Hi,
I finished the testing of the voltage regulators in a bad way
First of all my results to the ldovr 3045 is kind of similar to what was already reported: it seems like the 3045 has more definition and feels kind of more precise or sounds kind of faster (for a lack of better words)
But
The trade off is definitely imaging, the soundstage collapses a bit and all in all I would give the favor to the lf50 for most of the songs I listened to.
Now to the bad part: as I wanted to try some other voltage regulators 50cent to 45€ I must have made a mistake and fried something (probably when connecting one of the 1117 sparkos labs regulator which has a different output pin design but I couldn’t reconstruct my failure
Fact is I now have only about 6ohm resistance between ground and the lf50 out connection and when I connect the lf50 in a regular way it heats up and shuts off so I guess I fried the 74vhct244
Any tips on what to do now? Desoldering the chip seems not like an easy task but with 4decks swapping the mainboard isn’t done easy either as my soldering equipment and skills don’t reach into the smd and desoldering regions ^^
I still hope it’s just a resistor or a cap I fried but I’m kind of getting desperate here
TLDR: lf50 > 3045 and be careful mixing different regulator models
I finished the testing of the voltage regulators in a bad way
First of all my results to the ldovr 3045 is kind of similar to what was already reported: it seems like the 3045 has more definition and feels kind of more precise or sounds kind of faster (for a lack of better words)
But
The trade off is definitely imaging, the soundstage collapses a bit and all in all I would give the favor to the lf50 for most of the songs I listened to.
Now to the bad part: as I wanted to try some other voltage regulators 50cent to 45€ I must have made a mistake and fried something (probably when connecting one of the 1117 sparkos labs regulator which has a different output pin design but I couldn’t reconstruct my failure
Fact is I now have only about 6ohm resistance between ground and the lf50 out connection and when I connect the lf50 in a regular way it heats up and shuts off so I guess I fried the 74vhct244
Any tips on what to do now? Desoldering the chip seems not like an easy task but with 4decks swapping the mainboard isn’t done easy either as my soldering equipment and skills don’t reach into the smd and desoldering regions ^^
I still hope it’s just a resistor or a cap I fried but I’m kind of getting desperate here
TLDR: lf50 > 3045 and be careful mixing different regulator models
Soooo, pretty much the best case I could hope for: The short between ground and 5V is gone after desoldering the ground pin of the octal buffer. Thanks for the design choice not to put the 74vhct under the DAC boards Doede, so people like me, who are in the frying business can fix their mistakes more easily 
On the positive: I can test now how well things work without a line driver, now that i have to remove it anyway.
Anybody a tip where I could buy a single new 74vhct244a as i can only find B2B and bulk sellers :/
@Doede: Is there a possibilty you have some more of those chips lying around which i could buy from you? :>
greetings Oli
On the positive: I can test now how well things work without a line driver, now that i have to remove it anyway.
Anybody a tip where I could buy a single new 74vhct244a as i can only find B2B and bulk sellers :/
@Doede: Is there a possibilty you have some more of those chips lying around which i could buy from you? :>
greetings Oli
Soooo, pretty much the best case I could hope for: The short between ground and 5V is gone after desoldering the ground pin of the octal buffer. Thanks for the design choice not to put the 74vhct under the DAC boards Doede, so people like me, who are in the frying business can fix their mistakes more easily
On the positive: I can test now how well things work without a line driver, now that i have to remove it anyway.
Anybody a tip where I could buy a single new 74vhct244a as i can only find B2B and bulk sellers :/
@Doede: Is there a possibilty you have some more of those chips lying around which i could buy from you? :>
greetings Oli
Hi Oli, thanks for your experiments and sharing of the results, even when the end result is RIP
So, I am sure I will have some left over. There is just a new batch of mainboards underway to me, and there are always a few rest chips. By soldering the chip, make sure to use flux... check some MSD soldering tutorials in YouTube.
Send me PM, I will send you a chip of course for free....
Last edited:
Hello,
It shows again that not all LDO are state of the art. And it is a bit of a tricky business to just replace them. Some sellers like Belleson provide some extra information and will tell you how to do it and i managed to do it with help of Doede himself because i wanted to create some extra decoupling on the mainboard.
