Thanks all,
Sorry if my questions look like beginner's (which they kind of are
). Anyway, from the information gathered the measured differerence to Tripath is better SNR. There are also some subjective sonical differences. However, the class D amps are new to most of all, and there is little design information available (I've read about things like half-brigde etc. but don't know what it means in practise), so what i questioned was something related to design (class-T / class-D differences etc.). Maybe that's information can be searched somewhere else (google etc.).
Also it seems that the power supply plays a great role with UcD amps (maybe also with class-T amps). I have a Tripath amp (SoundPax 2x15W or so, depends on PSU) connected to a pair of Wharfedale Diamonds on my PC. The PSU is underpowered but I still like the sound, it also lacks the background noise of solid state / chip amps (or the noise is different). It would be nice if there were some application notes available for the applications like DACs (I think that if you can leave out or use a less steep antialiasing filter on DAC you'll get much better phase response, I think that's also the main reason for using sample rates greater than 44,1 kHz), active monitors and usual stereo amps. In this thread there was already mentioned the power supply calculations. It was also mentioned that with a 2-way active monitor system you should have similar amps for both bass and treble channels (but is the requirement similar for the PSU, I mean is it really 2x the channels or will 1.5x be enough).
Sorry if my questions look like beginner's (which they kind of are
). Anyway, from the information gathered the measured differerence to Tripath is better SNR. There are also some subjective sonical differences. However, the class D amps are new to most of all, and there is little design information available (I've read about things like half-brigde etc. but don't know what it means in practise), so what i questioned was something related to design (class-T / class-D differences etc.). Maybe that's information can be searched somewhere else (google etc.).Also it seems that the power supply plays a great role with UcD amps (maybe also with class-T amps). I have a Tripath amp (SoundPax 2x15W or so, depends on PSU) connected to a pair of Wharfedale Diamonds on my PC. The PSU is underpowered but I still like the sound, it also lacks the background noise of solid state / chip amps (or the noise is different). It would be nice if there were some application notes available for the applications like DACs (I think that if you can leave out or use a less steep antialiasing filter on DAC you'll get much better phase response, I think that's also the main reason for using sample rates greater than 44,1 kHz), active monitors and usual stereo amps. In this thread there was already mentioned the power supply calculations. It was also mentioned that with a 2-way active monitor system you should have similar amps for both bass and treble channels (but is the requirement similar for the PSU, I mean is it really 2x the channels or will 1.5x be enough).
It's hard to decide "in what order" the differences between tripath and ucd should be listed. The best thing to do is to get hold of the product literature of both types and make comparisons. The differences that I set most stall by are output impedance (frequency response at various loads) and THD vs frequency.
Full bridge/half bridge are normally not inherent differences in people's design strategies. Most common class D control schemes work in full and half bridge alike.
Power supplies effects are extremely important for so called "fully digital amplifiers" (TI, ApogeeDDX, Pulsus etc). For analog class D amps like UcD, Mueta, Tripath, ICEpower etc power supply sensitivity is chiefly comparable to that of "normal" linear amps ie. not important for measured performance, surprisingly important to sound.
The reason why the sonic effect of power supplies is discussed relatively often in this thread is because the ucd crowd (including the designer) are mostly audiophiles, to whom it means a lot to get the last drop of quality out of a circuit.
Full bridge/half bridge are normally not inherent differences in people's design strategies. Most common class D control schemes work in full and half bridge alike.
Power supplies effects are extremely important for so called "fully digital amplifiers" (TI, ApogeeDDX, Pulsus etc). For analog class D amps like UcD, Mueta, Tripath, ICEpower etc power supply sensitivity is chiefly comparable to that of "normal" linear amps ie. not important for measured performance, surprisingly important to sound.
The reason why the sonic effect of power supplies is discussed relatively often in this thread is because the ucd crowd (including the designer) are mostly audiophiles, to whom it means a lot to get the last drop of quality out of a circuit.
And the winner is ............
Yes, Ghemink has a Tripath Based Marantz and he also has a quite expense Accuphase E407 that is considered a high-end integrated amp. Well, I can tell you that the UcD180 blows them both away.
Finally I had a decent test setup for my UcD180 finished (see attachment for a low-res and heavily compressed picture). And tonight I compared the above 3 amps. The power-amp stage of the Accuphase, the Tripath based Marantz and the UcD180 amps. They played on my B&W CDM9NT which is not my prefered system (I prefer my active 3-way), but it is the only way I can compare these 3 amps.
