Hi Abra,
WHich voltage ? For the C1+ ?
I'm verry interrested to know how it performs instead LiPo ! In theory LiPoFe4 should be better (lessier noise floor and better ESR !
There are some little moto batteries in LiPoFe4 not so expensive while more expensive than the little LiPo external charger for cellulers which are very cheap (around 20 euros with led indicator of the capacity level).
At the FIFO owners : anyone can listen a difference with a Pi board outputting with I2S (40 pins plug) and the ufl wires (IanCanada little adaptator) and simple wires (no impedance matched) ?
WHich voltage ? For the C1+ ?
I'm verry interrested to know how it performs instead LiPo ! In theory LiPoFe4 should be better (lessier noise floor and better ESR !
There are some little moto batteries in LiPoFe4 not so expensive while more expensive than the little LiPo external charger for cellulers which are very cheap (around 20 euros with led indicator of the capacity level).
At the FIFO owners : anyone can listen a difference with a Pi board outputting with I2S (40 pins plug) and the ufl wires (IanCanada little adaptator) and simple wires (no impedance matched) ?
We have tried two kind of batteries: Lithium Ion vs. LiFePo4 with the Cubietruck/Cubieboard3.
The CT has two power inputs: a normal 5V and a battery input at 3.6V for single cell Lithium Ion batteries. It is intended to use as smart UPS. When the LiOn is not properly charged, and the 5V input is available, a control chip automatically charges the battery from the 5V input. If the 5V input is not available, and the battery is properly charged, the control chip uses this battery as power supply. It continuously checks the state of the battery, and if the voltage is too low, shuts down the SBC.
And it was a problem at the test. We know this issue
The LiFePo4 nominal voltage is 3.2V, which is lower, than the LiOn has. But after a fresh charging, it started with 3.43V as I measured. After a few minutes of playback the voltage dropped down to its nominal voltage, and the control chip (who expected LiOn higher voltages) initialised a shut down.
Just as we expected. But according the practice, we had 3-4 minutes only to listen to the LiFePo4, which is not enough fairly compare the different /batteries. W had only unchecked impressions how it sounds.
Now the question is how to modify control chip/RuneAudio to handle LiFePo4 correctly, or to find own control circuit for it. Maybe the control chip can be omitted from the chain temporarily, maybe the software can be configured to discard the shut down interrupt, or the shut down voltage can be set to lower value. Any idea?
The CT has two power inputs: a normal 5V and a battery input at 3.6V for single cell Lithium Ion batteries. It is intended to use as smart UPS. When the LiOn is not properly charged, and the 5V input is available, a control chip automatically charges the battery from the 5V input. If the 5V input is not available, and the battery is properly charged, the control chip uses this battery as power supply. It continuously checks the state of the battery, and if the voltage is too low, shuts down the SBC.
And it was a problem at the test. We know this issue
The LiFePo4 nominal voltage is 3.2V, which is lower, than the LiOn has. But after a fresh charging, it started with 3.43V as I measured. After a few minutes of playback the voltage dropped down to its nominal voltage, and the control chip (who expected LiOn higher voltages) initialised a shut down.
Just as we expected. But according the practice, we had 3-4 minutes only to listen to the LiFePo4, which is not enough fairly compare the different /batteries. W had only unchecked impressions how it sounds.
Now the question is how to modify control chip/RuneAudio to handle LiFePo4 correctly, or to find own control circuit for it. Maybe the control chip can be omitted from the chain temporarily, maybe the software can be configured to discard the shut down interrupt, or the shut down voltage can be set to lower value. Any idea?
Last weekend the linear power supply was finished by my audio buddy. Contains two pcs. of LT1764, one big 10000uF and two 47uF non-polar Black Gate electrolitics, Cree silicon carbide Schottkies, can operate at 6A/5V theoretically, more than enough for the CubieTruck. I had bought the Black Gates a few decades ago from Japan, when it was in production. They are original, not the faked ones sold on ebay at very high price nowadays.
It was listened against the Lithium Ion battery. The linear power supply was the better by a rather large margin.
Overcharging our third PSU candidate the LiFePo4 battery, we were capable to run the LiFePo4 battery also as power supply (the CubieTruck expects a Lithium Ion with a bit higher voltage). We had about 5 minutest now to listen to the LiFePo4, when it's output voltage dropped below the Cubietruck excepted as minimum, and the Cubie switched off.
Not enough for a real test, but i was very surprised how the 10Ah single cell LiFePo4 performs against a decent linear PSU. Maybe it's better or not - a later thorough listening can decide -, LiFePo4 seems to be competitive.
Yesterday the Cubie's input power control chip was reset for the LiFePo4 voltage range succesfully. Today I got a SMS: after 21 hours of continous run, the battery with 3.23V is till runs at it's nominal voltage. Shotdown will be at 2.6V. It's capacity seems to be enough for a very long run. Takes too long time to finish this initial test, but very adequate for normal usage.
