Those 5v should come from any of the dac's power supplies. Once you switch the dac on, the battery should be automatically switched from the charger to the clock.
Not sure how strict is the stated 3.6v max from the Crystek datasheet. Exceeding this slightly won't blow anything out short term but longevity is bound to suffer. A solution would be to use two batteries in series and then regulate down to 3.3v but there have been reports that any regulator will make the sound suffer.
I intend trying the trickle-charge option with the battery permanently connected to a regulated and current limited source of 3.3v. It will be interesting how this will work out sonically as it is the least painful solution.
Hi Eldam,
Thanks for your answer. To be more presice:
1) raspberry pi -> i2s & power (over GPIO) to the Fifo v2.
2) Fifo -> isolator board
3) isolator board -> clock board
4) clock board & fifo -> s/pdif board (output)
5) (spdif output to DAC)
I am powering the RPi with a lineair power supply. If I understood the documentation correctly, all is powered from the Fifo board, however, when using the isolator board (which I will be doing), then the clock board needs its own power like battery, or 3.3V with a LDO voltage regulator
My question is simple: When I am using the isolator board, Is the S/PDIF board being powered by the Fifo? or should it be powered by a separate power supply?
If there is a good solution to keep a battery in place rather than creating a lineair PSU for this, I would be happy. Because now, the whole battery management is a little bit weird to me too. It's not clear on how it works.
Last edited:
I didn't know a Spidf output board was existing to be feeded from the Clock Board (or the Spidf in board could be managed as an output board as well if the case) ! Which explain my non understanding about your initial question!
Well for the battery charger embeded solution for a standalone LifePo4 cell, I found some cheaper product, an eval board at 22 euros w/o VAT : it needs a minimum 4V input : MCP73X23EV-LFP - MICROCHIP - CARTE D'EVAL MCP73X23 OVP CHARGEU DE BAT | Farnell element14 : it seems to has the complet profile charge embeded with protection of the cell as well (automatic shutdown after 6 hours !)
and the cheap chip alone : MCP73123-22SI/MF - MICROCHIP - CHARGEUR BATTERIE LI IRON PHOSPH 10DFN | Farnell element14 ! (very little !)
Thanks Analog SA, clearer to me now about the 5V of the shematic.
Well for the battery charger embeded solution for a standalone LifePo4 cell, I found some cheaper product, an eval board at 22 euros w/o VAT : it needs a minimum 4V input : MCP73X23EV-LFP - MICROCHIP - CARTE D'EVAL MCP73X23 OVP CHARGEU DE BAT | Farnell element14 : it seems to has the complet profile charge embeded with protection of the cell as well (automatic shutdown after 6 hours !)
and the cheap chip alone : MCP73123-22SI/MF - MICROCHIP - CHARGEUR BATTERIE LI IRON PHOSPH 10DFN | Farnell element14 ! (very little !)
Thanks Analog SA, clearer to me now about the 5V of the shematic.
Last edited:
I don't know either.. that's why I am asking about the s/pdif board, haha. Not sure how that works ;-)
I already got the LDO and the battery manager unit.. just to play around with. I am going to see how that works. Pitfall of not knowing electronics and have any tools to test properly unfortunately.
I already got the LDO and the battery manager unit.. just to play around with. I am going to see how that works. Pitfall of not knowing electronics and have any tools to test properly unfortunately.
Hi Ernst,
My multi-channel FIFO will have programmable delay time. If you are in video mode, you can use very short delay time, says 0.1 second. That will 100% solve your problem. That delay time will be independent from Fs.
For now, try to use higher Fs or up-sample your I2S, that can be better for video mode.
Regards,
Ian
Hi Eldam,
I usually use one 5V power supply from DAC or system to control the battery board. When the 5V is on, battery will be in working mode. When it is off, battery will be switched to charging circuit.
Battery power is totally independent from the control voltage and ground.
You don't need any manual on/off.
Hope this idea works for you.
Happy New Year!
Ian
Hi Gitaarwerk,
S/PDIF board will be powered by FIFO board. If you really want an independent power supply for it, you have to disconnect the power from FIFO in advance.
Regards,
Ian
If you don't use DIT function of S/PDIF board, you are good with FIFO power supply.
