Hifimediy and Nick,
Whatever happens with this board, I hope you keep experimenting with the DDX digital chips. In the near-future, almost all music and video will be PC-based and a lot of us would like to see a good digital amp. Analog is so twentieth-century.
One thought: you're spent a lot of time working on the hardware interface, but it sounds like the chip is actually software-based. I don't know what the software support for DDX is like, but a USB interface (like MiniDSP) might be a useful addition. Maybe even a critical one.
Whatever happens with this board, I hope you keep experimenting with the DDX digital chips. In the near-future, almost all music and video will be PC-based and a lot of us would like to see a good digital amp. Analog is so twentieth-century.
One thought: you're spent a lot of time working on the hardware interface, but it sounds like the chip is actually software-based. I don't know what the software support for DDX is like, but a USB interface (like MiniDSP) might be a useful addition. Maybe even a critical one.
I've encountered a similar problem in HW.
It sounds a bit like a floating port (i.e one that is not terminated high or low) We had a similar issue when upgrading from a pin-compatible 1Mbit to 2Mbit Ram chip on one of our products.
One of the pins that was NC on the 1Mbit chip was "address bank selector" on the 2Mbit version. As the HW design was for the former, the gate was left "in the air" this caused the DDR to write/read randomly between bank 1 and 2. BAD chip!
However: as SoldierAnt says; the STA320 seems to be SW based, but make sure there are no HW selector for high/low bitrate..
Also: a USB-interface to set custom EQ biquads, enable/disable "night mode", etc would be very much appreciated!
It sounds a bit like a floating port (i.e one that is not terminated high or low) We had a similar issue when upgrading from a pin-compatible 1Mbit to 2Mbit Ram chip on one of our products.
One of the pins that was NC on the 1Mbit chip was "address bank selector" on the 2Mbit version. As the HW design was for the former, the gate was left "in the air" this caused the DDR to write/read randomly between bank 1 and 2. BAD chip!
However: as SoldierAnt says; the STA320 seems to be SW based, but make sure there are no HW selector for high/low bitrate..
Also: a USB-interface to set custom EQ biquads, enable/disable "night mode", etc would be very much appreciated!
Nakamoomin's suggestion has the ring of plausibility; hopefully the problem is hardware-related, and easily fixed.
Even so, it's worth pondering what would happen if a software bug ever surfaced. How would any of us diagnose the problem, or update/modify the firmware? Swapping out resistors or capacitors is relatively easy, but a problem with the code would be difficult for users to troubleshoot, much less fix.
So a USB port and software interface ought to be standard-equipment on a product such as this one. Besides, if people had the ability to customize their board through software, they wouldn't need to pester you so much to keep making new versions.
It lives...
and makes music!
Initial impression:
- Strength: Sounstage and imaging,PRAT, midrange clarity
- Weakness: treble sharpness. Perhaps I need to play some more?
Excellent build quality
Comparison system: Audiosector USB DAC -> B1 preamp -> Modded sure amp (Dayton input & output caps, air core coils), not as good PRAT but smoother.
Speakers are Pispeakers model 2 (Suits me better than RS Natalie P.)
Thinking about replacing the coil with Wurth. I think the stock coils are 11uH, which Scott Endler recommended replacing them with 4.7uH 9.3A Wurth.
At first I had some trouble getting it to play. Things to watch out for:
- Main connection to the Connex SMPS is GND - AC - AC
- Out from SMPS is GND - POS into GND - V2 in the DDX320
- DIP switch set to ON-ON-ON-ON (pass thru). The default setting is OFF-OFF-OFF-OFF which is 320Hz High pass. Make the sound tiny and stuck to the speakers
and makes music!
Initial impression:
- Strength: Sounstage and imaging,PRAT, midrange clarity
- Weakness: treble sharpness. Perhaps I need to play some more?
Excellent build quality
Comparison system: Audiosector USB DAC -> B1 preamp -> Modded sure amp (Dayton input & output caps, air core coils), not as good PRAT but smoother.
