I don't know what your preference with DAC's are, but there's a pre-order going on now for the TDA1541A Red Baron 5.0
I have one populated, and signed up for two more.
I'll use one or two (in parallel) in my modular DAC project. I'm also building a dual WM8740 on perf board and two PCM1794 PCB's(no point in perf boarding those as the PCB's where very cheap and had I2S input).
I've got quite a few of the BC550/560's and maybe 25pcs each of the BD139/140, might not be enough to get a close enough match...they're cheap though.
I match resistors, diodes etc too actually
Been doing that for a while, even when it's not necessary it provides an extra insurance that I get the right part in the right spot.
The Philips TDA1541A chip. I've heard it must be really good. But I will go for something more modern and newly available. The TDA chip is often counterfeit bs here in China and the somewhat more modern ES9018 has really nice specs and if combined with the right circuit and components there is probably little competition. I have the ES9018K2M in my laptop and I've heard the Aune X1S with the same chip and these puppies are really good even with minimal implementation. The ES9018 requires somewhat better care in terms of power and stuff.
On Taobao they sell a lot of kits and DIY builds for a price in between $150-200. They look quite nice.. Maybe I'll be looking into building something like that. I think in the mean time I will hook myself up with some SMSL M8 carrying the same chip and I'll just build a better power supply for it.
On Taobao they sell a lot of kits and DIY builds for a price in between $150-200. They look quite nice.. Maybe I'll be looking into building something like that. I think in the mean time I will hook myself up with some SMSL M8 carrying the same chip and I'll just build a better power supply for it.
So you reckon, given the optimum implementation, the ESS can equal the TDA1541A?
It depends.. the TDA is supposed to be a really good chip but it's old and discontinued. It's never as clean as the flagship ES9018 will be. I heat the TDA is REALLY good for CD players since it is designed for stereo jobs in mainly CD players. The ES9018 is waaaay more modern having alles kinds of modern features better suited to combine with modern day applications and sources. The ES9018 is multichannel where the TDA is only stereo. I think the limitations of the TDA is also 16 bit. Against 32 bit on the ES9018. The TDA can't do dsd and so on.
I don't own CD's. I have always found them a nuisance to use. So as soon as good software came available I started storing CD's in WAV and FLAC and collecting high quality digital albums. So you see my collection also consist out of a lot of 24 bit music (not that I could noticeably hear the difference with 16 bit) I could not use this music anymore.
If I would build a DAC I would build it future proof. The Es9018 has all features and functions for a time to come.
I don't own CD's. I have always found them a nuisance to use. So as soon as good software came available I started storing CD's in WAV and FLAC and collecting high quality digital albums. So you see my collection also consist out of a lot of 24 bit music (not that I could noticeably hear the difference with 16 bit) I could not use this music anymore.
If I would build a DAC I would build it future proof. The Es9018 has all features and functions for a time to come.
The Lehmann original case is smaller than Jay's Audio one.
Probably the one you're describing is more similar to the original but bad quality.
The ferrite bead alone should work, otherwise use a plain 1 Ohm metal film resistor.
With the groove you've made the CU tape would not have a constant distance from signal traces, resulting in strong impedance changes.
Search for 'Microstrip' for further infos.
From what I read it seems a single case of a probably defective PCB... never heard of other cases.
Those boards are pretty identical clones of original Lehmann boards, do you really think that a commercial amp with such a great hum/grounding problem would have ever hit the market and receive such high praise?
Their only use is for improving Mundorf's performance, if you don't hear any difference you can safely omit them.
Strange, BTW, in my case they were quite audible.
Simply trim leads diameter with a nail's file, did it several times with success.
IMHO they're the only ones that somewhat are decoupling the opamp, the elcos are probably too far to be effective.
Vishay V730P are far better, IMHO.
Haha, yeah probably! I think the ES9018 will be sufficient for quite a while. The ES9018 to my ears has a really pleasant sound when implemented right.
