classd4sure said:
You too have Emu? 1212M has no audiodock, its DAC is a good Cirrus but is internal to PC. The external one is a WOLFSON 8740 EV board (it's used in Arcam DACs) and has two dedicated PS I built myself (+5V for chip and +/-12V for opamps) and differential outputs. It shares with PC only safety ground, as it's connected to Emu through Toslink cable.
Maybe opamps specs have the signal level in mV? (Or maybe I better move to digital source threads and search
)Thanks
Thomas
thomaseliot said:
The concept sounds good, just have to remember about the DC when you start shifting things around to different systems, or maybe put a warning note as a reminder. If your DAC design only has some DC drift, you could also consider keeping the caps in the amps and removing the ones in the DAC.
Not really sure what caps are going to be best since small signal caps and those used in passive XOs have different considerations.
I made some modifications to the outputfilter of AMP1 (same chip as AMP5, and the results are astonishing.
If you promise to test these modifications on your AMP5 I can send them to you; they are not expensive at all, they only require some cheap parts and some finger labour (copper bending). I'd like to know how the tripath compares to the ucd's with these modifications.
Please let me know.
If you promise to test these modifications on your AMP5 I can send them to you; they are not expensive at all, they only require some cheap parts and some finger labour (copper bending). I'd like to know how the tripath compares to the ucd's with these modifications.
Please let me know.
Hi Thomas,
As for coupling capacitors, in my uneducated opinion any good film capacitor will do. I once read a review of a high-end capacitor adding a 'lustre' that wasn't present when the capacitor was left out (which was possible in this particular case). This 'lustre' is not what I would go for. But if another test reveals otherwise my opinion could change. For my AMP1's I use 4,7uF Wima's, which give a pole at about 2 Hz; this may seem nonsense, but this is usually not the only high-pass filter in the chain.
As for the output filter, my modification is something that is difficult to convey in writing. If you give me a postal adres, I will send it to you, with an explanation of why and how.
Bert
As for coupling capacitors, in my uneducated opinion any good film capacitor will do. I once read a review of a high-end capacitor adding a 'lustre' that wasn't present when the capacitor was left out (which was possible in this particular case). This 'lustre' is not what I would go for. But if another test reveals otherwise my opinion could change. For my AMP1's I use 4,7uF Wima's, which give a pole at about 2 Hz; this may seem nonsense, but this is usually not the only high-pass filter in the chain.
As for the output filter, my modification is something that is difficult to convey in writing. If you give me a postal adres, I will send it to you, with an explanation of why and how.
Bert
bertus said:
Hi bertus,
thanks. My email is thomaseliot@libero.it
Looking forward for your mods.
Ciao
Thomas
Hi Berthus,
I agree that any good film cap should do although the big question is how do we define "good".
The WIMA MKS2 that you seem to be using is one of three I have noted people recommend (others being the Vishay SAL 123 RPM series and the Auricap), so I know these three are "good" although I am not sure is only the Auricap is considered "exotic".
I know the Auricap is considered very "transparent" and that Spragues PIOs add a tube-like sound which is therefore not transparent. I don't know if The WIMA and the Vishay are considered "transparent" by people who use them.
I agree that any good film cap should do although the big question is how do we define "good".
The WIMA MKS2 that you seem to be using is one of three I have noted people recommend (others being the Vishay SAL 123 RPM series and the Auricap), so I know these three are "good" although I am not sure is only the Auricap is considered "exotic".
I know the Auricap is considered very "transparent" and that Spragues PIOs add a tube-like sound which is therefore not transparent. I don't know if The WIMA and the Vishay are considered "transparent" by people who use them.
Hi
I am no expert on film caps. I received two pairs of free Claritycaps weeks ago.
I used a pair of PX series caps to replace BG N type as DAC's output coupling caps, wich I thought were very transparent. After apropriate burn-in I can say they have a beautiful midrange, good botom end and soft yet detailed enough HF.
They make the BG's sound as clinical, on midrange.
Well, the good news is that they are not expensive. See the higher grade SA series here:
http://www.diycable.com/main/product_info.php?cPath=35_41&products_id=299
I have not tried my SA yet but they are said to be even better.
