ProtoDAC TDA1387 X8 project

[michaelagiles]

Thank you for posting this.

The MA and Q7 manuals state "or other audio HAT on top of FifoPiMA." Do all other audio HATs use only 3.3VDC?

The TDA1387 will operate from supplies of 3.0 to 5VDC, but the typical current output at 3.0 VDC is only 0.6mA compared to 1mA at 5VDC (per chip), resulting in lower output voltage after I/V.

Do you know if the 5VDC from MA J3 is isolated from the Raspberry Pi? I think this also applies to Q7. It states that the 5VDC comes from the Q7 GPIO.

If you connect 5vdc  to J3 AND jumper J4 1-2,   3-4 this will power both the MA and the  Rpi with both 5v and 3.3v.  I am not sure what happens if you un-jumper J4 except I am certain you will need a 3.3v supply connected to J5 to power the MA.  I THINK the J3 5v will still power the Rpi but this needs to be better investigated. Don't take my word for it!

I have had another issue arise that maybe someone can help with. I have setup SOX resampling to selectively resample to 384 but adhere to the base frequency.  This works for 44.1, resampling to 352 but on 48 base data it still indicates 352 rather than the correct 384 when I check Audio Info. Is this what is actually happening or is it a reporting error?
Can it be corrected?
Pi 4b, FifopiMA, ProtoDAC, Pass B1 , Pass SLOB active Xover, Luxman and Ayima, OB Markaudio and Dayton Audio.

[michaelagiles]

Thank you for posting this.

The MA and Q7 manuals state "or other audio HAT on top of FifoPiMA." Do all other audio HATs use only 3.3VDC?

The TDA1387 will operate from supplies of 3.0 to 5VDC, but the typical current output at 3.0 VDC is only 0.6mA compared to 1mA at 5VDC (per chip), resulting in lower output voltage after I/V.

Do you know if the 5VDC from MA J3 is isolated from the Raspberry Pi? I think this also applies to Q7. It states that the 5VDC comes from the Q7 GPIO.

If you connect 5vdc  to J3 AND jumper J4 1-2,   3-4 this will power both the MA and the  Rpi with both 5v and 3.3v.  I am not sure what happens if you un-jumper J4 except I am certain you will need a 3.3v supply connected to J5 to power the MA.  I THINK the J3 5v will still power the Rpi but this needs to be better investigated. Don't take my word for it!

I have had another issue arise that maybe someone can help with. I have setup SOX resampling to selectively resample to 384 but adhere to the base frequency.  This works for 44.1, resampling to 352 but on 48 base data it still indicates 352 rather than the correct 384 when I check Audio Info. Is this what is actually happening or is it a reporting error?
Can it be corrected?
Pi 4b, FifopiMA, ProtoDAC, Pass B1 , Pass SLOB active Xover, Luxman and Ayima, OB Markaudio and Dayton Audio.

BTW I found a very noticeable deterioration with the 3.3 v supply to the ProtoDAC before I realized what was happening.
MG

[martin155]

Few posts back the @OnkelPH mentioned using Mundorf MCap EVO Oil 6,8 uf. I was wondering if there is somebody else who tried some interesting output caps with positive results?

[OnkelPH]

Few posts back the @OnkelPH mentioned using Mundorf MCap EVO Oil 6,8 uf. I was wondering if there is somebody else who tried some interesting output caps with positive results?

I’ve ordered a Q7 mk2 and found some second hand Mundorfs supreme silver oil 10uf that will arrive this week. I’m really excited to put this into the mix and hook it up to my Decware. Hoping to start listening at the weekend.

[hifinet]

Thank you for posting this.

The MA and Q7 manuals state "or other audio HAT on top of FifoPiMA." Do all other audio HATs use only 3.3VDC?

The TDA1387 will operate from supplies of 3.0 to 5VDC, but the typical current output at 3.0 VDC is only 0.6mA compared to 1mA at 5VDC (per chip), resulting in lower output voltage after I/V.

Do you know if the 5VDC from MA J3 is isolated from the Raspberry Pi? I think this also applies to Q7. It states that the 5VDC comes from the Q7 GPIO.