All amps were driven by the DEQX PDC unit without any speaker correction etc. So CD/SACD signal went into the PDC and I used the PDC as a pre-amp. I had all 3 amps switched on and I had 3 pairs of almost identical speaker cables. The Tripath and the UcD had the same cables (ortofon) and the same length (3 meters). The Accuphase also had ortofon cables, similar type but a bit cheaper, so the the cable may play a small role here. Then the UcD180 was hooked up via XLR using a 2m goertz micropurl silver (self assembled). The Marantz and the Accuphase were hooked up via RCA cables (good quality) of about 1m.
The ranking in order of sound quality was: 1. UcD180; 2. Tripath; and 3. Accuphase. The Tripath sounded better than the Accuphase, I knew that already since I had the Tripath (Marantz) already more than a year. However, the UcD180 sounded much better than the Tripath. Comparing these 3 amps, the Tripath seems marginally better than the Accuphase (but clearly noticable) and the UcD180 is a large step ahead of both others. It is really amazing. The UcD180 is very transparant, acoustical music sounded as it did never before. Female vocals are very good. Acoustic guitar is amazing. It is the transparancy, the air, the imaging that directly grabs you. You immediately know you have something special when you hear this amp. Bruno and Jan-Peter, congratulations with this very good amp.
As you can see in the picture. I use a quite solid power supply with 6x10.000uF in total for 2 channels. I`m using a 250W transformer. I don`t know how important those 4 ELNA caps are. they were expensive so they`d better sound good. I don`t know how it would sound without them. The UcD180 modules that I have in this setup are unmodified and not burned in. They sound great from the beginning. I started modifying another pair of UcD180 (using BG caps etc). This will sure take a couple of days too. I`ll compare the modded ones with the unmodded UcD180 as well.
Also I don`t know how big the difference in the cables are. The difference between the Tripath and the UcD180 is only in the interconnect cables (speaker cables exactly the same, even bought at the same time). The Tripath used RCA 1m while the UcD180 used 2m micropurl silver XLR. I need to assemble a couple more micropurl cables anyway for the active 3-way, so then I can feed the Tripath via the same XLR cables and see what happens then. I`ll sure do that. May take a couple of days.
Bottomline: these amps are very good, I was surprised to hear this much music and this much difference on my passive CDM9NTs.
Best regards
Gertjan
Jan-Peter said:
Yes, Ghemink has a Tripath Based Marantz and he also has a quite expense Accuphase E407 that is considered a high-end integrated amp. Well, I can tell you that the UcD180 blows them both away.
Finally I had a decent test setup for my UcD180 finished (see attachment for a low-res and heavily compressed picture). And tonight I compared the above 3 amps. The power-amp stage of the Accuphase, the Tripath based Marantz and the UcD180 amps. They played on my B&W CDM9NT which is not my prefered system (I prefer my active 3-way), but it is the only way I can compare these 3 amps.
All amps were driven by the DEQX PDC unit without any speaker correction etc. So CD/SACD signal went into the PDC and I used the PDC as a pre-amp. I had all 3 amps switched on and I had 3 pairs of almost identical speaker cables. The Tripath and the UcD had the same cables (ortofon) and the same length (3 meters). The Accuphase also had ortofon cables, similar type but a bit cheaper, so the the cable may play a small role here. Then the UcD180 was hooked up via XLR using a 2m goertz micropurl silver (self assembled). The Marantz and the Accuphase were hooked up via RCA cables (good quality) of about 1m.
The ranking in order of sound quality was: 1. UcD180; 2. Tripath; and 3. Accuphase. The Tripath sounded better than the Accuphase, I knew that already since I had the Tripath (Marantz) already more than a year. However, the UcD180 sounded much better than the Tripath. Comparing these 3 amps, the Tripath seems marginally better than the Accuphase (but clearly noticable) and the UcD180 is a large step ahead of both others. It is really amazing. The UcD180 is very transparant, acoustical music sounded as it did never before. Female vocals are very good. Acoustic guitar is amazing. It is the transparancy, the air, the imaging that directly grabs you. You immediately know you have something special when you hear this amp. Bruno and Jan-Peter, congratulations with this very good amp.
