Develop ultra capacitor power supply and LiFePO4 battery power supply

Sorry for the slow reply, been a busy week.

AlanElsdon, I still owe you a separate response. In the next few days. BUT VERY GLAD you got it all running!

redjr, did you get it to work ok? If not, when I last used my modified HiFiBerry DAC+Pro it was with a standard 49/44 clocks Allo.com Kali reclocker on an RPi running PiCorePlayer. I used one of the basic I2S DAC drivers. While I had done a number of power modifications to the unit, as I remember (and from looking at it) it worked ok with no mods to the clocking and/or clock power schemes. Same on needing no mods to clocks or clock power when I tried both Allo's original and upgraded 2nd version Boss DACs along with their Piano 2.1 on a Kali.

In most cases I preferred the Kali to using the RPi clocks whether in Master or Slave mode. Only Allo's 2nd version Boss sounded best in Master mode.

I'm not familiar with RopieeeXL, so can't advise on what driver/overlay is equivalent. You may have to also play with bit depth.

BTW, comparing the Kali against Ian's FiFiPiQ3, I preferred the Q3 with the same type of clocks over the Kali. While the Kali was groundbreaking when released, if Allo ever becomes active again an update would be appropriate... if for nothing else, likely there are newer versions of the I2S signal chips that will provide lower output jitter.

Greg in Mississippi

P.S. Descriptions of my mods to the HiFiBerry DAC+Pro are in this thread, starting with post #371 for some of the latest info and working towards the front of the thread for the rest...

https://www.diyaudio.com/community/threads/hifiberry-dac-pro-hw-mods-anybody.285811/page-19
 
@greg, Ian informed me that the hifiberry dac+ dsd board only works in master mode and therefore will not work with any of the FifoPi versions. It wasn't a show stopper. I just wanted to see if it worked with some of the new platforms I'm toying with. I'm just getting started with some of Ian's LifePO4 products. The PurePi sounds very nice indeed.
 
redjr,

Glad you're liking Ian's goodies and that they are working well for you.

I wrote and posted a long email saying that the DAC+ Pro DID work in Slave mode... then noticed you were talking about the DAC+ DSP.

Though I don't know much about that model, I'd agree with Ian on this one, that it needs to be in Master mode with their driver and won't work with a reclocker like Ian's FiFoPi.

So I deleted that post.

One thing you may be able to try may be a setting in the driver... when I first tried to run the DAC+ Pro in Slave mode, it was with their driver and a Slave setting checked. I never got that to work.

That MAY be an option for the DSP and MAY work.

But since I see it is not a critical thing for you, I suspect it just makes sense to go onto the next thing.

Best!

Greg in Mississippi
 
Hi,
just yesterday finished my first UCPure supply. As I could not get some 3000F Ultracaps new I bought some used Maxwell BCAP3000 from eBay.
Now I am not sure If all is right:
For the 3.3V UCPure version I took two caps with balancing board and connected everything to the PCB as stated in the mnánual. Then started to charge with a laptop charger (19v, 3.42A).
The red charging LED lit right up and after ca. 50 min. the orange "full" LED came on. When I set both Switch to on position the blue and green LEDs. cam eon too.

Now if I measure the Voltage at J9 (output) it is 3.3V as expected.

But at J7 (to UC pack) ist is also 3.3 V? shouldn't it be 5.4V if the two caps are "full" an put in series? As the two Caps are rated 2.7V/3000F each?

thx
Branko
 
redjr,

Glad you're liking Ian's goodies and that they are working well for you.

I wrote and posted a long email saying that the DAC+ Pro DID work in Slave mode... then noticed you were talking about the DAC+ DSP.

Though I don't know much about that model, I'd agree with Ian on this one, that it needs to be in Master mode with their driver and won't work with a reclocker like Ian's FiFoPi.

So I deleted that post.

One thing you may be able to try may be a setting in the driver... when I first tried to run the DAC+ Pro in Slave mode, it was with their driver and a Slave setting checked. I never got that to work.

That MAY be an option for the DSP and MAY work.

But since I see it is not a critical thing for you, I suspect it just makes sense to go onto the next thing.

Best!

Greg in Mississippi
Greg,

Thanks for your clarification post. In Ropieee, there are 11 HifiBerry overlays that look to be one for every DAC hat model they offer. I did try several of them, but none seemed to work. I wonder if the HifiBerry would work with the FifoPi MA since it works in MASTER mode? Or, can't you have two devices that operate in master mode? I'm not too fluent in the inner workings of the FifoPi and clocks and how they interact. 😉

redjr
 
But at J7 (to UC pack) ist is also 3.3 V? shouldn't it be 5.4V if the two caps are "full" an put in series? As the two Caps are rated 2.7V/3000F each?

Please be careful: ultracaps can be very risky devices! If you don't have a proper understanding of what capacitors/ultracapacitors are, how they work, what are their limits/specifications and the basic electrics/electronics theory about them, maybe it would be better to avoid playing with fire...
 
Last edited:
Please be careful: ultracaps can be very risky devices! If you don't have a proper understanding of what capacitors/ultracapacitors are, how they work, what are their limits/specifications and the basic electrics/electronics theory about them, maybe it would be better to avoid playing with fire...
I tried to read a lot about them. And I am careful with them. As far as for the UCPure it should be/is quite easy to assemble it.
But I do not have the tools to properly measure this Ultracaps.
Therefor I would like to compare the behavior of my UCPure with others to have an idea if they work properly.
 
