My new Boss - another PI-HAT

Status
Not open for further replies.
I mailed cdgames/Allo and received a reply: soldering a 0 resistor on the R828 place is indeed the solution.

Hi All,
To work from external 5V ( BOSS J4 connector) need to solder 0 ohm resistor on R828 Location.
-Remove (desolder) R826 - 0 Ohm
-Solder 0 ohm on R828 Location.
Regards


Attached were updated BOSS External Power Connector Instructions which, undoubtly, will be published on their website monday.

I followed the updated instructions - I moved the 0 resistor from R826 to R828 - and now it works!
 
Yep. I reported that already. Not nice, rather cosmetics. In bcmrpi_defconfig still both are missing!
However.
What matters is that it works now!

Cool, just backing you up and I know the piCore, Moode, Allo (Ioan), etc. guys see this thread.

As for 384K, are the patches in a separate GitHub branch yet to start (preparing for a PR into official)?

Was just gonna do a piZERO/piZEROW compile, thanks re: bcmrpi_defconfig.
 
Last edited:
@DJ le Roi,

Oops, you was going to let you know this alternative before soldering, both will work fine.

You can power the boss from it's header: (pin14,15 GND) (pin1,2 5V)

I use a zero with 5V, 3.3V not connected.
 

Attachments

  • boss_ext_power.JPG
    boss_ext_power.JPG
    430.3 KB · Views: 578
  • IMG_0873.jpg
    IMG_0873.jpg
    1,023.5 KB · Views: 561
Another set of my comments to put Soundcheck to sleep...

It is hard to watch everyone else having fun with their Boss DAC while mine is just sitting here while I'm in a very busy work+family+stereo-repair-upgrade-for-a-friend time. Hopefully when I can get to it in early May, all the driver & patch stuff will be settled.

I was very happy to see how Ioan and Allo put it together... the SMD PPS (acrylic) film caps have worked very well for me in a number of builds/mods and I was happy to see them adopt the .33F cap from Joe Rasmussen's thread here:

http://www.diyaudio.com/forums/digi...entations-alternative-post-dac-filtering.html

I hadn't tried them yet at that time, but have since and been very happy with the results. They are now part of my standard toolkit for builds.

Some thoughts...

1. I saw comments on burn-in time on the Boss. My experience with the PPS caps is about a week or a little less until they settle in, BUT the .33F caps take about 2-3 weeks before a sandy-textured background finally fades away. Best to keep the unit on and playing music that whole time... you will be rewarded. And when I get more time, I'm adding one of these to both the analog and digital supplies of my modified HFBD+P, following in Allo's and Joe R's footsteps.

2. While I like what Allo did to filter the power from the Pi, I can't imagine it won't sound better with separate power. My rule of thumb is that once you have noise in your setup, filtering can only reduce it, not eliminate it... and the filtering has tradeoffs. I have been amazed with how good my modified HiFiBerry DAC+ Pro is sounding with separate power to the digital and analog sides of the DAC chip through OPC's paralleled LT3042 boards (with about 17000uf per side), all being powered by 3 Uptone Audio LPS-1 floating UltraCap supplies (1 also for a Kali). This rig is pushing into the same realm as I have been getting from my somewhat modified Soekris DAM DACs... time to do the next stage of work on them!

3. Even well-powered as above, I still prefer the also PCM5122-based HFBD+P with an also well-powered Kali in Slave mode... more defined and transparent top-to-bottom. My theory is that as good as the clock starts on the DAC board, sending it through the R-Pi's circuits and software likely add significant jitter before the I2S signals get back to the DAC. As much as I like the idea of a fully synchronous setup, at least with the HFBD+P, the Kali reclocker is still a better option. Curious what people will find with the Boss with it's already better clocks.

4. Since Ultrafi's big writeup on the HFBD+P mod thread, I have tried his then-recommended cheap SMPS to power my R-Pi. I could hear it sounding more dynamic, but also more grungy. I heard the same on the above-mentioned stereo gear repair/upgrade when I compared a slightly modified standard switching regulator board in his Sony HAP Z1-ES to my linear-regulator-modified board. The unit's switcher was again more dynamic sounding, but with added grunge. BUT after I put one of those .33F caps at the processor-board end of the cable from my linear-regulator board, the dynamics were equal between the 2 while retaining the greater purity of the linear setup. I've since tried adding one of these .33F caps at the R-Pi end of the cable from the 5v linear supply powering the R-Pi, now it keeps up with the switcher in dynamics while pushing well ahead in cleanliness. I have not yet tried the .33F cap on the output of one of these switching supplies, maybe that's the way to nirvana. That'll come later when I have time. I can recommend adding a .33F cap to the supply line for the Pi, BUT only with these caveats at this time:

a. Only do it with a supply that has an output regulator that has current limiting, you might burn up others

b. I haven't tried it with an SMPS yet, it may work or it may not AND possibly could damage the unit, try at your own risk.

AND I have not tried Soundcheck's preferred iFi units, so have no point of comparison with them... maybe they best all comers?