I'm thinking about building this thing:
https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.13.lWHPf4&id=25873632122
It has the same design as the Lehmann and is exactly as long.. I would looking nice under it. Next to the design of the housing, it's a fairly simplistic design that has some headroom for some modifications by choosing better components. The power supply seems good and I could do some opamp rolling.
Tell me what you think.
I only have the implementation in my mobile phone to go on, but to my ears ES9018 doesn't do the treble right, too noisy. So I'm not really the right person to ask about implementations of the 9018 as I avoid them. I note the one you linked is coming from Jiaxing which is where the guy behind the raindrop hui designs on eBay lived. I'm curious to know if its designed by the same man so I'll see if L.K.S has any TDA1541 designs, which was his forte in days gone by.
Ok made some final changes.
Soldered in LXZ 470uf 25V caps. Also tried other caps. doesn't make a big difference. They made more of a difference before the PSU mod. The LXZ 470uf 25V seem to be the fastest sounding and quite clear. The 560uf 35V had more 'BOOM BOOM POW' it seemed but at the loss of some details. Don't know why. The other caps were similar sounding in general.
Soldered in the Elna Silmic II 100uf 25V near the opamp. The Panasonic Pureisms have better detail but they are also harsh sounding. Just a bit to harsh for my taste. And it seems that after a while they become worse sounding than fresh. (had this with the 470uf 25V from the previous board). And that said the Elna Silmics only had a couple of hours burn in time and I know they get better after burn in. More open and detailed. So the Elna's seemed to be the best choice.
Other choices were the Nippon ASF and AWJ caps but they were nowhere as clear and I didn't give them the time to run in. Maybe I will try this some time. But for now I'm happy.
Soldered in 0,1uf AV100 MKP caps in between the Silmics. When installed they opened up the sound a bit and from all the caps I have they did the best. I also tried the Wima FKP4 0.022uf caps and they also helped opening up the sound a bit but they also made it a bit flat and colder. They were second best of all other caps.
Soldered in 0.027uf Shin Shin polystyrenes above the input cap bypass positions. They did help the sound open up a bit more too. Other choice was some orange mkp caps I had. They were just as good but in the end there was not much difference in between caps so I used the best looking and fitting ones. The red Shin Shins.
Opamp rolling really depende on which caps installed to get a balanced sound. I haven't figured it out yet. OPA111VMs sound best with acoustic music for example. Really clear and detailed. Holographic soundstage etc. But at some hiphop or other music they became REALLY harsh and ear hurting. OPA111AMs are a bit less detailed but really shine with rock music and metal. Warm and wide. OPA627BPs are a bit colder and a bit more tight than the OPA111AMs but miss depth. They are the best allrounders probably. OPA2604AP (good version) is a really good chip. Very pleasant sound and also allround but lacking a lot of detail and channel seperation. The sound gets a bit more homogenic and a bit boring.
I wanted to use the BCL also as a pre amp but this is not going to work. The speakers need a much clearer source than my headphones. My headphones hurt my ears with a clear source like the Aune T1. Everything gets too harsh. So I use the Sanskrit 6 for the headphones right now and the Aune T1 for my active speakers. I will put my other BCL clone together again for a nice A/B review between standard version and my version, but also to hook my speakers on to.
It is Chinese spring festival here now, so I have to wait a bit to start building my new DAC. But I've already contact a shop and made a deal that I get a nice discount on the parts unassembled after the spring break. I have one week of holiday left to build it. It will be the one from the link I send you with some customized parts choices. And I will share the layout here for tips to improve the design.
https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.13.jhWLET&id=25873632122
Will cost me Y1800 unassembled. Which is €245 or $275. I'll put a bit of extra money in it on better parts here and there and some improvements but I think for about $350 this will be a really nice DAC that will fit the 2 Lehmann BCLs very nicely in therms of design and performance.
