Do you have a rough idea of battery life when powering Osc, DAC and Fifo?
Whichever rail drains fastest I guess would be the limiter?
Also about daughter board, is the output stage mentioned part of your DAC or the daughter board is an option for those wishing to add external output stage?
Whichever rail drains fastest I guess would be the limiter?
Also about daughter board, is the output stage mentioned part of your DAC or the daughter board is an option for those wishing to add external output stage?
Last edited:
The cycle life of the A123 ANR26650 battery is > 4000 cycles.
If we assume 2 hours every day of operation I believe you have to recharge the batteries every 3-4 days.
So from 12000 to 16000 days, around 30 years.
More realistic, if we assume recharging every day you can calculate at least 11 years.
If we assume 2 hours every day of operation I believe you have to recharge the batteries every 3-4 days.
So from 12000 to 16000 days, around 30 years.
More realistic, if we assume recharging every day you can calculate at least 11 years.
No, the footprints are suitable for battery holder only (1.2 x 2.4 mm hole).
We have used the HIFI 2000 1GX283N case
HIFI 2000 Boitier Galaxy GX283 3mm 80x230x230 - Audiophonics
We are upgrading the firmware in order to support Lithium Titanate batteries.
The daughter board TWRPS-LBS-D will accept 26650 or 18650 LTO batteries.
There will be a jumper to select the proper configuration according to the installed batteries (LiFePo4 or LTO).
The result will be a mixed setup with LiFePo4 batteries on the main board (for oscillators and digital devices) and LTO batteries on the daughter board (analog devices like the DAC reference).
I have just ordered a few LTO battery samples.
Attachments
Hello,
I am happy to see Supersurfer is still using my old Bourdereau.
Take a look at their turntables.
Greetings, Eduard
P.s recently Doede the DDDAC designer also has done some test with supercaps. In the eighties the French used dry lead batteries, combined with normal BIG caps and NEC supercaps ( just new on the market then)
They said they needed the caps for noise from the batteries.
So it could be like Supersurfer said a mix of lifepo4 and supercaps can be beneficial.
IF the Italian boards will allow gradually charfing at start up like the Canadian board t6hat would be perfect.
I am happy to see Supersurfer is still using my old Bourdereau.
Take a look at their turntables.
Greetings, Eduard
P.s recently Doede the DDDAC designer also has done some test with supercaps. In the eighties the French used dry lead batteries, combined with normal BIG caps and NEC supercaps ( just new on the market then)
They said they needed the caps for noise from the batteries.
So it could be like Supersurfer said a mix of lifepo4 and supercaps can be beneficial.
IF the Italian boards will allow gradually charfing at start up like the Canadian board t6hat would be perfect.
Attachments
Yes, it was not really for the noise that a simple wima cap between 0.1 or 0.47 uF seen for that purpose was seen. It was more about the bad ESR of lead batteies iirc (in fact all sort of batteries were used from truck to motorcycle, also dry like you mentioned : they used a lot of lytics tank as fast tank for the load not the batterie to be seen and to have a big inertia wheel, it was not rare to see them // with 5V memory caps in serie (that have a very bad ESR as well but were here for the inertia wheel effect like in a car) and big MKP film cap for the 'punch" (good low ESR) at the end and of course a local little fast cap or not (because oscillation) near some oaps. That were for CD players, but firstly it was for amps before by Hiraga that certainly saw that in Japan
Last edited:
Hello,
In the mc preamplifier i was talking about electrolytic caps they were using 6 times 470000 microfarad in the Autograph model of the Prepre.
The 5 volt memory caps was indeed used for the purpose you mentioned.
Todays supercaps are totally different.
If remember well they just used a 2,2 microfarad MKC cap as cherry on a pie.
Greetings, Eduard
P.s i think now ideas have changed and more people start using supercaps.
I think our Italian friend told us somewhere that not every part in the digital chain will work perfectly with a similar supply. I could imagine having a '' kind of energy storage '' close to the '' consumer '' and then a small decoupling in the form of a tiny LDO or shunt really NEXT to where it is '' consumed ''
If things get so big that all distance are becoming bigger it might not be beneficial.
If the big part of the supply with lifepo4 and/or supercaps takes care of 95 % of the noise will the sonic '' footprint '' of an active circuit CLOSE to the load be smaller because there is hardly anything to filter?
Greetings, eduard
In the mc preamplifier i was talking about electrolytic caps they were using 6 times 470000 microfarad in the Autograph model of the Prepre.
The 5 volt memory caps was indeed used for the purpose you mentioned.
Todays supercaps are totally different.
If remember well they just used a 2,2 microfarad MKC cap as cherry on a pie.
Greetings, Eduard
P.s i think now ideas have changed and more people start using supercaps.
I think our Italian friend told us somewhere that not every part in the digital chain will work perfectly with a similar supply. I could imagine having a '' kind of energy storage '' close to the '' consumer '' and then a small decoupling in the form of a tiny LDO or shunt really NEXT to where it is '' consumed ''
If things get so big that all distance are becoming bigger it might not be beneficial.
