Develop ultra capacitor power supply and LiFePO4 battery power supply

Ready to bring the #1 test system into the house now!

@aguaazul

Yes, it is possible to include this new continuous DC 5V feature to MKII by upgrading the firmware.

Regards,
Ian

Ian,
I have 2 full IanCanada Prototype DIY Experimental systems that I have available to run.

I'd like to bring #1 in to the house to use with a Balanced Switch I have to choose between my (1) HiFiBerry, (2) DacMagic Pro & your (3) IanCanada Prototype #1.

Any chance we can get our hands on the firmware to keep the new continuous DC 5V feature on the MKII?

Thanks,

Aguaazul
 
Ian,
I have 2 full IanCanada Prototype DIY Experimental systems that I have available to run.

I'd like to bring #1 in to the house to use with a Balanced Switch I have to choose between my (1) HiFiBerry, (2) DacMagic Pro & your (3) IanCanada Prototype #1.

Any chance we can get our hands on the firmware to keep the new continuous DC 5V feature on the MKII?

Thanks,

Aguaazul

MKII can have the new continuous 5V function by upgrading the software.

Ian
 
I dont believe the battery outputs are current limited, havent tried welding with them though.

The batteries will deliver 50A constant, 100A spikes on their own so the limit is likely the pcb track width.

Adequate as Rolls Royce used to say

Just spotted in the LifePO4mk II manual that 5A fuses are fitted, these protect the relay contacts which have a 8 amp max rating. Therefore maximum power draw for a single LifePO4 is 13.2V * 5A = 66 watts.

Is running an 8 amp fuse recommended?

Can bridging of LifePO4's be done in parallel to double the amperage?
 
My Mk1 board came with 5A fuses fitted but supplied with 2 spare 8A fuses.

The Mk11/111 boards will be higher due to the battery holders.

Why the desire for such high current are you building a mobile arc welder ?

Rather than power a 5v raspberry pi, I'm going to power an Intel NUC. Higher powered processors sound better. I'll use the 5v rail to power a PCI-express card.

More than enough power for a NUC, but I am also considering a higher 80-90watt processor. Not unusual for very high end music servers to be built around a 12v 10amp rail (obviously I'll be using a multiple of 3.3).

So can you bridge in series as well as parallel?
Anyone know the height?
 
Higher powered processors sound better.
Hi stretchneck
Interesting... Did you compare higher vs lower powered processors to come to that opinion?
Reason I ask... there are a bunch of us here that have a legacy of tuning digital sound systems based on the teachings of a guy named CICS with his CPLAY Win98 based playback software. Back in the day, we were all striving for minimal number of tasks running to enable playback and tended to underclock the cpu in the belief that less processor power resulted in better sound. I'm wondering if we were misguided in our efforts. The use of Single Board Computers like the Rpi with minimalist OS kind of follows that mantra. So curious to learn your observations. Also aware there are many ways to achieve audio nirvana.
 
The difference between ultra capacitor upgrading solutions

I developed a full series of ultra capacitor upgrading solutions. Here I will explain the differences in between: UcConditioner, ConditionerPi, UcMateConditioner and UcHybrid.
The main difference could be that they are suitable for different applications.

1. If you have my LifePO4 power supply, no doubt UcMateConditioner and UcHybrid are the first solution.
UnMateConditioner can upgrade the 5V output into ultra capacitor voltage rail.
UcHybrid will upgrade each LifePO4 battery cell into ultra capacitor hybrid cell. If more than one battery cells is needed for voltage rail higher than 3.3V, multiple UcHybrid boards have to be installed.
UcAdapter PCB is designed to mount UcMateConditioner and UcHybrid directly to the LifePO4 power supply, either MKIII or MKII.

They are designed together as a total solution dedicated for the LifePO4 power supply.

2. UcConditioner would be a universal ultra capacitor conditioner. If you have 5V or 3.3V power supply other than the LifePO4 power, UcConditioner 5V/3.3V would be your best choice.
UcConditioner 5V: will upgrade any 5V power supply (1.5A or higher) into ultra capacitor power supply.
UcConditioner 3.3V: will upgrade any 3.3V power supply (1.5A or higher) into ultra capacitor power supply.

Only 5V and 3.3V version are developed so far. No control signal is needed.

