Thanks @sa3999 You are correct I have the HDMIPiPro II. Sorry for the short hand and not correctly stating the product. I have been looking at the documentation. I see the switches to flip to on, but it looks like there is a cover on the switches which make it impossible to move the switches. It least that is what I see. It is probably something simple that I don't know about. Again thanks for your help
Dave
Dave
I m using a quad TDA1541A in fact. Based on the very nice design of @ryanj 3D pcb design.
And it is the best dac I have ever built and heard. I have had some quite expensive high-end dacs next to it, with the most expensive being a dCS Vivaldi with master clock unit. The performance of the dCS was quite shameful next to my old Philips dac chips. It was no comparison, the owner of the Vivaldi said within 20sec after start playing that the difference is ridiculously big and I agreed. The focus, naturalness and lifelike experience was much bigger with my diy dac.
Attachments
@dcallende
Dave, removing and dumping the yellowish film from the DIP switch block takes a second at most. You just need something sharp or pointy to get under it's edge to pry it off. I used the blade of utility knife. Can't be that difficult.
The sticky film on the switch block is to ensure picking from the carrier tape in the feeder and holding by the nozzle in the pick&place machine in SMT assembly process.
Last edited:
Lso,
I have used Ian's OPA861 I/V in various configurations since Ian started development of it before he released it publicly.
His original developmental board came with an on-board TL431-based shunt regulator. It sounded just ok no matter how I powered it.
Later I bypassed the onboard regulator & tried some alternatives. Some made it sound better, but still just better 'ok'.
I tried +-3.3V from Ian's larger LiFePO4 Battery Supply board & the SQ jumped to 'very good'! I recommend using either that board or a couple of Ian's single 3.3V LiFePO4 battery boards as a good baseline power setup. In that configuration it is as good or perhaps better than any other I/V stages I've tried with Ian's DAC boards.
I am currently awaiting the components to power Ian's OPA861 I/V with +-5V from 2 UCPures. I will report on that after I have it running.
Greg in Mississippi
Instead of removing the thin Kapton tape, I just use a bulletin-board push-pin to poke through the tape & move the switch. Doing that leaves some amount of protection over the switch.
Greg in Mississippi
That is good to hear.
My R26 is lent out to a friend, but I expect it back in the next few weeks.
I fed it from either I2S-over-HDMI from an SDTrans384 SD Card player or USB from a Sonore Ultra-Rendu. I was surprised that the Ultra-Rendu sounded overall better than the SDTrans384 (each fed from very good power supplies), but realized the SDTrans384 uses over 10 years old LVDS chips to fed the I2S-over-HDMI output & they will have higher jitter than the best recent chips. AND that the Ultra-Rendu designer, John Swenson, knows how to do a high-quality USB output right.
OTOH, I expect to have an Ian HDMIPro II on hand about the time when the R26 returns & I am very curious how the R26 will sound fed from that, though it doesn't have the ability to turn off the I2S-over-HDMI incoming PLL like Ian's Holo Spring, so I doubt it will equal what Ian heard through his Holo Spring.
When I lent the R26 out I was comparing it to an Ian stack with a FiFiPi Q3, ReclockPi, & NDK SDA clocks. I overall preferred the Ian stack though the R26 did have some charms.
By the time it returns I will have updated that stack to the FiFoPi Q7II with SCPure clocks. A rematch against the R26 fed from an HDMIProII will be interesting.
Greg in Mississippi
Spacejack,
I questioned my sanity too when I first heard the warmup improvement even though I had read about it. Hearing that (along with many other 'ah-ha' moments) helped me learn to listen critically, verify, & trust my ears even if common audio lore said different. & also trust my ears when something that some others say should make a difference doesn't to my ears & in my setups.
Learning to do that has helped me take them to the levels they are at now & much above what I ever expected.
Greg in Mississippi
So, having a 'high end' DIY streamer, based on StationPi Pro, FifoPi Q7 and TransportPi AES, what could be an equally good DAC HAT or board that could complement the whole thing? I know there is ProtoDAC, which has received several positive reviews. But I wonder whether there is anything else, too. I'd prefer a NOS R2R rather than FPGA-chip-based DAC.
Thanks in advance for any recommendations.
Thanks in advance for any recommendations.
Thanks for the feedback, can you drive the opa861 with +/- 3,3 vdc? I have a 2 LiFEPO4 mini 3,3vdc in spare
+-3.3V works very well, sounds very good, & the volume drop from +-5VDC was unnoticeable to me.
Just be sure to set it all up so they turn on simultaneously, I had to replace some OPA861s after a non-simultaneous power-up.
Using Ian's larger LiFePO4 board ensures that, but as I remember you can chain multiple LiFEPO4 Mini 3.3VDC together.
I looked at using a single LiFePO4 Mini 6.6VDC for this with a center-tap, but the circuit does not work for this, the +- between the cells is not at 0V when turned off.
