an excellent and further modded LPSU, GPIO contacts and the little fuse for 3.3V cured by fantastic Q45T, double shielded, isolated with standalone Ian's isolator, etc etc - is it is probable that either the PurePi platform (even with UcConditioner) or the FifoPiMa itself could be the limiting factor here?
Either way I'm planning to find out for myself eventually, but it's a mystery that each of the dozens and dozens of tweaks to my streamer or digital network generated perceivable gains in SQ and enjoyment factor, just not this clock upgrade (yet). I could maybe try to remove my "noise eating" tweaks from my DAC to see if some delta appears then.
On the technical side like soundstage, layering, separation or details,, they're now pretty much on par, really can't tell them apart. Except for tonality, where I clearly hear added midrange (whole spectrum) about 2-4 dBs and lowered bass, but not much, just on the verge of perception (1-2 dBs). Just based on my tonality preference, I'd still pick Accus now, but it doesn't mean that SC-Pure is worse at anything. And based on my experience, some elements can improve even after 200 hours, therefore no final judgment on my side yet.
So if you prefer SC-Pure so early in their break-in process, that could implicate only two things (or their combination):
1. Your Accus were given sub-optimal conditions to operate and/or perhaps are fakes (btw I never found any official document/link/prospect that Accusilicon was ever producing "338" clock range, you did?)
2. I'm giving suboptimal conditions to SC-Pure to operate. So thinking hard, where could they be hiding?
Regarding point 1 and I'm curious about all of you reading here: Have you shortened the pins of Accus clocks, so that they rest fully on the boards socket? At least for FifoPiMa, I had to do it, otherwise the clocks would be hanging in the air above the socket, which would undoubtedly affect their performance to a rather large degree.
Regarding 2: Even though I put utmost care to minimize the noise coming in and out of the Rpi (3B , v1.2), which is fed by an excellent and further modded LPSU, GPIO contacts and the little fuse for 3.3V cured by fantastic Q45T, double shielded, isolated with standalone Ian's isolator, etc etc - is it is probable that either the PurePi platform (even with UcConditioner) or the FifoPiMa itself could be the limiting factor here?
Either way I'm planning to find out for myself eventually, but it's a mystery that each of the dozens and dozens of tweaks to my streamer or digital network generated perceivable gains in SQ and enjoyment factor, just not this clock upgrade (yet). I could maybe try to remove my "noise eating" tweaks from my DAC to see if some delta appears then.