Speakers are Pispeakers model 2 (Suits me better than RS Natalie P.)
Thinking about replacing the coil with Wurth. I think the stock coils are 11uH, which Scott Endler recommended replacing them with 4.7uH 9.3A Wurth.
At first I had some trouble getting it to play. Things to watch out for:
- Main connection to the Connex SMPS is GND - AC - AC
- Out from SMPS is GND - POS into GND - V2 in the DDX320
- DIP switch set to ON-ON-ON-ON (pass thru). The default setting is OFF-OFF-OFF-OFF which is 320Hz High pass. Make the sound tiny and stuck to the speakers
Attachments
I used the Connexelectronics power supply but I think you can use a 20-0-20 VAC transformer rated about 300VA without any problem.
Duc
If you ask me. Full digital amps have the capacity to blow away standard DAC-amp setups at a fragment of cost.
I do own a TAS5706 amp since almost two years and wouldn't swap it against my TP Sabre+amp anymore.
I'd more than understand why the audio industry is not interested to go the full digital route.
There'd be a high risk that you'd kill large parts of the DAC market as well as amp segment as well as analog IC cable segment.
(Of course that would apply also to your own portfolio.)
Last Sunday I decided to order one of your products. Unfortunately I'd been late. All modules were gone. ( I had PMed you guys)
Some feedback:
(A friend of mine gotta DDX320. His findings match my experiences with my TAS5706)
1. What's needed is a highest quality digital transport, since the input stage
is not able to recover rather average quality input signals at highest quality.
The same rules apply as for top quality DAC input/recovery/reclock stages.
2. Your amp seem to lack a high quality spdif pulsetransformer on the input. That one is needed to keep the noise out and ususally makes a hell of a difference.
3. As usual and being discussed a hundred million times: A top PS is key to success.
Full digital amps are IMO the best kept secret in the audio world. And I'm sure many manufactures want to leave it this way.
Please put some more effort on getting your problems fixed and you might want to have a look at improving your input stage.
Rev 2 of your product could be a real winner.
Put me on the waiting list. Good luck.
Cheers
Last edited:
Hi folks.
I just installed a DDX320. Got one ( or better a friend of mine) from the last 45$ batch.
After a couple of hours my first impressions are pretty positive.
Out of the box it sounds pretty good. The sound signature is pretty similar
to my TAS5706. The greater muscles of the DDX make the amp more
dynamic even on my 98db/SPL speakers.
Compression at higher levels disappeared. Great stuff.
Right now It's breaking in. It still seems to improve from hour to hour.
I'm sure the DDX got greater potential then my 5706 installation. It'll stay.
(The pity:I'd need one more amp for my woofers - hopefully we'll see a new batch coming out soom)
I'm wondering now what power supply to go for?!?!? Right now I'm using a decent switching single voltage 24V 6.5A device. Nothing spectecular.
What should I Iook at when selecting the best possible PS for the DDX:
SMPS or linear PS
Single voltage or double voltage
Power ratings ( bigger = better ??).
What brand!?!? Many people seem to rave about Connex SMP500R 24 V ( single or double voltage?) Are there alternatives?
Please let me know any other tweak potential.
Thx.
Cheers
I just installed a DDX320. Got one ( or better a friend of mine) from the last 45$ batch.
After a couple of hours my first impressions are pretty positive.
Out of the box it sounds pretty good. The sound signature is pretty similar
to my TAS5706. The greater muscles of the DDX make the amp more
dynamic even on my 98db/SPL speakers.
Compression at higher levels disappeared. Great stuff.
Right now It's breaking in. It still seems to improve from hour to hour.
I'm sure the DDX got greater potential then my 5706 installation. It'll stay.
(The pity:I'd need one more amp for my woofers - hopefully we'll see a new batch coming out soom)
I'm wondering now what power supply to go for?!?!? Right now I'm using a decent switching single voltage 24V 6.5A device. Nothing spectecular.