See them reviewed on the link I posted before:
http://www.humblehomemadehifi.com/Cap.html
I hope this helps.
M
I am no expert on film caps. I received two pairs of free Claritycaps weeks ago.
I used a pair of PX series caps to replace BG N type as DAC's output coupling caps, wich I thought were very transparent. After apropriate burn-in I can say they have a beautiful midrange, good botom end and soft yet detailed enough HF.
They make the BG's sound as clinical, on midrange.
Well, the good news is that they are not expensive. See the higher grade SA series here:
http://www.diycable.com/main/product_info.php?cPath=35_41&products_id=299
I have not tried my SA yet but they are said to be even better.
See them reviewed on the link I posted before:
http://www.humblehomemadehifi.com/Cap.html
I hope this helps.
M
Hi Max, thanks for the data about your experience. Are the PX and SA from Claritycaps as well? (There is no information on them on the diycable link).
Regarding the other link you posted, it is my understanding that a test of caps in a passive crossover for a loudspeaker filter (high power and high voltage) is not useful to draw conclusions on caps used as input (low power and low voltage). In a filter we actually expect the cap to do something to the sound (filter it) while as input we place the filter frequency very low and we are only interested on the transparency (ie want it to sound as if the cap was not present). Let me know if you think I am mistaken.
Guy
Regarding the other link you posted, it is my understanding that a test of caps in a passive crossover for a loudspeaker filter (high power and high voltage) is not useful to draw conclusions on caps used as input (low power and low voltage). In a filter we actually expect the cap to do something to the sound (filter it) while as input we place the filter frequency very low and we are only interested on the transparency (ie want it to sound as if the cap was not present). Let me know if you think I am mistaken.
Guy
Further comparison after tweaking UCD180 and AMP5
After the hints gathered on posts, I've upgraded both UCD180ad and AMP5:
UCD180ad:
1. Removed input caps
2. Connected dedicated +/-12v from supplyHG to UCD input buffer
a third upgrade, hinted by Jan-Peter, was too expensive: a second supplyHG and another transformer for a full dual mono (€200 + €40). Besides AMP5 is stereo with only one on board supply and on 96dB efficiency speakers in a 16 m2 room there would not be so much difference.
AMP5:
1. replaced 2 input caps with cheap polypropilene 4.7uF caps
2. replaced 4 ouptut caps with Wima 0.22uF caps
3. replaced cap for VN10 inductor with 1000uF (to suppress a 1kHz noise)
4. replaced coils for output filter (thanks to Bertus that has kindly sent to me the couple)
5. added some caps for EMI suppression on input and output cables (see Tripath application note AN17).
RESULTS
UCD180ad:
improvements are udible, but small. Sounds are more clear, without veils, and the floor noise seems better. For the rest impressions are similar to the first comparison: coherent and precise sounds with good instrument timbers.
AMP5:
here results are more apparent, even surprising. Input and output caps have changed a bit the character of the amp: sounds are more "educated". Maybe is the transparency gained in the high range to give a less coloured timber to strings and piano.
Replacements of cap for VN10 inductor has no effect on sound, while EMI suppression caps are bad. Yes, the noise on higher range is lower, but at expenses of dynamics.
The very surprise are the output coils made by bertus (maybe he can explain better the priciples underlying) here they are:
Well, with these coils AMP5 is another class amp! Timber of instruments become magically right like UCD. Each sound is more distinct and each instrument has more space and air from each other instead of piling up around your ears. Overall the surprise is the gained easyness of sounds. If you want to stress your system try the Kodaly's Sonata for cello solo op. 8 played by Rostropovich, at live volume. For each note, from very high to suddenly lowest cello string, you must ear all the shading resonances of its harmonic case, without cutoffs.
CONCLUSION
From the first listening remains the fact that from Tripath come out more sounds for a better resolution. With improvements AMP5 has come very close to UCD for intrinsic quality of instruments sound, so....
When with small tweaks you get big results it means that there is still much to discover. This is the case for AMP5 based on Tripath. I don't think to go much further improving UCD.