Soldered in LXZ 470uf 25V caps. Also tried other caps. doesn't make a big difference. They made more of a difference before the PSU mod. The LXZ 470uf 25V seem to be the fastest sounding and quite clear. The 560uf 35V had more 'BOOM BOOM POW' it seemed but at the loss of some details. Don't know why. The other caps were similar sounding in general.
Soldered in the Elna Silmic II 100uf 25V near the opamp. The Panasonic Pureisms have better detail but they are also harsh sounding. Just a bit to harsh for my taste. And it seems that after a while they become worse sounding than fresh. (had this with the 470uf 25V from the previous board). And that said the Elna Silmics only had a couple of hours burn in time and I know they get better after burn in. More open and detailed. So the Elna's seemed to be the best choice.
Other choices were the Nippon ASF and AWJ caps but they were nowhere as clear and I didn't give them the time to run in. Maybe I will try this some time. But for now I'm happy.
Soldered in 0,1uf AV100 MKP caps in between the Silmics. When installed they opened up the sound a bit and from all the caps I have they did the best. I also tried the Wima FKP4 0.022uf caps and they also helped opening up the sound a bit but they also made it a bit flat and colder. They were second best of all other caps.
Soldered in 0.027uf Shin Shin polystyrenes above the input cap bypass positions. They did help the sound open up a bit more too. Other choice was some orange mkp caps I had. They were just as good but in the end there was not much difference in between caps so I used the best looking and fitting ones. The red Shin Shins.
Opamp rolling really depende on which caps installed to get a balanced sound. I haven't figured it out yet. OPA111VMs sound best with acoustic music for example. Really clear and detailed. Holographic soundstage etc. But at some hiphop or other music they became REALLY harsh and ear hurting. OPA111AMs are a bit less detailed but really shine with rock music and metal. Warm and wide. OPA627BPs are a bit colder and a bit more tight than the OPA111AMs but miss depth. They are the best allrounders probably. OPA2604AP (good version) is a really good chip. Very pleasant sound and also allround but lacking a lot of detail and channel seperation. The sound gets a bit more homogenic and a bit boring.
I wanted to use the BCL also as a pre amp but this is not going to work. The speakers need a much clearer source than my headphones. My headphones hurt my ears with a clear source like the Aune T1. Everything gets too harsh. So I use the Sanskrit 6 for the headphones right now and the Aune T1 for my active speakers. I will put my other BCL clone together again for a nice A/B review between standard version and my version, but also to hook my speakers on to.
It is Chinese spring festival here now, so I have to wait a bit to start building my new DAC. But I've already contact a shop and made a deal that I get a nice discount on the parts unassembled after the spring break. I have one week of holiday left to build it. It will be the one from the link I send you with some customized parts choices. And I will share the layout here for tips to improve the design.
https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.13.jhWLET&id=25873632122
Will cost me Y1800 unassembled. Which is €245 or $275. I'll put a bit of extra money in it on better parts here and there and some improvements but I think for about $350 this will be a really nice DAC that will fit the 2 Lehmann BCLs very nicely in therms of design and performance.
Something interesting happened. the inner Silmic capacitor at the opamp started swelling up and I could hear a noticeable difference in volume balance. The right channel (swollen cap) was noticeable softer and less open. It was like the cap had to work to hard. So I figured that I probably damaged the AV 0.1uf cap when bending the lead to short to the case to make it fit. I replaced both caps with the original Wima FKP4 caps. After that I replaced the Silmics with Nippon ASF 470uf 50V caps to give them a chance.
They sound fine. But now I notice I have a channel inbalance. It's not the same as it was but left is definitely a bit louder. If I correct the balance on in the windows settings to R80% L100%, the stereo image is in the middle again.
The inner cap gets a bit warmer, but this may also come because it's right next to the transistors, voltage regulators and touching an LXZ psu output cap.. I will also put the Wima FKP4 caps back at the input position an retrace my steps to find out what is fooling around.