Ian
Hi Eldam
Not using the management board, so cannot really tell you much. From what i recall it automatically switches the battery on when power is applied.
Not sure if this has been discussed - it probably has - but it's still worth discussing: how do you intend to get the battery charged to 3.3v +/-0.3v?
Most of these batteries leave the charger at something above 3.7v which apparently is not great for the oscillator circuits lifespan. And how do you monitor the battery has not been excessively discharged? It is a project in its own right.
There is a similar current discussion on the asylum where Jkeny advocates using the battery in the manner of a capacitor and constantly trickle charging it. This solves most issues and you will need the management board to make sure the battery is disconnected when power is off, but of course you no longer have an isolation from the mains. Jkeny says it doesn't matter soundwise and he may be right.
From my limited experiments with a carefully charged up to 3.5v battery i can say i am stunned. Not so much by the improvement - it has been expected - but by the fact that the battery preserves it's own "sound", the same it has powering analogue electronics and even my tt motor. Too weird, innit? This goes further towards components having a sound irrespective upon where and how are they used. Not gonna please the engineers much
Hi analog_sa,
I use once cell of A123 LifePo4 battery. After being fully charged, the voltage is around 3.4V, I use the voltage range above 3.2V, which will cove 80% of battery capacity. 3.7V would be too high.
Happy New Year.
Ian
Thank you Ian,
Great.
So the genuine A123 stays at 3.4 V fully charged : so safe from the first beginning !
...and can drop enough if one want to use your Tsa7 ldo after.
OK. I relink the datasheet for this post to serve as a memo : http://a123batteries.com/v/vspfiles/images/pdf/26650.pdf
I will try myself with the jumper instead as I have not your ldo. I read you tried both and after one comeback, you now prefer without ldo after the A123 cell....
Btw I have an equivalent Tsa7 board for the main plug I setuped at 3.3 V to wait the reception of the A123 cell.
@ analog sa & Gitarrwerk :
I found some other LiFePo4 but any with a so low 6 m ohms ESR if the A123 statment is true ! I plan to purchase mine at the adress of DVBprojekt member gave in this thread (thanks!) : A123 Systems ANR26650M1B, 2500 mAh, Lipopower, LiNANO : said to be genuine as many counterfacts exists !
For North America there is http://www.a123systems.com/solutions-commercial-products-group-vacps.htm
I would like to find a 26650 size box with spring for the cell (I didn't find it yet), though you can purchase also with the U or Z shape (if such charger board as the one above I linked is your choice)
Great.
So the genuine A123 stays at 3.4 V fully charged : so safe from the first beginning !
...and can drop enough if one want to use your Tsa7 ldo after.
OK. I relink the datasheet for this post to serve as a memo : http://a123batteries.com/v/vspfiles/images/pdf/26650.pdf
I will try myself with the jumper instead as I have not your ldo. I read you tried both and after one comeback, you now prefer without ldo after the A123 cell....
Btw I have an equivalent Tsa7 board for the main plug I setuped at 3.3 V to wait the reception of the A123 cell.
@ analog sa & Gitarrwerk :
I found some other LiFePo4 but any with a so low 6 m ohms ESR if the A123 statment is true ! I plan to purchase mine at the adress of DVBprojekt member gave in this thread (thanks!) : A123 Systems ANR26650M1B, 2500 mAh, Lipopower, LiNANO : said to be genuine as many counterfacts exists !
For North America there is http://www.a123systems.com/solutions-commercial-products-group-vacps.htm
I would like to find a 26650 size box with spring for the cell (I didn't find it yet), though you can purchase also with the U or Z shape (if such charger board as the one above I linked is your choice)
Last edited:
Thanks for your answers Ian... looks like i only now need to decide how I want to clock board to be powered. The rest is all simple enough. Cheers!
Looks good!
I'm going to invest some time in how I want the clock power to be =). Cheers for the link!
Last edited:
Ian, up to now VLC player does the job with negative delay. Will see if the multichannel Fifo is necessary and affordable for me. Cheers Ernst
Jan, or may consider to build a XVCO solution with a softer frequency control. I build this some decade ago and had buffer with 2 samples and over/under runs did not happen very often. It was build on DAC & XVCO and a PIC who controlled using optical isolation the DAC-XVCO...
Hp