The noise handled by the ground box is very small, but it can actually be measured with living organisms. However, this noise cannot be measured with general instruments. It's like you feel dizzy every day, and when you are checked by precision instruments in the hospital, most of the time the doctor says you are not sick, and then you have to go home very depressed, and then when the condition is more serious, it is finally discovered that it is XXX disease.
By the same logic, if the system noise is serious, it is easy to hear it. If it is not so serious, it is impossible to hear it directly, unless the grounding box removes this small amount of noise to make the sound better. Grounding noise will only be discovered on the basis of comparison, and a very subtle sound heard before is actually noise.
The above-mentioned living organisms for detecting ground noise is probably to input the signal current to the living frog and measure his brain wave changes. Since this kind of experiment requires grounding, a grounding rod inserted into the ground, so the comparison can be made with a grounded box and found to have variations similar to grounded rods.
You have a wonderful wife who would accept you adding something like this to the ceiling. 😂
If the system noise is serious, it is easy to hear it. If it is not so serious, it is impossible to hear it directly, unless the grounding box or something else removes this small amount of noise to make the sound better. Grounding noise will only be discovered on the basis of comparison, and a very subtle sound heard before is actually noise.Or, whether or not there is an audible difference.
Hi,
We both presented decent options for others to consider and let's just leave it at that.
I just like to use official brands where I can. The Hirose socket on the PCB matches my official Hirose connector perfectly. I tried non-official connectors and it's torture. That's why I shared my experiences. If someone new whom is assembling the PCBs and connectors run into trouble, there is an alternative. One can easily apply too much pressure with pliers with getting the non-official connectors to stick. I wouldn't want to go through that again or for anyone else, that's why I recommend just throwing a official Hirose connector in your next Mouser / Digikey order to try out. It's low risk, high reward.
So let's just stop here, we gave people options so our job is done.
I'm trying to cut Cooper Bussbar in (2) Cooper Washers as sets of M12 Cooper Washers are too expensive at this time. With the UcBalancer available to order in GB2023, Cooper Bussbars are not needed anymore so it's a no loss proposition. I'm glad we can agree on Nylon M12 Nuts at least. So we have similar end goals and safety concerns, just different ways to get there. Good Weekend.
Last edited:
Hi all,
I also want to build this streamer.
And I want the best modules available also the power supply.
Please give me the good direction witch modules to buy.
The streamer and dac I never switch off also this streamer must be on 24/7
My dac is the Denafrips Terminator Plus (with all the inputs available and clock output)
Now I use the usbridge signature with the shanti power supply as my current streamer and my server is the nuc8i7beh passive cooling with roon rock.
Regards,
Rudy
I also want to build this streamer.
And I want the best modules available also the power supply.
Please give me the good direction witch modules to buy.
The streamer and dac I never switch off also this streamer must be on 24/7
My dac is the Denafrips Terminator Plus (with all the inputs available and clock output)
Now I use the usbridge signature with the shanti power supply as my current streamer and my server is the nuc8i7beh passive cooling with roon rock.
Regards,
Rudy
Attachments
@BMW850
I have a similar system than yours
Here are the streamer/transport options.
I recommend the option #C flagship
https://www.diyaudio.com/community/...fight-the-jitter.192465/page-414#post-7263122
If it is possible I suggest you to change the LifePO4 Mini 3.3V (#42A) and UcConditioner 3.3V (#26A) to
#41C UcPureMkIII for a great sound quality upgrade.
HdmiPiProDDCListeningRoom by Ian, on Flickr
I have a similar system than yours
Here are the streamer/transport options.
I recommend the option #C flagship
https://www.diyaudio.com/community/...fight-the-jitter.192465/page-414#post-7263122
If it is possible I suggest you to change the LifePO4 Mini 3.3V (#42A) and UcConditioner 3.3V (#26A) to
#41C UcPureMkIII for a great sound quality upgrade.
HdmiPiProDDCListeningRoom by Ian, on Flickr
Hi @iancanada , hello @Gabster 2000, hello world!