I am hopeful I can get the Boss sounding nearly as good or even possibly better than what I'm using now. AND I'm still very eager to hear the isolator boards from both Allo and Ian Canada... I suspect they will result in another lowering of the noise from the Pi and increase in SQ.

Following closely, thanks for all the great info so far!

Greg in Mississippi
 
Another set of my comments to put Soundcheck to sleep...
Just woke up. I even missed my Yoga sessions. Good to have u back. 😀


Obviously there's a recipe - common DIY sense - "How to design a DAC properly" to achieve more then just midfi qualities.

With any of the HAT projects I've done now. It's always been the same tweaks that lead to rather serious enhancements (in audiophile terms).

I think the Boss is a much better HAT-DAC compared to any other of them out there.

However. It also comes with quite some tweak potential.
What we discuss now was kind of expected from the very beginning.

* external supply
It's been confirmed by lukivision
I'd never buy that insane Uptone stuff. It's IMO just overkill.
An inmate sells the same LDO at a fragment. Then hook it up to a
LiFePO4 or whatever and you'll be set.
I've done that. I don't even think that a high quality LDO is the
best choice at the pre-reg position!
I also tried 4 different linear supplies. Incl. Sbooster
stuff. I still prefer my ifi iPower setup.
I do add some minor modifications to the iFi.
1. shorten the cable
2. add a large (>10000uF) low ESR buffer cap

Unfortunately the Allo Boss is not designed to make external feed
easy. ( I think I mentioned that before - wrong side, no jumper, bad connector choice - on one ear Allo seems to be deaf. 😀 )

* Kali
Adding the Kali can have several effects - beside the jitter focus.
Kali jitter will be higher then the onboard jitter btw. according to spec.
PS, EMI/RFI shield,... might do some good somwhere to eliminate the
potential jitter disadvantage.


* As usual - flimsy RCAs.
For those who're serious about audio - pass them by!
Hifiberry offers some extra holes to solder wire right to the board.
IMO the way to go. Neither would I hook up jumper wires to
the J20 header.


* Output filter.
Parts play a role here too. On the Mambo project I passed the resistor by.
And yes, resistors in the path mess around with the sound. (Greg - I even followed your advise to confirm that this is not what I want!)
I also stayed with SilverMica caps.



* DACs with on-dac opamps.
It's crucial to choose the right decoupling/buffer caps and LDOs.
That's why IMO boutique output stages can have that advantage.
Look at the HATs out there. Design by datasheet! What do you expect?
On my Mambo project I said something like: >70% of sound improvements
was contributed by the right choice of caps and LDOs (input from Subbu DAC project).
To put it into perspective. Kali contributions might be rated about 15%.
And than a supercap looks good on the first glance. It can deliver energy.
And for sure it's the perfect part to put into a marketing brochure.
But. Does it really sound good!?!? Is it needed?

* DAC SW filters/DSP
O dear. Here's where IMO that TI PCM51xx falls short.
Best to go the 352,8/384kHz route to pass the filters by.
Same with internal VC. Best to leave it at 100%.



Bottom line. Everything matters and nothing is perfect!

With a little tweaking the Boss is really a very nice device.
If DimDim ever has the guts to go that route he'll be surprised.



Have a good one.
 
Ah. One more.

PiCorePlayer 3.2 Beta has been released.

It comes with the Boss driver and up to 384k support!

Unfortunately they can't keep up the pace with Allo.
The kernel still seems to be 4.9.17 ( from an Allo driver perspective 4.9.20 would currently be the kernel of choice).
But, as far as I recall, they patched the Boss driver to a newer (but not the latest though) version. It'll do.
Keep the fingers crossed that the final PcP version comes with the latest kernel and driver!

Enjoy.
 
Last edited:
Thanks soundcheck for testing. pCP 3.20 uses the 4.9.16 patch set provided at Allo's github.. This is a complete driver set, so there should be no worries if we decide to stay at 4.9.17

Yes all wifi is disabled by default, but a user can easily install wifi and enable rpi3B and ZeroW wifi. We do disable rpi3 BT by default.
 
Hi folks.

Just to give u a heads up.

If you'd see below Boss related error message in the dmesg output of kernel 4.9.20:

Code:
[    5.831797] snd-allo-boss-dac soc:sound: ASoC: CODEC DAI pcm512x-hifi not registered - will retry
[    5.836090] snd-allo-boss-dac soc:sound: snd_soc_register_card() failed: -517
[    6.087757] 1-004d supply AVDD not found, using dummy regulator
[    6.089958] 1-004d supply DVDD not found, using dummy regulator
[    6.092024] 1-004d supply CPVDD not found, using dummy regulator
[    6.194833] snd-allo-boss-dac soc:sound: pcm512x-hifi <-> 3f203000.i2s mapping ok


I've been told by Allo "That's how it is".
Always good to see that this kind of stuff is known behavior.
Basically it can be ignored. The DAC seems to works fine.
 
Status
Not open for further replies.