What should I Iook at when selecting the best possible PS for the DDX:
SMPS or linear PS
Single voltage or double voltage
Power ratings ( bigger = better ??).
What brand!?!? Many people seem to rave about Connex SMP500R 24 V ( single or double voltage?) Are there alternatives?
Please let me know any other tweak potential.
Thx.
Cheers
Last edited:
after being positive in the early hours, i have to report that i'm experiencing a pretty high noisefloor on this amp.
and i'm not talking about the general noise flaw here. that one i experienced too.
i tried two different power supplies. the better supply made the high noisefloor even more obvious.
i've been also running that amp on 24Vbatteries. same thing.
i also tried toslink vs. spdif. same game.
it starts with playback start. it seems to be related to the processing. if playback is stopped it's dead quiet.
it's on any samplerate i tried.
i'd appreciate any hints what to do next - maybe i've done something wrong.
though you can't do much wrong here -- i guess.
my tas5706 still plays dead quiet in comparison.
cheers
and i'm not talking about the general noise flaw here. that one i experienced too.
i tried two different power supplies. the better supply made the high noisefloor even more obvious.
i've been also running that amp on 24Vbatteries. same thing.
i also tried toslink vs. spdif. same game.
it starts with playback start. it seems to be related to the processing. if playback is stopped it's dead quiet.
it's on any samplerate i tried.
i'd appreciate any hints what to do next - maybe i've done something wrong.
though you can't do much wrong here -- i guess.
my tas5706 still plays dead quiet in comparison.
cheers
Last edited:
Hi guys.
I'll do some surgery on the board.
1. First of all. Is there a layout of that board available to the public??
2. On a first visual examination I noticed two empty pads.
R3 is missing (close to the Torx).
C70 is missing (close to the STA320).
If HifiMeDiy is still following this thread, please let me know if that is an issue.
3. What I'll do next
a. getting rid of the relais
b. and plugging my Wuerth coils in. I'll put them underneath the board.
Somehow I don't want to have those coils sitting right next to the
surrounding resistors, elcos etc.
c. let see
Cheers
I'll do some surgery on the board.
1. First of all. Is there a layout of that board available to the public??
2. On a first visual examination I noticed two empty pads.
R3 is missing (close to the Torx).
C70 is missing (close to the STA320).
If HifiMeDiy is still following this thread, please let me know if that is an issue.
3. What I'll do next
a. getting rid of the relais
b. and plugging my Wuerth coils in. I'll put them underneath the board.
Somehow I don't want to have those coils sitting right next to the
surrounding resistors, elcos etc.
c. let see
Cheers
Last edited:
Thx Hifimediy. I guessed that.
Progress Report:
Stage 1:
I applied some tweaks. Relays removal, Würth coils ( mounted underneath the board), better chip decoupling.
Folks. 100% success. The sound was really nice at this stage. The noise went down
considerably.
I've been enjoying that modification level for exactly one day...
...when I decided to apply further mods that I still had in mind and were recommended by a friend of mine..
Stage 2:
1. replaced the 3.3.V with a low noise type -- with a temporary separate power-on switch.
2. removed the woofer relay and coils -- and left the outputs open!?!?!?! ( open output -- problem??)
3. Some more decoupling
THAT ONE FAILED!!!!!!!!!!!!!!! Noooo, I can't believe it.
After applying those final little tweaks everything seemed to come up again -- all voltages were there, LEDs on, PIC seemed to work...
...BUT ---- NO SOUND.
Folks that's driving me crazy. Two days ago I had a great performing amp in my
hands and 10 hours later I screwed it. Damned.
I've been looking for almost a day after the problem. All voltages seems to be OK. I also
put everything the way it was the last time it worked. Still no sound.
I'm running out of ideas.
I guess without a scope it'll be difficult to trace the issue down.
PLEASE: If anybody has an idea of what can go wrong or what and how to verify.
Please let me know.