I've got a very similar question like BMW850. Here is what Ian's link shows, which can be also found in some of the documents of his modules:
C. Flagship RPi I2S/DSD network streamer transport
1. HdmiPi Pro (#17B) or TransportPi AES
2. FifoPi Q7 (#1D)
3. StationPi SMT (#35C) OR Pro
4. LinearPi 5V(#32A)
5. UcConditioner 5V (#25A)
6. LifePO4 Mini 3.3V (#42A)
7. UcConditioner 3.3V (#26A)
8. MonitorPi (#49A) or Pro
For 6 and 7 comes the new UcPureMkIII.
The manual for the StationPi Pro states, that there are 2 inputs for 5V, one for the Raspberry Pi (dirty side), and another one. For which purposes are the (other, cleaner) 5V used or meant? (If e.g. clocks are using 5V, then a UcPureMkIII would make sense.)
Which power supply would then be ideal to power the Raspberry Pi? The listet LinearPi 5V?
Is it capable of powering a Raspi 4, because it's max I out is 2,5A?
If the LinearPi 5V can power the Raspberry Pi, this should be really top notch, or does an additional filtering has other benefits which could improve the streamer's overall performance (RFI, noise, xyz)?
Last question for now:
I plan using the StationPi Pro, Hdmi Pro, TransportPi AES, MonitorPi Pro, Sc-Pure Clocks, and maybe the touchscreen controller for the Station Pi Pro: What will be the average 3,3V power consumption?
Thank you very much!
Best regards,
Matthias
I've got a very similar question like BMW850. Here is what Ian's link shows, which can be also found in some of the documents of his modules:
C. Flagship RPi I2S/DSD network streamer transport
1. HdmiPi Pro (#17B) or TransportPi AES
2. FifoPi Q7 (#1D)
3. StationPi SMT (#35C) OR Pro
4. LinearPi 5V(#32A)
5. UcConditioner 5V (#25A)
7. UcConditioner 3.3V (#26A)
8. MonitorPi (#49A) or Pro
For 6 and 7 comes the new UcPureMkIII.
The manual for the StationPi Pro states, that there are 2 inputs for 5V, one for the Raspberry Pi (dirty side), and another one. For which purposes are the (other, cleaner) 5V used or meant? (If e.g. clocks are using 5V, then a UcPureMkIII would make sense.)
Which power supply would then be ideal to power the Raspberry Pi? The listet LinearPi 5V?
Is it capable of powering a Raspi 4, because it's max I out is 2,5A?
If the LinearPi 5V can power the Raspberry Pi, this should be really top notch, or does an additional filtering has other benefits which could improve the streamer's overall performance (RFI, noise, xyz)?
Last question for now:
I plan using the StationPi Pro, Hdmi Pro, TransportPi AES, MonitorPi Pro, Sc-Pure Clocks, and maybe the touchscreen controller for the Station Pi Pro: What will be the average 3,3V power consumption?
Thank you very much!
Best regards,
Matthias
So looking at the Uc power supplies, UcPure is the best. Uc Conditioner is next best. Then there are three Uc for the Pi: UcPi which has 2 smaller Ucs, PurePi with 1 Uc and two LiPO4 batteries, and ConditionerPi which uses 2 to 8 of the smaller Ucs. So for the 3.3V supply to the clean side of the FIFO Pi use a UcPure. For the 5V supply neither the UcPure or the Uc Conditioner manuals really recommend using for with the Pi, but we have three different options for the Pi. Why would you choose one of the Pi options over the others?
In case you also intend to add diffusers, i strongly recommend not to place them behind the loudspeakers. So far they can act at all, they will send you some frequencies directly and som indirectly back, because of their particular behaviour. Diffusors treat higher reflections as we knon, while the problem are the bass frequencies running out of speakers to their backside = the front of our sound wall. I recommend bass absorbers there (large, thick) and place the diffusers always at the opposite side of the loud speakers.
Hi Ian,
how do you feed primary power to your system? Do you use switching supply for the 3 LifePO4? Or AC toroid for the dual LinearPI ?
How is your setup here?
Regards
Branko
Please take my comments with a grain / pinch of salt. I haven't got around to the basic StationPi yet, but hoping to after assembling the UltraCap 5V this Summer. Plus, I'm really rusty at this hobby so I'm not caught up on the latest and my memory can be bad.