THOMAS
After the hints gathered on posts, I've upgraded both UCD180ad and AMP5:
UCD180ad:
1. Removed input caps
2. Connected dedicated +/-12v from supplyHG to UCD input buffer
a third upgrade, hinted by Jan-Peter, was too expensive: a second supplyHG and another transformer for a full dual mono (€200 + €40). Besides AMP5 is stereo with only one on board supply and on 96dB efficiency speakers in a 16 m2 room there would not be so much difference.
AMP5:
1. replaced 2 input caps with cheap polypropilene 4.7uF caps
2. replaced 4 ouptut caps with Wima 0.22uF caps
3. replaced cap for VN10 inductor with 1000uF (to suppress a 1kHz noise)
4. replaced coils for output filter (thanks to Bertus that has kindly sent to me the couple)
5. added some caps for EMI suppression on input and output cables (see Tripath application note AN17).
RESULTS
UCD180ad:
improvements are udible, but small. Sounds are more clear, without veils, and the floor noise seems better. For the rest impressions are similar to the first comparison: coherent and precise sounds with good instrument timbers.
AMP5:
here results are more apparent, even surprising. Input and output caps have changed a bit the character of the amp: sounds are more "educated". Maybe is the transparency gained in the high range to give a less coloured timber to strings and piano.
Replacements of cap for VN10 inductor has no effect on sound, while EMI suppression caps are bad. Yes, the noise on higher range is lower, but at expenses of dynamics.
The very surprise are the output coils made by bertus (maybe he can explain better the priciples underlying) here they are:
Well, with these coils AMP5 is another class amp! Timber of instruments become magically right like UCD. Each sound is more distinct and each instrument has more space and air from each other instead of piling up around your ears. Overall the surprise is the gained easyness of sounds. If you want to stress your system try the Kodaly's Sonata for cello solo op. 8 played by Rostropovich, at live volume. For each note, from very high to suddenly lowest cello string, you must ear all the shading resonances of its harmonic case, without cutoffs.
CONCLUSION
From the first listening remains the fact that from Tripath come out more sounds for a better resolution. With improvements AMP5 has come very close to UCD for intrinsic quality of instruments sound, so....
When with small tweaks you get big results it means that there is still much to discover. This is the case for AMP5 based on Tripath. I don't think to go much further improving UCD.
THOMAS
Hi Thomas
How difficult was it to connect the aux 12V outputs of the HG-Supply to the opamp on the UCD? Application note UCD180AUX.PDF by Hypex shows how to do this, but it seems not that easy.
On the HG-supply, what does VIA mean? What drill size did you use to remove the VIA and from which side did you drill?
Is there a better point to connect the ground than the extremly small solder point shown in the app note?
In your opinion, what contributes more to the sonic improvement, the removal of the coupling caps or the connection of the aux supply?
Kurt
How difficult was it to connect the aux 12V outputs of the HG-Supply to the opamp on the UCD? Application note UCD180AUX.PDF by Hypex shows how to do this, but it seems not that easy.
On the HG-supply, what does VIA mean? What drill size did you use to remove the VIA and from which side did you drill?
Is there a better point to connect the ground than the extremly small solder point shown in the app note?
In your opinion, what contributes more to the sonic improvement, the removal of the coupling caps or the connection of the aux supply?
Kurt
Don't you guys know the latest tweak?
No solder connection here! Its wire wrapping without the wire!
The twine puts pressure on the coil into the pcb and thus makes the connection, the
twine however is made of a very special hemp based fibre and when the connection gets hot, everything is just wonderful.... but you do get hungry
Sorry, just my attempt at some humor whilst I am STILL awaiting my amp 5 kit.
I too was wondering about the coil, I thought the coils should be nice and neat, the finished product measuring 11uH.
Javin5 said:
Hi Kurt,
connecting aux supply is easier than reading the Hypex pdf. Just identify T2 and T3 transistors on your modules then look at this picture of one module modified:
Removing the VIA on the supplyHG is still easier: I used scissors
. Here a picture:
The input caps for sure. But the aux supply has udible noise reduction. If you have Hypex supplyHG you have already the aux supply on board: it's worth just connecting the 3 cables for each module.
Ciao
Thomas
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.