If you connect 5vdc  to J3 AND jumper J4 1-2,   3-4 this will power both the MA and the  Rpi with both 5v and 3.3v.  I am not sure what happens if you un-jumper J4 except I am certain you will need a 3.3v supply connected to J5 to power the MA.  I THINK the J3 5v will still power the Rpi but this needs to be better investigated. Don't take my word for it!

I have had another issue arise that maybe someone can help with. I have setup SOX resampling to selectively resample to 384 but adhere to the base frequency.  This works for 44.1, resampling to 352 but on 48 base data it still indicates 352 rather than the correct 384 when I check Audio Info. Is this what is actually happening or is it a reporting error?
Can it be corrected?
Pi 4b, FifopiMA, ProtoDAC, Pass B1 , Pass SLOB active Xover, Luxman and Ayima, OB Markaudio and Dayton Audio.

BTW I found a very noticeable deterioration with the 3.3 v supply to the ProtoDAC before I realized what was happening.
MG

This thread on diyAudio reports a similar problem with the FifoPi Q7.  The Q7 does not have similar J4 1-2 3-4 jumpers. I don't see the GPIO pin 2,4 issue addressed in the manual. 

I recently received a very kind donation of Ian Canada parts (which includes the FifoPi Q7) from a ProtoDAC kit customer. I will be investigating this problem soon.

[hifinet]

Normally with the MA, J3 (5V) only provides power to the Raspberry Pi below via J1. When J4 jumpers 1-2 and 3-4 are shorted, this apparently provides common J3 power (5V) to the MA and the top side GPIOs J2 and J3. J5 (3.3V "clean power") is then left unconnected. MA must have 3.3V regulators on board to supply power to the MA??

Question: What are the voltage readings of J4 jumpers when 1-2 and 3-4 are shorted? Are they both 5V or is 1-2 5V and 3-4 3.3V?

[michaelagiles]

Normally with the MA, J3 (5V) only provides power to the Raspberry Pi below via J1. When J4 jumpers 1-2 and 3-4 are shorted, this apparently provides common J3 power (5V) to the MA and the top side GPIOs J2 and J3. J5 (3.3V "clean power") is then left unconnected. MA must have 3.3V regulators on board to supply power to the MA??

Question: What are the voltage readings of J4 jumpers when 1-2 and 3-4 are shorted? Are they both 5V or is 1-2 5V and 3-4 3.3V?

My reading of the Ma manual is that the 3,3v regs are on the Rpi board and that this is available on pins 1 and 17 of the Rpi GPIO connector.  If the 1-2. 3-4 jumpers are NOT shorted then you must power the RPI directly and the Ma with 3,3 at ;5.

Pins 2.4 on the Rpi are 5v as you know and changed to 3,3 only on 1 of the Ma GPIO connectors. I had this confirmed directly by Ian. He was very responsive to direct email questions BTW.
.
I don't know the voltage on the jumper pins. And at the moment I am just enjoying the music!
MG

[hifinet]

In the Q3 manual, there is a chart of the GPIO on page 5. The Q7 manual doesn't have the chart, but it is probably the same. J7 is the isolated GPIO.

[img]<a data-flickr-embed=[/img]" />

The solution for Q3 and Q7 is to use a stacking header with GPIO pins 2 and 4 removed and power ProtoDAC directly from the 5V and GND pads between the RCA jacks with 5VDC.

Looking at the Gabby video again, it appears that he was powering ProtoDAC with 3.3V.

[hifinet]

I am powering my entire stack (RPi Zero W/Allo Kali/ProtoDAC) with the 5V PurePi output from the ultracapacitors. This connects to the barrel DC jack of Kali, which powers the stack. Kali does have some input filtering.

I really like the sound. Surprisingly, the soundstage is a tiny bit smaller than compared to the $15 Allo 5V SMPS. Everything else is much better. The haze of noise is gone. Better depth, definition, detail and bass.

I think the design of PurePi seems very odd. The LiFePO4 batteries provide power only for the 3.3V supply. The two 100F ultracapacitors in series provide power only for the 5V supply. The USB-C SMPS charges both.

So that means in normal operation, the batteries are providing power to the FifoPi Q7 and DAC HAT. The ultracapacitors are providing power to the Raspberry Pi. Very odd. Seems to be a real waste of ultracapacitors.

[LordShadow]

That is odd. From the impression I had the whole thing ran from the ultracaps and the batteries were just a backup but it seems I was completely off base on that.