@HIFIMEDIY: How far are you with your redesign/new batch of that amp.
I'd reorder immediately a new one. In case there'd be no progress on your side: What would be the minimum number of units that would justify producing a small
new batch of the old design.
Just some more comments:
Improvement proposals for rev 2 of the board, that you might consider:
(I'm not a real HW expert. So please apologize and/or correct me if I'm wrong or misinterpret things.)
While scanning the design for quite some hours now, some potential areas of improvement I'd like to bring up over here:
1. All digital and analog 3.3V pins sit on one rail. No ferrite beds or similar - just a little
decoupling.
We're talking about the STM controller, the SPDIF receiver, the clock, the PIC, the
control unit. With just a little extra effort this situation might be improved.
Though you're the experts.
(However . It is IMO still amazing how this unit performs -- after the tweaks.)
2. The PIC with its internal clock sits right beside the one and only 3.3V regulator.
I'd put some space in between and I'd also hook it up to a nicely isolated supply.
3. I don't think that the output stage is well done. All those coils and relays in close
interaction with buffer caps for the main amp power rail, etc. Design should be
more careful here. I do understand the space constraints due to the chosen
heatsink. Still, you shouldn't place those sensitive parts in such a noisy high-power
HF environment.
4. I'd love to get all the AC related stuff removed from the board. I'm sure a single pole
24 DC voltage supply would be widely accepted and maybe even preferred.
Production cost can be lowered resp. those budgets can be put on better 3.3V supply.
5. It would be nice for us tweakers to have mount points for external DC supplies, that
would replace onboard supplies easily. (Twisted Pear audio is doing this very nicely on the Buffalo DACs)
THX
Cheers
Progress Report:
Stage 1:
I applied some tweaks. Relays removal, Würth coils ( mounted underneath the board), better chip decoupling.
Folks. 100% success. The sound was really nice at this stage. The noise went down
considerably.
I've been enjoying that modification level for exactly one day...
...when I decided to apply further mods that I still had in mind and were recommended by a friend of mine..
Stage 2:
1. replaced the 3.3.V with a low noise type -- with a temporary separate power-on switch.
2. removed the woofer relay and coils -- and left the outputs open!?!?!?! ( open output -- problem??)
3. Some more decoupling
THAT ONE FAILED!!!!!!!!!!!!!!! Noooo, I can't believe it.
After applying those final little tweaks everything seemed to come up again -- all voltages were there, LEDs on, PIC seemed to work...
...BUT ---- NO SOUND.
Folks that's driving me crazy. Two days ago I had a great performing amp in my
hands and 10 hours later I screwed it. Damned.
I've been looking for almost a day after the problem. All voltages seems to be OK. I also
put everything the way it was the last time it worked. Still no sound.
I'm running out of ideas.
I guess without a scope it'll be difficult to trace the issue down.
PLEASE: If anybody has an idea of what can go wrong or what and how to verify.
Please let me know.
@HIFIMEDIY: How far are you with your redesign/new batch of that amp.
I'd reorder immediately a new one. In case there'd be no progress on your side: What would be the minimum number of units that would justify producing a small
new batch of the old design.
Just some more comments:
Improvement proposals for rev 2 of the board, that you might consider:
(I'm not a real HW expert. So please apologize and/or correct me if I'm wrong or misinterpret things.)
While scanning the design for quite some hours now, some potential areas of improvement I'd like to bring up over here:
1. All digital and analog 3.3V pins sit on one rail. No ferrite beds or similar - just a little
decoupling.
We're talking about the STM controller, the SPDIF receiver, the clock, the PIC, the
control unit. With just a little extra effort this situation might be improved.
Though you're the experts.
(However . It is IMO still amazing how this unit performs -- after the tweaks.)
2. The PIC with its internal clock sits right beside the one and only 3.3V regulator.
I'd put some space in between and I'd also hook it up to a nicely isolated supply.