I tried to read a lot about them. And I am careful with them. As far as for the UCPure it should be/is quite easy to assemble it.
But I do not have the tools to properly measure this Ultracaps.
Therefor I would like to compare the behavior of my UCPure with others to have an idea if they work properly.
Ok, i was referring to your question about voltage after charging being lower than capacitors' rated voltage:
But at J7 (to UC pack) ist is also 3.3 V? shouldn't it be 5.4V if the two caps are "full" an put in series? As the two Caps are rated 2.7V/3000F each?
If you understand well how caps (and ultracaps) in series work, you should also understand the difference between rated voltage and charge voltage, i.e. the difference between the max voltage the cap series can stand and the actual voltage you charge it with. Also, the differences between capacitance, charge and voltage, and the relation Q=C*V.
To be able to output 3,3V, UC pack have to be charged to 3,3V, as i guess that using an output regulator would ruin the exceptional UC output qualities.
Then if you charge a 2-cap series to 3,3V it can't store (and deliver) 5,4V voltage after charging, even if the cap series can stand 5,4V, as the entire series will act as a single, double-V-rated cap and will be charged with 3,3V IN ALL (i.e. about 1,65V each cap). This doesn't mean the caps series is not "fully" charged, only it is charged with 3,3V charges. To reach maximum charge (Q) you should charge it to 5,4V, then it would deliver 5,4V current - but obviously this wouldn't fit the 3,3V UCPure case. So probably the 19V input current you provide is regulated to 3,3V (or little more) by the UCPure charger section itself BEFORE reaching the UC pack. Directly providing 19V current to the 5,4V rated UC pack, without voltage control/regulation, would lead to a real disaster!
 
Ok, i was referring to your question about voltage after charging being lower than capacitors' rated voltage:

If you understand well how caps (and ultracaps) in series work, you should also understand the difference between rated voltage and charge voltage, i.e. the difference between the max voltage the cap series can stand and the actual voltage you charge it with. Also, the differences between capacitance, charge and voltage, and the relation Q=C*V.
To be able to output 3,3V, UC pack have to be charged to 3,3V, as i guess that using an output regulator would ruin the exceptional UC output qualities.
Then if you charge a 2-cap series to 3,3V it can't store (and deliver) 5,4V voltage after charging, even if the cap series can stand 5,4V, as the entire series will act as a single, double-V-rated cap and will be charged with 3,3V IN ALL (i.e. about 1,65V each cap). This doesn't mean the caps series is not "fully" charged, only it is charged with 3,3V charges. To reach maximum charge (Q) you should charge it to 5,4V, then it would deliver 5,4V current - but obviously this wouldn't fit the 3,3V UCPure case. So probably the 19V input current you provide is regulated to 3,3V (or little more) by the UCPure charger section itself BEFORE reaching the UC pack. Directly providing 19V current to the 5,4V rated UC pack, without voltage control/regulation, would lead to a real disaster!
Thank you for taking time to help.
This is exactly the way the UCPure should work: configured to 3.3V or 5.0V (with 2 caps) by jumpers/resistors it uses DC or AC for input 12-20V and has isolated regulation to charge the ultra caps.
In My understanding it should always give the 3.3V or 5.0 V. A blue LED indicates the "pure state" when the power comes from caps only. If the blue light goes of a new charging cycle starts and for short time the power is not "pure".

Here is my problem: When I connect the power supply with UCPure set to "Off" (as per manual) red LED come on (shows power in is working) after some time orange LED comes on (showing caps are "full"). Now I set the UCPure to ON and I get exact 3.3 V on output (blue LED is on; orange LED on.).
BUT! after some time voltage on output goes dove to 3.0V and blue LED stays on but orange "full" LED is off and comes not on again? Tried this without load and with load (power LED 3.5V 700 mA). With load the blue LED goes out for some time and comes up again but no orange LED and the Voltage fluctuates between 3.3 an 3.0V.

Now I don't know if it is a problem from caps or from the board? Don't know how to troubleshoot further. And would not like to buy an other pair of expensive caps with out need.

Branko
 
Last edited:
UcPure in 15V 6*3000F configuration

Now it turns even more crazy.
This is a UcPure in 15V configuration with 6*3000F ultracapacitors.

When powered on, only the fully charged ultracapacitor pack will be applied to the load. It will be a pure UC power supply and 100% isolated from any rest of the circuit.

No doubt this is the best 15V power supply in the real world. And it will be the best power supply for Andrea’s TWTMC clocks.

Why?

Actually ultra capacitor has two different working modes.

UcPure works in the first mode which I call it the class A mode. In this mode, only ultracapacitors apply to the load. No other active power supply is involved. The current goes only one direction out of the ultracapacitors into the load. Class A mode will ensure the best low ESR and low noise performance.

Another mode I call it the class B mode. In this mode, ultracapacitors work as class B shunt elements in parallel with another power supply or charger. The current goes both directions into and out of the ultracapacitor. Though the performance of this mode is still better than other active power supplies, but the dynamic ESR will still be higher than the class A mode because the ultracapacitor current crosses zero frequently. And even more, the noise of the power supply that attached can still be introduced into the voltage rail to make noise performance worse than the class A mode.

The sound quality difference between class A and class B can be heard clearly if we use them as power supplies for clock circuits such as the FifoPi clean side or the TWTMC oscillators.

UcPure is a heavy duty design for a durable operation. It can use both AC and DC as input. The quality of the input doesn't matter because the input will be disabled when UcPure is turned on.

Ian
Hi Ian, Can you give the link to buy some of the clear shrink wrap that is in the picture?