EDIT: After some checking.. I notice my hearing right is also not as good as left haha. Because It sounds softer on every headphone and amp I have.. So I guess it's also personal handicap... I'm just 26.. To much parties..
They sound fine. But now I notice I have a channel inbalance. It's not the same as it was but left is definitely a bit louder. If I correct the balance on in the windows settings to R80% L100%, the stereo image is in the middle again.
The inner cap gets a bit warmer, but this may also come because it's right next to the transistors, voltage regulators and touching an LXZ psu output cap.. I will also put the Wima FKP4 caps back at the input position an retrace my steps to find out what is fooling around.
EDIT: After some checking.. I notice my hearing right is also not as good as left haha. Because It sounds softer on every headphone and amp I have.. So I guess it's also personal handicap... I'm just 26.. To much parties..
Last edited:
Replaced the 470uf 25V caps with the 560uf 35V caps. Since the small ones really got quite hot. This wouldn't be good in the long run for the amp. I like the somewhat softened sound by it. Everything gets a bit warmer.. maybe slower but less harsh and sibilance is the thing I hate the most out of everything!
Also put the Wima FKP4 at the input again. They were not bad when I tested them. Some guy in the the Lovely cube Head-Fi topic said Wima FKP4 are notoriously bad for sound. I don't really agree. He said even 50 year old polystyrenes would beat them easily.. It's just not true. In this circuit the difference with or without these bypass caps is already really small and the difference between different caps was even less noticeable. There is a reason Lehmann put em there and thats because the don't harm the sound at all. So yeah the Wima FKP4 a considered a fine choice for these positions.
The amp also seems to have way more power.. or blocking less of the input signal. I don't have to use the full gain anymore for my Mad Dog planar magnetic headphones.
This is probably the result of the somewhat bigger power supply output caps. (This is probably also why the small ones got really hot with me having the dip switches on full gain.)
Also did a final good clean on the board. With a brush and some (Chinese chemical) PCB cleaner. Then with a cotton swap and the same chemical waste. Then with a toothpick with the PCB cleaner and finally a coton swab again.
Guess what.. DC offset lowered even more! OPA111VM has no offset OPA111AM went from -1mv/6mv to 0mv/2mv. The rest I haven't checked yet but they were already quite good. I think there was still some solder paste in the PSU section where I used a toothpick in between the small solder points.
Oh and the imbalance seems gone.
Quite happy with the sound now. Did an A/B comparison with my Aune X1SE/Pro Headamp which is able to drive 600 ohm headphones and is not a bad amp at all.. But the BCL blows it out of the water! Everything is deeper, fuller, more natural, wider etc. Much more alive and it feels like without breaking a sweat.
Also put the Wima FKP4 at the input again. They were not bad when I tested them. Some guy in the the Lovely cube Head-Fi topic said Wima FKP4 are notoriously bad for sound. I don't really agree. He said even 50 year old polystyrenes would beat them easily.. It's just not true. In this circuit the difference with or without these bypass caps is already really small and the difference between different caps was even less noticeable. There is a reason Lehmann put em there and thats because the don't harm the sound at all. So yeah the Wima FKP4 a considered a fine choice for these positions.
The amp also seems to have way more power.. or blocking less of the input signal. I don't have to use the full gain anymore for my Mad Dog planar magnetic headphones.
This is probably the result of the somewhat bigger power supply output caps. (This is probably also why the small ones got really hot with me having the dip switches on full gain.)
Also did a final good clean on the board. With a brush and some (Chinese chemical) PCB cleaner. Then with a cotton swap and the same chemical waste. Then with a toothpick with the PCB cleaner and finally a coton swab again.
Guess what.. DC offset lowered even more! OPA111VM has no offset OPA111AM went from -1mv/6mv to 0mv/2mv. The rest I haven't checked yet but they were already quite good. I think there was still some solder paste in the PSU section where I used a toothpick in between the small solder points.
Oh and the imbalance seems gone.