I think this issue is complicated. You would think you can just UcPure 5V to RPi, but the RPi I believe has protection mechanisms when the voltage is out of range. I think Gabby briefly touched on this on one of his YouTube videos. That's when I gave up on this idea is after the video. So at this time, I think one has to rule out UcPure 5V for RPi.
Since I already have a LiFePO4 6.6V -> LPS -> UcConditioner 5V, I believe Ian mentioned privately when I didn't have a DIYAUDIO account that I can just SHARE 5V to both the RPi side and the HAT side of the StationPi with this stack.
So that is what I will try late Summer. As for 3.3V on the HAT side, I don't think it's worth the effort since there's no core Audio components involved. I believe if you feed the HAT side 5V, it will downconvert where needed internally for 3.3V RPi purposes. You will need to feed the clean side of the Q3/Q7 anyways with UcPure 3.3V, so I don't see a need to feed 3.3V on the HAT side. There's just no upside to do so. The clean side 3.3V will then feed the SinePi if that's in your chain, so no need to independently feed the SinePi.
I think I'm going to also SHARE UcPure 3.3V to clean side and TransportPi II. Hopefully, we can compile what can be SHARED to save on cost and time.
"SHARED confirmed"???:
"LiFePO4 6.6V -> LPS -> UcConditioner 5V":
"LiFePO4 6.6V -> LPS -> UcConditioner 5V"(1) -> StationPi RPi side
"LiFePO4 6.6V -> LPS -> UcConditioner 5V"(2) -> StationPi HAT side
"UcPure 3.3V":
"UcPure 3.3V"(1) -> Q3/Q7 Clean side
"UcPure 3.3V"(2) -> TransportPi II
I'm hoping by end of Summer I can confirm this is working. SHARED seems quite important as you can streamline Power Solutions and not have to get a UcPure for every indivdual component.
I think this issue is complicated. You would think you can just UcPure 5V to RPi, but the RPi I believe has protection mechanisms when the voltage is out of range. I think Gabby briefly touched on this on one of his YouTube videos. That's when I gave up on this idea is after the video. So at this time, I think one has to rule out UcPure 5V for RPi.
Since I already have a LiFePO4 6.6V -> LPS -> UcConditioner 5V, I believe Ian mentioned privately when I didn't have a DIYAUDIO account that I can just SHARE 5V to both the RPi side and the HAT side of the StationPi with this stack.
So that is what I will try late Summer. As for 3.3V on the HAT side, I don't think it's worth the effort since there's no core Audio components involved. I believe if you feed the HAT side 5V, it will downconvert where needed internally for 3.3V RPi purposes. You will need to feed the clean side of the Q3/Q7 anyways with UcPure 3.3V, so I don't see a need to feed 3.3V on the HAT side. There's just no upside to do so. The clean side 3.3V will then feed the SinePi if that's in your chain, so no need to independently feed the SinePi.
I think I'm going to also SHARE UcPure 3.3V to clean side and TransportPi II. Hopefully, we can compile what can be SHARED to save on cost and time.
"SHARED confirmed"???:
"LiFePO4 6.6V -> LPS -> UcConditioner 5V":
"LiFePO4 6.6V -> LPS -> UcConditioner 5V"(1) -> StationPi RPi side
"LiFePO4 6.6V -> LPS -> UcConditioner 5V"(2) -> StationPi HAT side
"UcPure 3.3V":
"UcPure 3.3V"(1) -> Q3/Q7 Clean side
"UcPure 3.3V"(2) -> TransportPi II
I'm hoping by end of Summer I can confirm this is working. SHARED seems quite important as you can streamline Power Solutions and not have to get a UcPure for every indivdual component.
Last edited:
It depends on what you want to archive / what is your goal.
Sharing is always a compromise performance wise. This goes all the chain down to the transformer, separates are always better.
Of course there are instances where a shared supply is completely adequate and fine, like one switching supply for a desktop or notebook. But even for solid state class A amps you'll find a difference in 3D presentation if the transformers for the left and right channel are from high quality compared to one with two pairs of windings.