I saw this supply from Ian with some supercaps added. Of course these supercaps will give optimum results if they are as close as possible to the load. With some extra work that can be done. If some kid,cat or dog start pulling some wires they and the diy audio member will have some memorable experience. After testing put it in a sturdy enclosure!
Because i can make my own chassis at work that is easily said but i honestly think that wood is for furniture and birds not for electronic devices.
Wood and red copper is a nice combination as you can see in the picture . But all the electronic parts are mounted on the copper. There is one part for power supply and one part for the pre amp itself. The two copper plates are bend on all four sides and the wood it just there because it can handle some abuse and polished copper so easily damaged. The wood just joins the two copper plates and prevent stains on the copper during handling.
Greetings, Eduard
P.s Curious about the experiments that will happen in the weeks to come.....
It shows again that not all LDO are state of the art. And it is a bit of a tricky business to just replace them. Some sellers like Belleson provide some extra information and will tell you how to do it and i managed to do it with help of Doede himself because i wanted to create some extra decoupling on the mainboard.
I saw this supply from Ian with some supercaps added. Of course these supercaps will give optimum results if they are as close as possible to the load. With some extra work that can be done. If some kid,cat or dog start pulling some wires they and the diy audio member will have some memorable experience. After testing put it in a sturdy enclosure!
Because i can make my own chassis at work that is easily said but i honestly think that wood is for furniture and birds not for electronic devices.
Wood and red copper is a nice combination as you can see in the picture . But all the electronic parts are mounted on the copper. There is one part for power supply and one part for the pre amp itself. The two copper plates are bend on all four sides and the wood it just there because it can handle some abuse and polished copper so easily damaged. The wood just joins the two copper plates and prevent stains on the copper during handling.
Greetings, Eduard
P.s Curious about the experiments that will happen in the weeks to come.....
Attachments
A streamer to match the quality of DDDAC
Hi Doede and other entheusiast users of your stuff,
It has been a while since I was active in this thread, nice to see it is still going strong!
I have been experimenting with another dac and streamer for a while but finally went back to my trusted DDDAC, there is still nothing better in this regard.
I had a friend bring over a very expensive dCS dac/upsampler for a comparison. And what do you think? My DDDAC beat the dCS big time! More openness, more resolution, more dynamics, more spaciousness. The dCS was tame, flat and uninvolving in comparison.
I did some nice improvements on the streamer side that are worthwile to mention:
-I am using an Allo usbridge sig instead of the rpi (less noise) it can run all rpi software, I use moode currently.
-fifopi on battery supply
-Andrea Mori clock unit on battery supply (big improvement on all audible aspects!)
-battery supply with added supercaps (this makes the noise floor of the already quiet battery even lower, this is clearly audible)
I have been using the rpi as a streamer for a long while, first with the Allo isolator and kali reclocker. Changing to the fifopi made an improvement and changing to the usbridge sig made another significant improvement. This stuff is highly reccommended for all DDDAC users, you will not be dissapointed!
The improvements on the streamer part did make a big improvement in the overall sound of my set, there is certainly a lot to gain here.
Regards,
Hi Doede and other entheusiast users of your stuff,
It has been a while since I was active in this thread, nice to see it is still going strong!
I have been experimenting with another dac and streamer for a while but finally went back to my trusted DDDAC, there is still nothing better in this regard.
I had a friend bring over a very expensive dCS dac/upsampler for a comparison. And what do you think? My DDDAC beat the dCS big time! More openness, more resolution, more dynamics, more spaciousness. The dCS was tame, flat and uninvolving in comparison.
I did some nice improvements on the streamer side that are worthwile to mention:
-I am using an Allo usbridge sig instead of the rpi (less noise) it can run all rpi software, I use moode currently.
-fifopi on battery supply
-Andrea Mori clock unit on battery supply (big improvement on all audible aspects!)