If the big part of the supply with lifepo4 and/or supercaps takes care of 95 % of the noise will the sonic '' footprint '' of an active circuit CLOSE to the load be smaller because there is hardly anything to filter?
Greetings, eduard
Unfortunately the LiFePo4 battery supply system does not manage the soft charging for supercapacitors.
If you connect a fully discharged supercapacitor to the output of the board the firmware detects it as a short circuit and the system goes in safety protection.
Hello Andrea,
In post number 176 Stefan ( Supersurfer) wrote he realized an improvement by adding supercaps in parallel with lifepo4 batteries.
Doede ( the DDDAC man) has done some test with supercaps and his standard power supply. BUT before that one is connected he will first charge the supercaps with a pretty standard current limited battery charger when the voltage on the caps is close to the one delivered by the standard supply the charger will stop and the standard supply will keep the supercaps at the right voltage.
What is your idea about offering a cleaner input to the usual voltage regulators which are mounted close to the circuits they are supplying too? If they have to filter less noise will they '' behave '' better.
Greetings, eduard
In post number 176 Stefan ( Supersurfer) wrote he realized an improvement by adding supercaps in parallel with lifepo4 batteries.
Doede ( the DDDAC man) has done some test with supercaps and his standard power supply. BUT before that one is connected he will first charge the supercaps with a pretty standard current limited battery charger when the voltage on the caps is close to the one delivered by the standard supply the charger will stop and the standard supply will keep the supercaps at the right voltage.
What is your idea about offering a cleaner input to the usual voltage regulators which are mounted close to the circuits they are supplying too? If they have to filter less noise will they '' behave '' better.
Greetings, eduard
I don't know how Ian's supercapacitor boards work.
If they provide a current limiter to soft charge the capacitors then I believe they can be connected to our LiFePo4 battery system without problems.
Otherwise there is no way to get them working because our supply system will detect a short circuit and go in safety mode disconnecting the output.
Although I have not performed any test I'm not sure that high value capacitors are the right way to supply digital circuits.
I think the best way is using ceramic decoupling capacitors as close as possible to each digital IC.
Our oscillators and frequency doublers are relatively constant loads, so they don't need high capacitance.
We have tested them with LiFePo4 batteries and with a clean linear regulator, practically there is no difference.
The other devices.
If a FIFO buffer is used the source does not matter.
Using a FIFO buffer the source and the DAC operate in different time domains, so with proper isolation there is no need to improve the power supply of the source and of the input section of the FIFO (FPGA, micro and so on).
Indeed the input section of our new FIFO Lite is powered by a standard linear regulator, while the output section is optically isolated and powered by battery like the oscillators.
The same for the digital section of the DAC, ceramic decoupling capacitors as close as possible to each ICs.
Finally, the analog section of the DAC.
I cannot say about Delta-Sigma DAC because I have not the proper skills about them.
Using R2R DAC like our new DAC Lite there will be a strong energy requirement every time one or more switches change their state.
But this high amount of energy is only required for the switches that are changing their state and not for the whole DAC.
That's the reason why we have installed a film capacitor close to each ladder switch.
In other words, we believe that the most important power supply improvement comes from the decoupling capacitors placed close to each IC, both in the digital and the analog section.
Ultracapacitors make the chemical battery noise disappear
Hi Andrea,
I have experimented a lot with ultracapacitors, from 5F up to 3000F in size. It is not the capacity that has a positive influence on the power supply, it is the accompanying very low ESR.
5F esr=35mohm
325F esr=1,7mohm
3000F esr=0,29mohm
0,1uf Ceramic esr=100mohm
Using the 325F does deliver a clearly audible lower noise floor when used in parallel to the battery power supply when it is feeding a fifopi and your clock units (while the rest of the streamer and dac use wall power.
So my analysis is that it certainly lowers the already low battery power noise.
I reccomended you to experiment with these caps!
PS. The Bourdereau is still working nicely Eduard!
Regards,
Hi Andrea,
I have experimented a lot with ultracapacitors, from 5F up to 3000F in size. It is not the capacity that has a positive influence on the power supply, it is the accompanying very low ESR.
5F esr=35mohm
325F esr=1,7mohm
3000F esr=0,29mohm
0,1uf Ceramic esr=100mohm
Using the 325F does deliver a clearly audible lower noise floor when used in parallel to the battery power supply when it is feeding a fifopi and your clock units (while the rest of the streamer and dac use wall power.
So my analysis is that it certainly lowers the already low battery power noise.
I reccomended you to experiment with these caps!
PS. The Bourdereau is still working nicely Eduard!
Regards,
Attachments
Last edited:
Hi Stephan,
maybe we will give a try in the future, now we are busy (GB aside) with the development of the FIFOs and the DACs.
However LTO batteries has very low ESR as you can see in the attached sheet.