3. ConditionerPi is another story. It improves quality of the main 5V power rail directly to the GPIO of RaspberryPi at local. It uses low profile ultra capacitors so that can easily be integrated with RPi stack. It also has built-in on/off control switch.

The ultra capacitors on ConditionerPi are smaller then the other conditioner because the limitation of space. But even with other conditioner installed, it can still make additional improvement at locall. ConditionerPi will suitable for any RPi audio application. Especially it improves the sound quality of light weight RPi audio application with USB or other simple power supply right away.

Please let me know if there is any confusing.




UcConditioner_ConditionerPi_UcMate_UcHybrid1
by Ian, on Flickr



Ian
 
Last edited:
Member
Joined 2010
Paid Member
Ian, I am impressed about the quality of all your projects. Congratulation!

I never approach the batteries PSU before but I read carefully this long thread and I decide to follow this path to rebuild the power supply of my DAC.

There will soon be a GB for the MKIII and the UcConditioners you just mention?

Thanks and Regards,
Enrico
 
Hi stretchneck
Interesting... Did you compare higher vs lower powered processors to come to that opinion?
Reason I ask... there are a bunch of us here that have a legacy of tuning digital sound systems based on the teachings of a guy named CICS with his CPLAY Win98 based playback software. Back in the day, we were all striving for minimal number of tasks running to enable playback and tended to underclock the cpu in the belief that less processor power resulted in better sound. I'm wondering if we were misguided in our efforts. The use of Single Board Computers like the Rpi with minimalist OS kind of follows that mantra. So curious to learn your observations. Also aware there are many ways to achieve audio nirvana.

I have been studying this for a while now, and higher powered CPU's make a significant difference (but you need a great power supply otherwise the sound can become harsh). There is a huge thread on Audiophile style.... I can't possibly go through everything here but their member AustinPop posted the following recently:

A novel way to massively improve the SQ of computer audio streaming - Page 649 - Music Servers - Audiophile Style

I think this is a good summary of the direction of travel, you can see that higher powered computers are ranked to sound better. My own experiences are the same (e.g. Allo USB Signature didn't sound anywhere near as rich and powerful in sound as my NUC, the USB Sig sounds comparatively flat and lifeless in comparision... both where powered by good quality linear power supplies).

But all of the same rules apply, low latency, low OS overhead, storage type, networks clocks, reclocked USB. Operating system makes a big difference as well. If you have a gaming PC (something with a high powered CPU) then try booting it with a USB stick which has a copy of Euphony or Gentoo Player on it... you might be amazing by how good it sounds even though the noise floor and lots of other things might not be optimised.

I'll post more on my build once I know the answer to putting LifePO4 in series and also the height of the product... anyone able to let me know?

Another good thread, Nenon is at the cutting edge: Building a DIY Music Server - Music Servers - Audiophile Style

You would think that a 40 core dual processor setup would be absolutely ridiculous... but it apparently sounds amazing. Powering this beast is a real issue, so I'll only ever go down the single processor route. Some people have given up with DIY and reached deep into their pockets and have spent $20,000 on the SGM Extreme. Apparently nothing comes close in sound quality, and I can well believe it.

Having said all of this, the theory behind LifePO4 is very sound - very significant uptick in sound quality for a Raspberry Pi I'm sure and it's massively cost effective. I am guessing that LifePO4 and a USBSig, with my favourite software Gentoo Player would easly sound better than many expensive commercial products and would completely annihilate the sound you'd get playing iTunes from a standard PC or Macbook. I'm really keen to try it with my NUC.

State of the Art has moved on a great deal since these earlier experiments / practices you quote.
 
Last edited:
I developed a full series of ultra capacitor upgrading solutions. Here I will explain the differences in between: UcConditioner, ConditionerPi, UcMateConditioner and UcHybrid.
The main difference could be that they are suitable for different applications.

1. If you have my LifePO4 power supply, no doubt UcMateConditioner and UcHybrid are the first solution.
UnMateConditioner can upgrade the 5V output into ultra capacitor voltage rail.
UcHybrid will upgrade each LifePO4 battery cell into ultra capacitor hybrid cell. If more than one battery cells is needed for voltage rail higher than 3.3V, multiple UcHybrid boards have to be installed.
UcAdapter PCB is designed to mount UcMateConditioner and UcHybrid directly to the LifePO4 power supply, either MKIII or MKII.