-battery supply with added supercaps (this makes the noise floor of the already quiet battery even lower, this is clearly audible)
I have been using the rpi as a streamer for a long while, first with the Allo isolator and kali reclocker. Changing to the fifopi made an improvement and changing to the usbridge sig made another significant improvement. This stuff is highly reccommended for all DDDAC users, you will not be dissapointed!
The improvements on the streamer part did make a big improvement in the overall sound of my set, there is certainly a lot to gain here.
Regards,
Attachments
Hello,
So nice to see the resurrection of Stefan ( supersurfer). In the past he used to share nice ideas here about how to improve the dddac. He installed the Tent shunt real diy style and also messed around with chokes.
A week ago i send him a private message to tell him about the rising interest in the dddac thread regarding optimizing '' the source equipment ''
It seems that during his absence he gathered lots of experience about it.
Greetings, Eduard
So nice to see the resurrection of Stefan ( supersurfer). In the past he used to share nice ideas here about how to improve the dddac. He installed the Tent shunt real diy style and also messed around with chokes.
A week ago i send him a private message to tell him about the rising interest in the dddac thread regarding optimizing '' the source equipment ''
It seems that during his absence he gathered lots of experience about it.
Greetings, Eduard
Quick update, perhaps someone can make use of this ^^: I removed the 74vhc244 and connected the data, bclck and lrclock directly. And it seems to work perfectly fine, only problem is my Kali stopped working for whatever reason 
, so I had to connect the rpi outputs and now I can’t say anything about quality degradation/improvement (i2s connection is also pretty rudimentary at the moment, running through breadboards and some quick and dirty soldering etc...)
On the positive: I can order Ian’s fifo now bc I’ll need a new reclocker anyway
When I got the fifopi running I’ll will listen to it for a while with proper 74’less implementation, and after that I’ll put the 74vhc244 back in once i get a new one from doede (thanks again on this point ) Then I should be able to say if it’ll stay or if there is an improvement by not having the 74.. in there.
Greetings Oli
, so I had to connect the rpi outputs and now I can’t say anything about quality degradation/improvement (i2s connection is also pretty rudimentary at the moment, running through breadboards and some quick and dirty soldering etc...)On the positive: I can order Ian’s fifo now bc I’ll need a new reclocker anyway
When I got the fifopi running I’ll will listen to it for a while with proper 74’less implementation, and after that I’ll put the 74vhc244 back in once i get a new one from doede (thanks again on this point
) Then I should be able to say if it’ll stay or if there is an improvement by not having the 74.. in there.Greetings Oli
The shift registers after the one i replaced are, ...164mx something, but the first one: ....244 is there to switch between spdif and USB and acts as a line driver, so no change in time here i think.
Edit:
The 1794 operates in nos mode the way it is connected. The problem which doede solved with the shift registers is that that it needs right justified i2s and has no channel select anymore (I still try to wrap my head around the missing channel select my first thought is that that one would have to invert the word clock but that doesn't seem true. aaanyway no need to worry as whatever magic is happening the dddac plays in stereo
Edit:
The 1794 operates in nos mode the way it is connected. The problem which doede solved with the shift registers is that that it needs right justified i2s and has no channel select anymore (I still try to wrap my head around the missing channel select
my first thought is that that one would have to invert the word clock but that doesn't seem true. aaanyway no need to worry as whatever magic is happening the dddac plays in stereo
Last edited:
Ah I just understood with the help of http://www.dddac.com/documents/dddac1794s_nos_ver42.pdf , the data is shifted 32bits so it's shifted a whole "word". Lots to learn again, gotta love DIY and this community
I use picoreplayer on one of my rpi streamers but did not use it on the usbridge.
Moode is working perfectly though.
I have read that when using usbridge with I2S output it will work with every rpi distro, only if you want to use usb output on the usbridge a certain driver is needed.
Did you try the andrea mori clock against the accusilicon? Is all info on them in the "the well tempered master clock" thread?
Thanks and greetings Oli
Last edited:
I will probably compare them next saturday. I think however that the accusilicon will not stand up to the well tempered because the last one has a very good driscoll style squarer with good separated power supplies.
Lets wait and see.....
It is a diy thing from this forum: The Well Tempered Master Clock - Building a low phase noise/jitter crystal oscillator