Greg in Mississippi
Just be sure to set it all up so they turn on simultaneously, I had to replace some OPA861s after a non-simultaneous power-up.
Using Ian's larger LiFePO4 board ensures that, but as I remember you can chain multiple LiFEPO4 Mini 3.3VDC together.
I looked at using a single LiFePO4 Mini 6.6VDC for this with a center-tap, but the circuit does not work for this, the +- between the cells is not at 0V when turned off.
Greg in Mississippi
I have seen a similar phenomenon with the overvoltage led, but not the full led. The overvoltage led started very dim and progressed to brighter as the capacitors were charged. I rectified the overvoltage problem by discharging the capacitors carefully and reflowing the solder on their pins, then when i charged the UcConditioner i checked with voltmeter that both charged similarly.
@Greg Stewart I think that turning off the PLL on the Holo DAC's effectively stops the DAC from re-clocking the incoming signal from the streamer and the DAC then uses the clocking information provided by the I2S signal stream (I may be wrong about this as I have never used a Holo DAC - but that is how I understand it).
I did ask Gustard technical support whether the R26 always re-clocks incoming digital streams or whether this could be switched off using the I2S input on the R26. The response from Gustard technical support was as follows:
"When nos is enabled, the IIS clock will be used in pcm"
Therefore, if you have NOS enabled in PCM mode on the R26. then the R26 will not re-clock the incoming I2S signal. Effectively this achieves the same as switching off the PLL on the Holo Spring DAC (as Ian did).
Therefore, with the R26 on NOS mode, you get all of the benefit of using Ian's SCPure clocks on your FIFOPiQ7 (as I have also done).
Whether the R26 sounds as good as an Holo Spring.....whole different conversation and not for this board !
Rgds
P.S. Just as an interesting point, as I understand it, the Denafrips DAC's always reclock the incoming signal. Just one more aspect of the minefield out there when selecting a DAC !
@pkonstantinidis - I would also be extremely interested in any recommendations for a really top quality DAC HAT/Board
I've had the Gustard R26 DAC for over a year now and have been very happy and impressed by its performance. It just seems to get better and better as the quality of the input signal improves. However, I am now subject to a further infection of the upgrade disease and looking for a significant improvement for the DAC. Ideally, as a relatively simple DIY project.
I too am after an R2R DAC - but from my perspective I see a number of potential "compatability" issues in terms of a DIY DAC to work with Ian's Pi / FIFOPi / SCPure / HMDIPi steamer stack - items such as:
- Ensuring the DAC accepts I2S
- Ensuring the DAC doesn't reclock the I2S stream
- Ensuring the DAC accepts up to 768KHz upsampled signal
Rgds
@sa3999 I am also curious whether the TDA1541A does any upsampling/downsampling of the streamed input.
I understand that @Gabster 2000 is working on a DAC based on the TDA1541A and he has already reported great results in terms of sound quality. Hoping that he will be kind enough to share the technical details about that project. I would be interested in replicating Gabster's DAC project (perhaps slightly adjusted in terms of power supplies).
I understand that @Gabster 2000 is working on a DAC based on the TDA1541A and he has already reported great results in terms of sound quality. Hoping that he will be kind enough to share the technical details about that project. I would be interested in replicating Gabster's DAC project (perhaps slightly adjusted in terms of power supplies).
Read the manuals and get something built and working I suggest.......its all about the music no?
PCM channels are not switched for RPI, only DSD, so if you switch on a client you will end up in the same spot (just reversed situation)...
It appears to me that the issue is more related to medication, or the absence thereof.
How can one truly listen when nothing ever seems to function as it should? And, of course, it's all too easy to attribute the blame to someone else's actions or inactions.
The TDA1541A chip does no up sampling. It plays what it is provided up to it's limits. It was designed to be run oversampled. I run it NOS at 44.1 but many prefer 4x up sampled.
The definitive analysis on SOTA implementation is a thread by ecdesigns. https://www.diyaudio.com/community/threads/building-the-ultimate-nos-dac-using-tda1541a.79452/
@ryanj did a beautiful job designing a 4 layer pcb effectively implementing ecdesigns findings. His group buy is here. https://www.diyaudio.com/community/threads/diy-tda1541a-pcb-d3.328060/page-65#post-7545958 It is not an easy DIY soldering task. If you want to do a very good, but easy to solder job, find a sample of his first effort, Distinction D1.
Gabster is having fun, but at this stage his implementation is far from optimized. EG, ground planes, trace line positioning and positioning decoupling caps tight to the 14 pins is critical to good sound. Most of the old CDPs did a very poor job of implementing them.
@iancanada I set up the attached streamer with purepi --> rpi4 --> shieldpi --> Q3 (NDK clocks) --> reclockpi --> HDMIpro. Will isolatorpi add any benefit to this streaming stack when feeding a receiverpiDDC --> q7 (SC pure clocks) --> 9038dac hat? If so, where should it be installed? Right below HDMPIpro or lower down, above shieldpi?
Thank you.
Thank you.
