Soekris Dam Dac

[nige2000]

you could use nimh it has been talked about but no one has tested it

It is the synchronicity of the i2s lines primarily im concered about
Thats the 22 and 24mhz family of clks
In the pc usb is clocked at 24mhz usb frequency is multiples of 12mhz
Usb hub or regen is 24mhz
And xmos can be programmed to most multiples of 12mhz
But most manufacturers use 48mhz which is unfortunate coz we dont get a clean sweep

[randytsuch]

I'm looking through randy's journal right now. Do i understand correctly : the endgame is to have just one clock in the FPGA linked to the SDtransclone, instead of two ? But when using the PC there will be only 2 drummers, instead of three ?


I'm wondering why you can't use 1.2V NiMH for the 1.2v side.

My blog does not currently change the clocks, that is something new that I have not tried yet.
Curious to see what comes out of this.

NiMH are harder to charge, they need special charging circuits, versus lifepo4 batteries which can be easily trickle charged with a constant voltage circuit.
And not all batteries are equal, lifepo4 batteries seem to be "special" for some reason.
That's not to say that there could be a better way to generate 1.2V. I was fairly lazy with that one, I found a simple was to do it, and I just did it.

For projects like this, sometimes you can go through paralysis by analysis. You spend too much time trying to find the "perfect" implementation when you could just try it.

I think that's why both Nige and I (we independently did this) get a big block of wood, mount the board, and start wiring things up. Its easy to do, and easy to change.

Nige and Rick
I was thinking it would be good to add pictures of your implementations to my blog page, since we have all approached this in similar, but slightly different ways.
So, when you guys get a chance, if you could send me some pics that best illustrate your dacs and the mods made, I'll add them to my blog.
I know there are many pics in this thread, but I'll let you guys choose which ones you want me to use, and I'll just add them with the appropriate credit. (also allows me to be lazy :) )

Randy

[jrling]

rolandpsp's question got me thinking.

For our purposes there are three general types of rechargeable battery usable:

Ni-CD
Ni-MH
Li-ion

Without thinking about it, I assumed the order above was 'worst to best'. However ......

see here: http://www.ti.com/lit/an/snva533/snva533.pdf from a very authorative source.

Whilst Li-ion has many advantages, particularly power to weight ratio and ability to pass very high currents, both of those attributes are not so interesting to us. They are also expensive and potentially explosive.

What surprised me was that for a 1.2v source as required here, Ni-Cd had a significantly lower ESR (internal resistance) than Ni-Mh - completely the opposite to my intuition. Not by a small margin either approximately 10 times worse:

Internal resistance of batteries -
Type AA
Ni-Cad 0.03 -
Alkaline 0.12 -
Ni-Mh 0.35 -

This paper also shows the biggest advantage of Ni-Cd and Ni-Mh batteries: their discharge curve is extremely flat, closest to an ideal battery.
This important difference between the battery types means that Ni-Cd and Ni-Mh cells are well suited for use with linear regulators, but Li-Ion batteries require switching converters to obtain good energy conversion efficiency in the power supply.


Between Ni-Mh and Ni-Cd again to my surprise, Ni-Cd shows significantly lower self-discharge rates if not used. The Eneloop type Ni-Mh clearly are trying to address that disadvantage in Ni-Mh.

My Conclusion
For powering 1.2v electronics, Ni-Cd is the best option. The cheapest technology, easily charged and most importantly the lowest internal resistance (ESR) and therefore 'noise'.
Chuck out those dangerous LiPos!

Jonathan

[nige2000]

ive tried ni-cd in the past there not as good sounding as lifepo4

id kinda go with this order
lifepo4
li-ion
then the rest

maybe a little shunt reg will do the job here

[Val33]

Hi All,

New here, but have been reading all the threads on here and Diyaudio. I'm not new to messing with digital, having built a few DACs over the years and modded stuff too.

I love what you have been doing with the dam dac and am on the verge of ordering one to follow your lead.

I am considering feeding it from a Pi3 with Hifiberry dac+ pro Both with separated linear/battery power supplies, just taking the I2s from the dac+. I like the idea of the correct master clock used in this combo (very similar to the cronus/botic for the BBB which I also have, linked to a BIII, which I don't enjoy at all). In my reading I have seen Soekris say that the FPGA SLV pin would allow input of MCLK ( http://www.diyaudio.com/forums/digital- ... ost4219849 ) but he also does not say if it works! I can't find anywhere where anyone has actually tried this?? I would guess that the Dam Dac clock would need to be disconnected.

Thoughts?