They are designed together as a total solution dedicated for the LifePO4 power supply.

2. UcConditioner would be a universal ultra capacitor conditioner. If you have 5V or 3.3V power supply other than the LifePO4 power, UcConditioner 5V/3.3V would be your best choice.
UcConditioner 5V: will upgrade any 5V power supply (1.5A or higher) into ultra capacitor power supply.
UcConditioner 3.3V: will upgrade any 3.3V power supply (1.5A or higher) into ultra capacitor power supply.

Only 5V and 3.3V version are developed so far. No control signal is needed.

3. ConditionerPi is another story. It improves quality of the main 5V power rail directly to the GPIO of RaspberryPi at local. It uses low profile ultra capacitors so that can easily be integrated with RPi stack. It also has built-in on/off control switch.

The ultra capacitors on ConditionerPi are smaller then the other conditioner because the limitation of space. But even with other conditioner installed, it can still make additional improvement at locall. ConditionerPi will suitable for any RPi audio application. Especially it improves the sound quality of light weight RPi audio application with USB or other simple power supply right away.

Please let me know if there is any confusing.




UcConditioner_ConditionerPi_UcMate_UcHybrid1
by Ian, on Flickr



Ian

So for 13.3 volts where 8 batteries slots are occupied to supply J3 and J4, 8 populated UcHybrid modules will need to be added to the UcAdapter PCB... one for each battery?

Can J3 and J4 be run in parallel to give 13.3v at 10-16amps (depending on 5-8amp fuse being used)?
 
I have been studying this for a while now, and higher powered CPU's make a significant difference (but you need a great power supply otherwise the sound can become harsh). There is a huge thread on Audiophile style.... I can't possibly go through everything here but their member AustinPop posted the following recently:

A novel way to massively improve the SQ of computer audio streaming - Page 649 - Music Servers - Audiophile Style

I think this is a good summary of the direction of travel, you can see that higher powered computers are ranked to sound better. My own experiences are the same (e.g. Allo USB Signature didn't sound anywhere near as rich and powerful in sound as my NUC, the USB Sig sounds comparatively flat and lifeless in comparision... both where powered by good quality linear power supplies).

But all of the same rules apply, low latency, low OS overhead, storage type, networks clocks, reclocked USB. Operating system makes a big difference as well. If you have a gaming PC (something with a high powered CPU) then try booting it with a USB stick which has a copy of Euphony or Gentoo Player on it... you might be amazing by how good it sounds even though the noise floor and lots of other things might not be optimised.

I'll post more on my build once I know the answer to putting LifePO4 in series and also the height of the product... anyone able to let me know?

Another good thread, Nenon is at the cutting edge: Building a DIY Music Server - Music Servers - Audiophile Style

You would think that a 40 core dual processor setup would be absolutely ridiculous... but it apparently sounds amazing. Powering this beast is a real issue, so I'll only ever go down the single processor route. Some people have given up with DIY and reached deep into their pockets and have spent $20,000 on the SGM Extreme. Apparently nothing comes close in sound quality, and I can well believe it.

Having said all of this, the theory behind LifePO4 is very sound - very significant uptick in sound quality for a Raspberry Pi I'm sure and it's massively cost effective. I am guessing that LifePO4 and a USBSig, with my favourite software Gentoo Player would easly sound better than many expensive commercial products and would completely annihilate the sound you'd get playing iTunes from a standard PC or Macbook. I'm really keen to try it with my NUC.

State of the Art has moved on a great deal since these earlier experiments / practices you quote.

Wow this goes opposite to what I have learnt. A much more powerful PC should not be needed for just music where less is more.

You need less noise here of course and not more noise with all those CPUs and i cant think how that makes the sound better to be honest.

How can music need all that processig power....well we all know it does not.

The PSU is king here plus all the ususal isolations etc. I know its all the case about trusting what you here but.....

Once Ian releases the new PSU and super cap solutions i will be ordering :D
 
Try the experiment I mentioned - powerful processors tend to have lower latency due to increased cache. Noise floor and ripple are only some of the factors affecting sound.

Anyhow, these findings are proven by many other now. I still don’t understand why a 40 core processor sounds better... but it does.

I’m looking forward to LifePO mkIII as well, just waiting for my answers to come back.