As you can see in the picture. I use a quite solid power supply with 6x10.000uF in total for 2 channels. I`m using a 250W transformer. I don`t know how important those 4 ELNA caps are. they were expensive so they`d better sound good. I don`t know how it would sound without them. The UcD180 modules that I have in this setup are unmodified and not burned in. They sound great from the beginning. I started modifying another pair of UcD180 (using BG caps etc). This will sure take a couple of days too. I`ll compare the modded ones with the unmodded UcD180 as well.
Also I don`t know how big the difference in the cables are. The difference between the Tripath and the UcD180 is only in the interconnect cables (speaker cables exactly the same, even bought at the same time). The Tripath used RCA 1m while the UcD180 used 2m micropurl silver XLR. I need to assemble a couple more micropurl cables anyway for the active 3-way, so then I can feed the Tripath via the same XLR cables and see what happens then. I`ll sure do that. May take a couple of days.
Bottomline: these amps are very good, I was surprised to hear this much music and this much difference on my passive CDM9NTs.
Best regards
Gertjan
Attachments
I have a prototype system with the UCD 180 running, and it sounds pretty good. However, I want it to sound much better with some tweaks. I am waiting for AD8620 and 12v zeners to arrive in the mail so I can try that. I am also curiuos whether anyone else has tried replacing more capacitors in the circuit with black gates. The 470uF big caps could be replaced with black gates, although of a smaller value because the 470uF BG is too wide. Has anyone else tried this?
The BW is a natural result of the choice of loop gain (for LF loop control) and loop phase (for self-oscillation). No explicit band-limiting is performed.kwhilden said:
In order to obtain higher bandwidths some other parameters need to be changed too, most to the detriment of sound quality (unless you're ready to boost the switching frequency as well).
In any case, extrapolating from tests done in the context of SACD, it appeared no quality improvement is to be had from extending system bandwidth above 70kHz. It is improductive to set the bandwidth target above this value because one gets all the problems and none of the advantages.
switchmode power supply
What is the prevailing wisdom regarding using a switching power supply for a class D amplifier? Can this have a sonic benefit or detriment? What are the crucial factors that would make this a good or bad idea?
I certainly like the idea of a much smaller and lighter power supply.
Thankyou,
KW
What is the prevailing wisdom regarding using a switching power supply for a class D amplifier? Can this have a sonic benefit or detriment? What are the crucial factors that would make this a good or bad idea?
I certainly like the idea of a much smaller and lighter power supply.
Thankyou,
KW
Although strictly not this threads topic, Casework & front panels are always a problem – well at least for me. Found these details on another thread.
Just had a quick look on emachineshop.com site, they offer free CAD software for down load, you design your latest creation, gives you a quote – if you agree - hey presto – you latest pride and joy will be delivered to your door – now that’s what I call cool!
OK – when will we able to download software to design our perfect partner(s)? – with a few clicks, a knock at the door and perpetual happiness
I sometimes feel I was born half a century too early! and apparently I have my very first gray hair - only down hill from here am afraid....
John
Just had a quick look on emachineshop.com site, they offer free CAD software for down load, you design your latest creation, gives you a quote – if you agree - hey presto – you latest pride and joy will be delivered to your door – now that’s what I call cool!
OK – when will we able to download software to design our perfect partner(s)? – with a few clicks, a knock at the door and perpetual happiness
I sometimes feel I was born half a century too early! and apparently I have my very first gray hair - only down hill from here am afraid....John
ghemink,
That's an absolutely brilliant and interesting report! Got me thinking maybe I should try more DIY!
Is it possible for me to get a copy of your power supply schematics so I can try to build a similar power supply for a home audio stereo version of the UcD180? BTW is 250W transformer that you use sifficient for the UcD 180? I thought two UcD 180's would consume more power than 250W? Thanks.
That's an absolutely brilliant and interesting report! Got me thinking maybe I should try more DIY!
Is it possible for me to get a copy of your power supply schematics so I can try to build a similar power supply for a home audio stereo version of the UcD180? BTW is 250W transformer that you use sifficient for the UcD 180? I thought two UcD 180's would consume more power than 250W? Thanks.
Re: switchmode power supply
pro: a smps is several times lighter
pro: well regulated smps voltage allows full peak audio power even at low line
pro: with power factor correction smps may deliver twice the power of standard supply without exceeding line current rating
con: smps common mode noise from switching frequency may generate audible beat products with class d amp's switching frequency
con: unlike line frequency transformers, a generic smps may cycle or shut down during sustained audio power peaks
With proper care in design or selection a switching power supply is the perfect compliment to a class d amplifier.