Thanks

Val

[randytsuch]

Hi All,

New here, but have been reading all the threads on here and Diyaudio. I'm not new to messing with digital, having built a few DACs over the years and modded stuff too.

I love what you have been doing with the dam dac and am on the verge of ordering one to follow your lead.

I am considering feeding it from a Pi3 with Hifiberry dac+ pro Both with separated linear/battery power supplies, just taking the I2s from the dac+. I like the idea of the correct master clock used in this combo (very similar to the cronus/botic for the BBB which I also have, linked to a BIII, which I don't enjoy at all). In my reading I have seen Soekris say that the FPGA SLV pin would allow input of MCLK ( http://www.diyaudio.com/forums/digital- ... ost4219849 ) but he also does not say if it works! I can't find anywhere where anyone has actually tried this?? I would guess that the Dam Dac clock would need to be disconnected.

Thoughts?

Thanks

Val

He said "might".
I'm pretty sure that it is an output right now, and he would need to create a new FPGA version to make it an input.
If it was an input, Nige would have tried it already ;)

[Val33]

He said "might".
I'm pretty sure that it is an output right now, and he would need to create a new FPGA version to make it an input.
If it was an input, Nige would have tried it already ;)

Thanks for the reply, but the next pin is labeled as FPGA MCLK OUT.

Just wondered if anyone had confirmed that it did not work?

Val

[gstew]

Hi All,

New here, but have been reading all the threads on here and Diyaudio. I'm not new to messing with digital, having built a few DACs over the years and modded stuff too.

I love what you have been doing with the dam dac and am on the verge of ordering one to follow your lead.

I am considering feeding it from a Pi3 with Hifiberry dac+ pro Both with separated linear/battery power supplies, just taking the I2s from the dac+. I like the idea of the correct master clock used in this combo (very similar to the cronus/botic for the BBB which I also have, linked to a BIII, which I don't enjoy at all). In my reading I have seen Soekris say that the FPGA SLV pin would allow input of MCLK ( http://www.diyaudio.com/forums/digital- ... ost4219849 ) but he also does not say if it works! I can't find anywhere where anyone has actually tried this?? I would guess that the Dam Dac clock would need to be disconnected.

Thoughts?

Thanks

Val

I am running a DAM from an R-Pi I2S AND also have a HiFiBerry DAC+ Pro I've modified a good bit and am also playing around with.

First the DAM from an R-Pi using the generic I2S driver (I'm running PiCorePlayer) sounds VERY, VERY NICE. While I am quite sure that the experiments Nige is performing regarding better clocking will result in better sound, the stock clocking setup fed from a Pi is pretty darned good!

Then you can read about my experiences with the HiFiBerry so far here: http://www.diyaudio.com/forums/digital- ... ody-3.html . Details on what I've done to the thing so far are scattered through the thread.

Bottom line is I am not too impressed with the HiFiBerry yet, stock or modified, even though I've spent a good bit of time and effort working the device and want to be more impressed. I'm not convinced that the clocking scheme HFB uses is the best and I am very sure that the clocks onboard are not as good as the not-the-best variable frequency clock on the DAM. Unless you are prepared to hack-in better clocks on the HFB DAC+ Pro, I suspect you will do as well or better with the stock DAM clock scheme on a stock Pi. The I2S feed from the HFBD+P MIGHT be more bit-perfect than that of the stock Pi, but I'm not convinced it will sound better than the stock Pi feed on a DAM still using the stock isolation and reclocking. Still, I haven't tried that hookup yet. It is on my list of things to try, but not very high at this time, sorry.

Greg in Mississippi

[randytsuch]

He said "might".
I'm pretty sure that it is an output right now, and he would need to create a new FPGA version to make it an input.
If it was an input, Nige would have tried it already ;)

Thanks for the reply, but the next pin is labeled as FPGA MCLK OUT.

Just wondered if anyone had confirmed that it did not work?

Val

I haven't, but I have not seen anywhere where he said you could feed a clock to the FPGA.
Everything I've seen is that you can't feed an external clock into this DAC.

[nige2000]

He said "might".
I'm pretty sure that it is an output right now, and he would need to create a new FPGA version to make it an input.
If it was an input, Nige would have tried it already ;)

Thanks for the reply, but the next pin is labeled as FPGA MCLK OUT.

Just wondered if anyone had confirmed that it did not work?

Val

I haven't, but I have not seen anywhere where he said you could feed a clock to the FPGA.
Everything I've seen is that you can't feed an external clock into this DAC.

id imagine theres a clock switch option to be implemented yet in future fpga firmware