Pros for shared power supply are:
Cheaper, less room needed in case, less work with installing and wiring, easier to keep distance to delicate areas like audio, clocks, etc.
Yes, I can see that POV from a Audiophile / Purist perspective to have seamless separation. But in the context of Ian gear, his power supplies are so clean and powerful with zero feedback so it's a bit of new frontier so one has to adjust their thought process accordingly.
For example, if you feed both StationPi 5V with "LiFePO4 6.6V -> LPS -> UcConditioner 5V" there's no feedback from the RPi side or the dirty side. In traditional power systems, the dirty side can contaminate the RPi side if you SHARE. In the context of Ian gear, I don't think this is an issue. Plus he would not recommend if he was not confident it would be a solid solution.
Because power is basically one-way, I don't think performance will be affected. There would be plenty of power to both sides of the StationPi.
I understand the hesitancy of this and hope others can weigh in, but we need to tackle this from a Passive Power Supply perspective not tradtional Active Power Supply. There's no control loop regulation back and forth between the dirty side back to the "LiFePO4 6.6V -> LPS -> UcConditioner 5V" stack back to the RPi side or RPi side to the "LiFePO4 6.6V -> LPS -> UcConditioner 5V" back to the dirty side. It's all one-way. This may not apply if one is running "LPS -> UcConditioner 5V". I don't know how that works as I have a Passive Power Supply underneath.
I believe you plan SC-PURE. The SC-PURE 3.3V will be powered from Q7 clean side via UcPure 3.3V, not the StationPi dirty side. The dirty side is just for mundane RPi / HAT tasks, not core Audio tasks like clocks.
I welcome any corrections and as a Purist I can see this POV. If I wasn't confident this may work out, I would plan a different architecture. But only time will tell. I believe it is just wasted resources if one doesn't consider SHARING. Do you really want to invest in a UcPure + 2 Eatons + Accessories just to power a TransportPi? Cost/Benefit...
Last edited:
Like i said: It depends on what you want to archive / what is your goal.
I will be waiting for @iancanada to chime in, as I would like to know which modules are using the (clean) 5V. If I know this I can decide what to use.
As a comment, as long you have anything than an ideal current sink, like a resistor, there are effects on the power supply. E.g. if the current draw is not linear, than the power source will react electrically to the changing sink.
So you have to estimate, how big this influence is in the power source, how big the sink's immunity is agains these fluctuations, and last, how you qualify these effects in the end.
So, if you listen to 50$ speakers, all this don't matter, but if you try to compete with a 5000$ or more expensive streamers, or to coin it differently, to find our where the limits of a DIY streamer are, than these thought are worthwhile.
I have a quite expensive DAC, and I would like to find out, if I like digital playback more (is more enjoyable/ brings more joy or positiv energy) than analog playback. For this the (for me very) expensive DAC, which I want to combine with the best DIY streamer possible, with a "audiophile" network switch, and so on.
So, Cost/Benefit depends on the point of view. But if you compare the needed budget for a crazy all-in-DIY-streamer to a 2k, 5k or even 10k streamer (there are even more expensive ones!), this is "cheap". Of cause, it's still a lot of money, I know that.
Last point:
If someone plans on a limited budget or case, this is something different. But as long as the "clean" 5V side is feed into sensitive modules in the streamer, I would rather choose 2 independent, cheaper power supplies than a better one and share it.
But then: I haven't tried both solutions, and I haven't seen any measurements of any sort. (Ah, Gabster showed in one video how he measured with a soundcard THD analyzer (ARTA or REW) the noise of a power supply. But to be able to measure the digital noise, I guess we have first to invest in some serious testing equipment.)
All the best
Matthias
I will be waiting for @iancanada to chime in, as I would like to know which modules are using the (clean) 5V. If I know this I can decide what to use.
As a comment, as long you have anything than an ideal current sink, like a resistor, there are effects on the power supply. E.g. if the current draw is not linear, than the power source will react electrically to the changing sink.
So you have to estimate, how big this influence is in the power source, how big the sink's immunity is agains these fluctuations, and last, how you qualify these effects in the end.