Regards -- analog(spiceman)
As I see it:kwhilden said:
pro: a smps is several times lighter
pro: well regulated smps voltage allows full peak audio power even at low line
pro: with power factor correction smps may deliver twice the power of standard supply without exceeding line current rating
con: smps common mode noise from switching frequency may generate audible beat products with class d amp's switching frequency
con: unlike line frequency transformers, a generic smps may cycle or shut down during sustained audio power peaks
With proper care in design or selection a switching power supply is the perfect compliment to a class d amplifier.
Regards -- analog(spiceman)
audionutter said:
Hi Audionutter,
The power supply that I used is the predator power supply that I bought at lcaudio. The soft delay circuit next to the transformer is also from LCaudio. What I added is the 4 big ELNA Cerafine caps, 10.000uF 63V. I`m suspecting that these caps play a role in the good sound that I`m getting. The reason I use only a 250W transformer is because these amps are meant for use with a tweeter and two midranges drivers so I figures that a 250W amp is more than enough to blow up the tweeter and midranges at the same time
. Besides that, those power supply caps hold a lot of reserve power. Those two UcD180`s can of course consume more than 250W together, however, who is running a continuos sine wave to run both of them at 180W (4ohm). I guess with a normal music signal that stays below clipping level, this 250W transformer should be more than enough. Please add a bigger one if you feel that is needed. Best regards
Gertjan
Thanks for the feedback on switching power supplies. It seems like there are a LOT of companies making them, but I'm having a hard time figuring out how to decide what will work best. Can anyone offer suggestions of companies that produce a high quality SMPS? Are there any particular specs of SMPS's that are important for sound quality of the final product?
Thanks again!
KW
Thanks again!
KW
possible SMPS
kwhilden,
One possible SMPS is the kitoffered by A & T labs.
http://www.a-and-t-labs.com/K6_Sw_Amp/index.htm
George
kwhilden,
One possible SMPS is the kitoffered by A & T labs.
http://www.a-and-t-labs.com/K6_Sw_Amp/index.htm
George
Hi,
maybe Jan or Bruno are working on smps for the UCD ? compared to an ucd a psu should be easier ;-) and there is perhaps a market for this... i 've searched over the net and the only PSU i can found with shematics was a-and-t labs, but nobody seems to use it on the net...
I want to take some modules at Hypex, and i'm really thinking of an SMPS for this modules to increase noise rejection, the question is does an UCD180 could work well with an SMPS ?
few years ago i've made an Aleph 30 and we have done a simple test with a friend to compare several amps : just take a film on dvd with someone speaking (we have done the test with an audio cd of poetry and increase the volume until you can understand all the words on the loudspeaker. replace the dvd with a sine on a test cd, and with a multimeter see how many volts you have on the loudspeaker terminals. i don't remember exactly the results but the aleph wins and starts to be understandable with a very low tension. I hope the UCD could bring me the same kind of power feeling at low volumes like a single ended Class A, the feeling of deep bass at low volume is also very interesting.
Pierre-Antoine.
maybe Jan or Bruno are working on smps for the UCD ? compared to an ucd a psu should be easier ;-) and there is perhaps a market for this... i 've searched over the net and the only PSU i can found with shematics was a-and-t labs, but nobody seems to use it on the net...
I want to take some modules at Hypex, and i'm really thinking of an SMPS for this modules to increase noise rejection, the question is does an UCD180 could work well with an SMPS ?
few years ago i've made an Aleph 30 and we have done a simple test with a friend to compare several amps : just take a film on dvd with someone speaking (we have done the test with an audio cd of poetry
and increase the volume until you can understand all the words on the loudspeaker. replace the dvd with a sine on a test cd, and with a multimeter see how many volts you have on the loudspeaker terminals. i don't remember exactly the results but the aleph wins and starts to be understandable with a very low tension. I hope the UCD could bring me the same kind of power feeling at low volumes like a single ended Class A, the feeling of deep bass at low volume is also very interesting.Pierre-Antoine.
One can add a simple non-dissapative voltage equalization circuit across the power supply rails to a half bridge class d amp in order make the two rails track each other via active charge transfer. It takes a transformer and two mosfets so the magnetics and mosfet count is no better than going with a full bridge output on the class d amp, but the added parts needn't be as beefy as those for the existing half bridge, and if the amp is already purchased, it may be the best solution.IVX said:
For what typical rail voltages and maximum audio output power levels should such an equalization circuit be designed to meet the needs of most of the diy-ers around here? (+/-80Vmax and 500Wmax?)