Quite happy with the sound now. Did an A/B comparison with my Aune X1SE/Pro Headamp which is able to drive 600 ohm headphones and is not a bad amp at all.. But the BCL blows it out of the water! Everything is deeper, fuller, more natural, wider etc. Much more alive and it feels like without breaking a sweat.
Last edited:
The amp Is exactly doing what it should now. The sound is pleasant, warm but clear and detailed. Very wide soundstage and holographic imaging.
Next step will be building the DAC based on the ESS Sabre ES9018 32-bit chip with Amanero USB interface. I will receive it in two weeks.
I've tested the amp with severel albums which are well recorded and have wide dynamic range:
Nirvana - MTV Unplugged in New York
Muse - Absolution
R.E.M. - Automatic for the People
Dr Dre - 2001
Plastikman - EX (Guggenheim)
Katie Melua - Piece by Piece
These albums all sound very good. When you get better gear.. lesser albums tend to sound really BAD with capital letters.. the new Dr Dre album 'Compton' is a big pile of BS because it is recorded for his stupid beats headphones. The albums is way to bright and sibilant. Just painful to listen to. What a waste. I thought he was one of the only HipHop entertainers that did care about sound quality.. He took a lot of years for this album.. (It does sound good on my active speakers btw. Haha. So probably it's recorded towards speakers.. )
Next step will be building the DAC based on the ESS Sabre ES9018 32-bit chip with Amanero USB interface. I will receive it in two weeks.
I've tested the amp with severel albums which are well recorded and have wide dynamic range:
Nirvana - MTV Unplugged in New York
Muse - Absolution
R.E.M. - Automatic for the People
Dr Dre - 2001
Plastikman - EX (Guggenheim)
Katie Melua - Piece by Piece
These albums all sound very good. When you get better gear.. lesser albums tend to sound really BAD with capital letters.. the new Dr Dre album 'Compton' is a big pile of BS because it is recorded for his stupid beats headphones. The albums is way to bright and sibilant. Just painful to listen to. What a waste. I thought he was one of the only HipHop entertainers that did care about sound quality.. He took a lot of years for this album.. (It does sound good on my active speakers btw. Haha. So probably it's recorded towards speakers.. )
After some burn in time and another listening session I conclude that the Nippon ASF caps are really soft caps. They have good detail but are not the most detailed. The sound on the other hand is really pleasant. It's a great combination with the Mad Dog headphones. They are already a bit dark and the BCL just makes them run real smooth but very exact. The sound is a bit buttery but not muffled. Probably the 560uf PSU caps are also responsible for this effect a bit. It is a really natural and comfy sound.
I replaced the output resistors for RA resistors that finally came in. They are just for looks.. since I don't hear any noticeable difference. They do their thing very well. Just as well as the Vishay Dale MBEM0414. Well they kinda look cool and they are marked 'USA'.
Can't wait to build a matching DAC for the system to see if things improve a bit more. Because the Sanskrit 6 is transparant and good but it just lacks in soundstage and imaging. The Aune T1 is more holographic but lacks in the lower regions and can be a bit harsh (with all tubes I have). So yeah a good DAC could make it even a better experience.
I think for now that finishes this blog a bit. I tried to share as much info on the build including all hickups and bumps in the road. It was a very nice learning experience. I want to thank everyone for helping out here.
On to the Next project!
I replaced the output resistors for RA resistors that finally came in. They are just for looks.. since I don't hear any noticeable difference. They do their thing very well. Just as well as the Vishay Dale MBEM0414. Well they kinda look cool and they are marked 'USA'
. Can't wait to build a matching DAC for the system to see if things improve a bit more. Because the Sanskrit 6 is transparant and good but it just lacks in soundstage and imaging. The Aune T1 is more holographic but lacks in the lower regions and can be a bit harsh (with all tubes I have). So yeah a good DAC could make it even a better experience.
I think for now that finishes this blog a bit. I tried to share as much info on the build including all hickups and bumps in the road. It was a very nice learning experience. I want to thank everyone for helping out here.