I also discovered a new (for me at least) battery type called Lithium Titanate (LTO). It has impressive parameters, and doesn't contains ferromagnetic materials like the LiFePo4. But it operates in the 2-2.4 voltage range, so it seems to be difficult if not impossible to apply a simple single cell LTO PSU solution for the Cubie's nice battery input.
It was listened against the Lithium Ion battery. The linear power supply was the better by a rather large margin.
Overcharging our third PSU candidate the LiFePo4 battery, we were capable to run the LiFePo4 battery also as power supply (the CubieTruck expects a Lithium Ion with a bit higher voltage). We had about 5 minutest now to listen to the LiFePo4, when it's output voltage dropped below the Cubietruck excepted as minimum, and the Cubie switched off.
Not enough for a real test, but i was very surprised how the 10Ah single cell LiFePo4 performs against a decent linear PSU. Maybe it's better or not - a later thorough listening can decide -, LiFePo4 seems to be competitive.
Yesterday the Cubie's input power control chip was reset for the LiFePo4 voltage range succesfully. Today I got a SMS: after 21 hours of continous run, the battery with 3.23V is till runs at it's nominal voltage. Shotdown will be at 2.6V. It's capacity seems to be enough for a very long run. Takes too long time to finish this initial test, but very adequate for normal usage.
I also discovered a new (for me at least) battery type called Lithium Titanate (LTO). It has impressive parameters, and doesn't contains ferromagnetic materials like the LiFePo4. But it operates in the 2-2.4 voltage range, so it seems to be difficult if not impossible to apply a simple single cell LTO PSU solution for the Cubie's nice battery input.
Cubietruck works well with LiFePo4 battery
Cubietruck worked 28 hours with the 10 Ah single cell battery when terminated at 2.8V. Seems to operate well after the AXP209 control chip was set specially to the LiFePo4 voltage requirements. We don't know why it was stopped at 2.8V when it was programmed to stop at 2.6V? But the essence is it worked properly, and also stopped when the battery was discharged.
Cubietruck worked 28 hours with the 10 Ah single cell battery when terminated at 2.8V. Seems to operate well after the AXP209 control chip was set specially to the LiFePo4 voltage requirements. We don't know why it was stopped at 2.8V when it was programmed to stop at 2.6V? But the essence is it worked properly, and also stopped when the battery was discharged.
Is not the clean battery power fed into a regular SMPS in the AXP209, supplying the SoC AXP209 - linux-sunxi.org
Thank you for the link. The .pdf datasheet also mentioned there was used for changing the voltages. And in practice, the 2.8V is also appropriate.
The LTO is currently in experimental phase, under testing. Hard to find especially with appropriate charger. Low output voltage requires battery packs, making life to be more complicated.
The internal impedance is tipically below 1 mOhm, but it may depend on a few thing, see here an example: 60 Amp Hour Cell | Altairnano
Sorry I have no assured information of noise.
At last I have chosen the linear power supply with Black Gate caps for a Cubietruck.
The LiFePo4 battery power supply couldn't exactly match the the speed of the BG capacitor, the LPS is also more balanced with it's deeper low frequency energy and more impact, and there is no need for regular charging.
I still think the battery supply is a good PSU, and if you haven't an Electron Transfer black gate on the shelf, it worth a try.
The LiFePo4 battery power supply couldn't exactly match the the speed of the BG capacitor, the LPS is also more balanced with it's deeper low frequency energy and more impact, and there is no need for regular charging.
I still think the battery supply is a good PSU, and if you haven't an Electron Transfer black gate on the shelf, it worth a try.
You may want to read up some of the threads here. While certainly specific to the vendor (Twisted Pear Audio), the fact, that they experimented with the RasperryPi and settled on the BeagleBoneBlack, two other SBC single board computer, may have lead to interesting findings. It seems they have done a lot to the clocking, like creating an external clock, etc.
Maybe you will find some interesting info, even if it does not solve your problem, per se.
Maybe you will find some interesting info, even if it does not solve your problem, per se.
44.1 Better/as Good ?
Peterma: I realize this/your reply was 2016 -and you'd be right (concerning 44.1 vs. HiRez), but omg, todays Internet Radio streams (of 128 kb/s) sound amazing ! (Using RP3B, Gustard U-12/i2S to X-20PRO DAC -using AQ 'Forest' HDMi cable -much better than model ('Pearl?) below it.
(Interestingly, AQ's own Carbon/Vodka HDMi cable sounded considreably worse -in my application. Awaiting a 1m AQ 'Cinnamon' HDMi cable. )
HiRez, MQA (in 2018) ? Gimm'e a break.
pj
Peterma: I realize this/your reply was 2016 -and you'd be right (concerning 44.1 vs. HiRez), but omg, todays Internet Radio streams (of 128 kb/s) sound amazing ! (Using RP3B, Gustard U-12/i2S to X-20PRO DAC -using AQ 'Forest' HDMi cable -much better than model ('Pearl?) below it.
(Interestingly, AQ's own Carbon/Vodka HDMi cable sounded considreably worse -in my application. Awaiting a 1m AQ 'Cinnamon' HDMi cable. )
HiRez, MQA (in 2018) ? Gimm'e a break.
pj
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.