3. I don't think that the output stage is well done. All those coils and relays in close
interaction with buffer caps for the main amp power rail, etc. Design should be
more careful here. I do understand the space constraints due to the chosen
heatsink. Still, you shouldn't place those sensitive parts in such a noisy high-power
HF environment.
4. I'd love to get all the AC related stuff removed from the board. I'm sure a single pole
24 DC voltage supply would be widely accepted and maybe even preferred.
Production cost can be lowered resp. those budgets can be put on better 3.3V supply.
5. It would be nice for us tweakers to have mount points for external DC supplies, that
would replace onboard supplies easily. (Twisted Pear audio is doing this very nicely on the Buffalo DACs)
THX
Cheers
Hi folks.
I don't give up on it. My device is still not singing.
I measured some pins at the STA320 while data are delivered from the source. It seems to stay quiet.
DDX1A/B = 0V ( seem to be kept low)
DDX2A/B = 0V
RESET= 3.3V
EAPD=0,9
TRI_STAE=3,3
OCDETECT=3,3
FAULT=3,3
TWARN=3,3
Can anybody confirm above values to be OK??? What would I have to check next??
BTW: The 5V reg, is it only used as 3,3V pre-regulator. If yes. Could it be removed in case I'm using an external 3.3V supply??
Thx for your support.
Cheers
I don't give up on it. My device is still not singing.
I measured some pins at the STA320 while data are delivered from the source. It seems to stay quiet.
DDX1A/B = 0V ( seem to be kept low)
DDX2A/B = 0V
RESET= 3.3V
EAPD=0,9
TRI_STAE=3,3
OCDETECT=3,3
FAULT=3,3
TWARN=3,3
Can anybody confirm above values to be OK??? What would I have to check next??
BTW: The 5V reg, is it only used as 3,3V pre-regulator. If yes. Could it be removed in case I'm using an external 3.3V supply??
Thx for your support.
Cheers
Hi,
RESET= 3.3V
EAPD=0,9
TRI_STAE=3,3
OCDETECT=3,3
FAULT=3,3
TWARN=3,3
These levels seem ok.
And 5V is not used for other device except a LED (for power).
DDX1A/B = 0V
DDX2A/B = 0V
Means no audio signal output from sta320, so may be you have to check the I2S signal from CS8416 to sta320 first.
RESET= 3.3V
EAPD=0,9
TRI_STAE=3,3
OCDETECT=3,3
FAULT=3,3
TWARN=3,3
These levels seem ok.
And 5V is not used for other device except a LED (for power).
DDX1A/B = 0V
DDX2A/B = 0V
Means no audio signal output from sta320, so may be you have to check the I2S signal from CS8416 to sta320 first.
THX for the feedback. Step by step I'm hopefully getting there.
I'll try first to check the I2S bus with a voltmeter. Should be sufficiant to see if anything is ongoing.
I'll get a scope next week. That'll be much better.
Cheers
Hi there.
Just a stupid question -- as I said I do not have any background in HW design.
Why is the thermal pad of the 3,3V regulator connected to the positive 3.3.V rail on
the board?
I'd expected a ground connection.
Anyhow. I had my friend check his setting. It's the same over there. It doesn't seem to be a problem at least.
Cheers
Just a stupid question -- as I said I do not have any background in HW design.
Why is the thermal pad of the 3,3V regulator connected to the positive 3.3.V rail on
the board?
I'd expected a ground connection.
Anyhow. I had my friend check his setting. It's the same over there. It doesn't seem to be a problem at least.
Cheers
Last edited:
A multimeter is useless for digital signal I think.
It's normal, it's 1117-3.3 regulator has output on the thermal pad, just like LM317.
A multimeter is useless for digital signal I think.
No. I don't think it's useless. I can clearly see if and when the playback is started. The data line level goes up on the voltage as soon as I press play.
Obviously I can't verify what kind of data in what shape are delivered.
My scope just arrived. My troubleshooting options should slightly improve
soon.
Cheers
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.