On to the Next project!
Cidious, you pay a lot of attention to capacitors so maybe it will be of use to you.
Just yesterday did some measurements on a friend's DIY capacitor testing rig. Although it can't go higher than about 15khz yet so we choose lower frequency to be sure:
Z - raw impedance
D - dissipation
R - Real (total) ESR
P - phase (at frequency)
V - measured real value (at frequency)
Test frequency - 13600Hz
These 3 are 100uF/25V caps:
Elna Silmic II Z=0,23; D=2,27; R=0,22; P=-26; V=122
Nichicon KZ Z=0,155; D=1,3; R=0,127; P=-37; V=120
Panasonic FM Z=0,14; D=1,1; R=0,103; P=-42; V=119
And some BG-PK 220uf/4V i had lying around:
BG-PK Z=0,39; D=6,6; R=0,38; P=-8,5; V=200
Test frequency - 1000Hz
Elna Silmic II Z=1,7; D=0,149; R=0,252; P=-81; V=93
Nichicon KZ Z=1,7; D=0,09; R=0,167; P=-84; V=92
These are 220uF/10V for comparison:
OSCON SP Z=0,139; D=1,1; R=0,103; P=-42; V=119
Test frequency - 93Hz
Elna Silmic II Z=17,3; D=0,04; R=0,7; P=-87; V=97
Nichicon KZ Z=17,6; D=0,03; R=0,6; P=-87; V=96
BG-PK 220/4 Z=9,1; D=0,11; R=1; P=-83; V=186
OSCON SP Z=6,2; D=0,03; R=0,18; P=-88; V=273
*Not including FM after first test because it was very close/same as KZ in all others.
*Note that BG-PK is 4V cap, 25V should be at least 2 times better ESR wise so not too bad...
I don't know who spread the gossip about "audio" capacitors having much worse ESR than industrial capacitors. They are a bit worse than panasonic FM, but much better than FC series which was considered "very low ESR" before FM came out..
Best of luck in your future diy&tweaks
Just yesterday did some measurements on a friend's DIY capacitor testing rig. Although it can't go higher than about 15khz yet so we choose lower frequency to be sure:
Z - raw impedance
D - dissipation
R - Real (total) ESR
P - phase (at frequency)
V - measured real value (at frequency)
Test frequency - 13600Hz
These 3 are 100uF/25V caps:
Elna Silmic II Z=0,23; D=2,27; R=0,22; P=-26; V=122
Nichicon KZ Z=0,155; D=1,3; R=0,127; P=-37; V=120
Panasonic FM Z=0,14; D=1,1; R=0,103; P=-42; V=119
And some BG-PK 220uf/4V i had lying around:
BG-PK Z=0,39; D=6,6; R=0,38; P=-8,5; V=200
Test frequency - 1000Hz
Elna Silmic II Z=1,7; D=0,149; R=0,252; P=-81; V=93
Nichicon KZ Z=1,7; D=0,09; R=0,167; P=-84; V=92
These are 220uF/10V for comparison:
OSCON SP Z=0,139; D=1,1; R=0,103; P=-42; V=119
Test frequency - 93Hz
Elna Silmic II Z=17,3; D=0,04; R=0,7; P=-87; V=97
Nichicon KZ Z=17,6; D=0,03; R=0,6; P=-87; V=96
BG-PK 220/4 Z=9,1; D=0,11; R=1; P=-83; V=186
OSCON SP Z=6,2; D=0,03; R=0,18; P=-88; V=273
*Not including FM after first test because it was very close/same as KZ in all others.
*Note that BG-PK is 4V cap, 25V should be at least 2 times better ESR wise so not too bad...
I don't know who spread the gossip about "audio" capacitors having much worse ESR than industrial capacitors. They are a bit worse than panasonic FM, but much better than FC series which was considered "very low ESR" before FM came out..
Best of luck in your future diy&tweaks
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.