On the face of it, "getting an SMPS to work" may be simpler than UcD (as the large amount of working power supplies of various quality shows), making a really good SMPS is not to be sneezed at.pa said:Hi,
maybe Jan or Bruno are working on smps for the UCD ? compared to an ucd a psu should be easier ;-) and there is perhaps a market for this... i 've searched over the net and the only PSU i can found with shematics was a-and-t labs, but nobody seems to use it on the net...
I want to take some modules at Hypex, and i'm really thinking of an SMPS for this modules to increase noise rejection, the question is does an UCD180 could work well with an SMPS ?
few years ago i've made an Aleph 30 and we have done a simple test with a friend to compare several amps : just take a film on dvd with someone speaking (we have done the test with an audio cd of poetry
Pierre-Antoine.
Off the shelf (OTS) SMPS's tend to have the following failings:
1)Maximum peak power=average power
A 100W power amp draws 200W peaks. An OTS supply will therefore have to be selected for 200W. Of course it'll be overrated then, too large and too expensive. I'm unaware of SMPS's where thermal design and regulation (sagging instead of cutting out) are optimised for audio use. As a result, using a standard SMPS will make your product lighter but much bigger than a 50Hz transformer!
2)Terrible EMI. The layout quality standard of most SMPS's is years behind on that used by seasoned class D designers.
3)Ridiculously large Y caps between primary and secondary (made necessary by incorrect choice of transformer form factor and pinout).
4)PFC, if present, overshoots at around 17Hz (current feedforward never heard of, although many of these controller chips have dedicated input pins for this)
All methods necessary to make good SMPS's are known and understood, but only by a small number of experts (I have good reasons to believe analogspiceman is one of them).
Unfortunately, these true experts number hardly more than the number of class D experts. Until one of these few takes time out to do something special for audio, we're still better off using linear supplies.
I am personally trying to find time to make such a "good smps", more specifically a single stage PFC+galvanic isolation, the aim being to put the storage on the secondary side, get high efficiency, good pfc but only basic regulation.
Unfortunately my experience with high-voltage (primary side) design is not very great so a product won't be ready anytime soon. In the meantime - all eyes on analogspiceman
Cheers,
Bruno
Bruno Putzeys said:
On the face of it, "getting an SMPS to work" may be simpler than UcD (as the large amount of working power supplies of various quality shows), making a really good SMPS is not to be sneezed at.
Off the shelf (OTS) SMPS's tend to have the following failings:
1)Maximum peak power=average power
A 100W power amp draws 200W peaks. An OTS supply will therefore have to be selected for 200W. Of course it'll be overrated then, too large and too expensive. I'm unaware of SMPS's where thermal design and regulation (sagging instead of cutting out) are optimised for audio use. As a result, using a standard SMPS will make your product lighter but much bigger than a 50Hz transformer!
2)Terrible EMI. The layout quality standard of most SMPS's is years behind on that used by seasoned class D designers.
3)Ridiculously large Y caps between primary and secondary (made necessary by incorrect choice of transformer form factor and pinout).
4)PFC, if present, overshoots at around 17Hz (current feedforward never heard of, although many of these controller chips have dedicated input pins for this)
All methods necessary to make good SMPS's are known and understood, but only by a small number of experts (I have good reasons to believe analogspiceman is one of them).
Unfortunately, these true experts number hardly more than the number of class D experts. Until one of these few takes time out to do something special for audio, we're still better off using linear supplies.
I am personally trying to find time to make such a "good smps", more specifically a single stage PFC+galvanic isolation, the aim being to put the storage on the secondary side, get high efficiency, good pfc but only basic regulation.
Unfortunately my experience with high-voltage (primary side) design is not very great so a product won't be ready anytime soon. In the meantime - all eyes on analogspiceman
Cheers,
Bruno
I really dont see the big problem adapting ucd tec to psu transformer primary side driver, having optical sec side feedback as the voltage tracking, other insulation feedback is also possible. Just like tracking audiosignals, tracking of a reference voltage is not much different is it? iven sort of paralelling is not far away, like synkroniced switching
Triac switching is an other option.
Obvius you can make an audio band "regulator", so the psu regulator will bee just as good or ?
In the meantime - all eyes on analogspiceman
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.