So, if you listen to 50$ speakers, all this don't matter, but if you try to compete with a 5000$ or more expensive streamers, or to coin it differently, to find our where the limits of a DIY streamer are, than these thought are worthwhile.
I have a quite expensive DAC, and I would like to find out, if I like digital playback more (is more enjoyable/ brings more joy or positiv energy) than analog playback. For this the (for me very) expensive DAC, which I want to combine with the best DIY streamer possible, with a "audiophile" network switch, and so on.
So, Cost/Benefit depends on the point of view. But if you compare the needed budget for a crazy all-in-DIY-streamer to a 2k, 5k or even 10k streamer (there are even more expensive ones!), this is "cheap". Of cause, it's still a lot of money, I know that.
Last point:
So, if I do everything possible for the reclocking etc, and then, I use a non-optimal power supply to the output to this signal (HdmiPi Pro/ or TransportPi AES), and there I again get bits of current/ voltage noise and EMI back in the digital signal, which is then transported to the DAC, which is converting the digital bits with this added noise into analog, I don't see the point.
If someone plans on a limited budget or case, this is something different. But as long as the "clean" 5V side is feed into sensitive modules in the streamer, I would rather choose 2 independent, cheaper power supplies than a better one and share it.
But then: I haven't tried both solutions, and I haven't seen any measurements of any sort. (Ah, Gabster showed in one video how he measured with a soundcard THD analyzer (ARTA or REW) the noise of a power supply. But to be able to measure the digital noise, I guess we have first to invest in some serious testing equipment.)
All the best
Matthias
Yes, that's admirable you want to build a pristine chain. I was determined to pursue that route when I first started before I realised how robust Ian's Passive Power Supplies with Solid Core Cabling are as a solution. Plus, add in that it's a continuous process so there may be slight changes / modifications over time. It's near impossible to get it completely right the first time. So, it might be StationPi today but another solution tomorrow. This makes me not put all my eggs in one basket to power the StationPi to maintain flexibility.
So I can just stay put powering the TransportPi II via the StationPi HAT side or use the extra juice from the UcPure 3.3V powering the Q7 clean side. I choose the latter as I can't justify giving the individual component (TransportPi II) it's only power supply as much as I love the TransportPi II. The TPII freaking handles 5.6448 MHz without a sweat, so I treasure it immensely but to give it it's own UcPure doesn't compute. I do though have the luxury of powering the external clock with a Passive Power Supply @ 16.5V, so the UcPure 3.3V that powers the Q7 clean side doesn't have to work as hard.
I'm still amazed with all the products Ian has designed.
It will be interesting how this plays out.
So I can just stay put powering the TransportPi II via the StationPi HAT side or use the extra juice from the UcPure 3.3V powering the Q7 clean side. I choose the latter as I can't justify giving the individual component (TransportPi II) it's only power supply as much as I love the TransportPi II. The TPII freaking handles 5.6448 MHz without a sweat, so I treasure it immensely but to give it it's own UcPure doesn't compute. I do though have the luxury of powering the external clock with a Passive Power Supply @ 16.5V, so the UcPure 3.3V that powers the Q7 clean side doesn't have to work as hard.
I'm still amazed with all the products Ian has designed.
It will be interesting how this plays out.
Last edited:
Does this mean that Ian admits that LifePO4 is not a good power supply? I have already mentioned my experience what any kind of battery will be appeared similar to the ripple current as its capacity decreases when Ian developed the LifePO4 power module.@BMW850
I have a similar system than yours
Here are the streamer/transport options.
I recommend the option #C flagship
https://www.diyaudio.com/community/...fight-the-jitter.192465/page-414#post-7263122
If it is possible I suggest you to change the LifePO4 Mini 3.3V (#42A) and UcConditioner 3.3V (#26A) to
#41C UcPureMkIII for a great sound quality upgrade.
by Ian, on Flickr
There is no need to continue to use batteries, because when the current passes through the supercapacitor, the quality of the current depends purely on the supercapacitor and has nothing to do with the front-end power supply. You will know it after you try it. A supercapacitor with a large capacity is like a large pool. After the dirty current is input, it will be purified.