[Headamp] upgrading a Lehmann BCL clone

I am looking for resistors. I can only get the DALE RCN55 in these values or the RLR20 (they are the 1/2 versions of the RLR07s I think.. not sure)

Any help?

After removing both coupling caps (the main 2.2uF one and the .22 bypass) you must restore continuity. To do it youl'll need to solder a piece of wire to the pads of one of the two capacitors position (AKA short it).

BTW verify before that your source/preamp have DC blocking caps.

I was only thinking of removing the small .22nf caps since I just invested in the Mundorfs and I want to run the amp from different DAC's including the internal one from my laptop so DC blocking will be questionable probably. I want the system to be able to connect to any device.

When I remove the small .22nf caps. I am not supossed to bypass this position with copper why do I? That would put the Mundorfs out of play I guess.
 
Last edited:
You can also improve the PS with two simple mods:

Replace the LM317 voltage setting resistors with zeners (BZX85 13V)
Replace LM317 C ADj with 10uf 35V Silmics II Black Gold.

Why did you choose Elna Silmic II caps for the powersupply? They are known for being slow and muddy there.. I tried them for the power supply of my DAC and the sound became slow and muddy. Replaced them for Panasonic FR's and everything opened up again. They where quite big though.

Why do you use radial caps here anyway instead of the originals? Can you explain to me why you chose these changes?
 
Ordered some of these: http://www.vishay.com/docs/31025/erc.pdf

Couldn't find the RLR07 and PTF56 in the values I want.

The shop told me they have the same characteristics but the military grade thingy made them more true to the specs.

I am also trying to find me a pair of Mundorf Mlytics 4700uf. The shop that has them normally is out of stock on the 40V but he does have the 63V version in stock. They are 5mm higher and will maybe just fit. I will have to see if I can fit them in.

attachment.php


Next to that, when I look to the specs I see they differ. The 63V has the better specs so far I can understand the datasheet. But someone told me that bigger capacitors can make a powersupply slow sometimes. I am not sure if he meant capacity big or also the physical size within a product line. Others told me that the voltage is just an indicator how much the capacitor can take before it starts malfunctioning, and that you can always upgrade caps with the same capacity to a higher voltage spec if it fits. That size and price was the only factor to choose caps as small as possible..

Can someone enlighten me?
 

Attachments

  • mundorf_specs.png
    mundorf_specs.png
    44.7 KB · Views: 2,047
Last edited:
Upgraded my opamp to dual discrete OPA111VM on a high quality UFEEL adapter. UFEEL is based in Hong Kong and makes some good stuff. So I also ordered their UFEEL updater which is a small adapter that makes puts the opamps in Class A but not by the method of 2 simple resistors but a more sophisticated like described here I think: Biasing Op-Amps into Class A

attachment.php

attachment.php


Result:
The upgrade to 2 OPA111VM's was already groundbreaking. The sound openend up with about 30-40%. I suddenly felt I was in the room with the Nirvana MTV Unplugged album. Stereo image was WAY better and after a while of listening I forgot the headphones I was their sitting behind my laptop working. This was with both DT1770 and DT770 a huge improvement. With the new opamps the really stubborn DT770 250 ohm totally came alive almost to the level of the DT1770 before the opamp upgrade even the mids that were always on holiday came forward, ok highs are still to harsh for me but it became a really fun session. The DT1770 also improved a lot. It sounded a bit more open too and also more extended.

Then I tried the UFEEL updater and things even got a bit better. I pulled a veil of a bit. Lost a bit of warmth but got more control back for it. Everything is clear and tight now. Really nice addition in combination with the Aune T1 tube DAC with Amperex tube.

Whole opamp upgrade is probably the best upgrade I can do for a device like this.

I also put the Nippon Chemi-con LXZs back on advice of you guys. Not sure if I really like them as much as the Panasonics. The Panasonics are a little thinner but also are more detailed and open. The combination with the T1 Tube DAC compliments this by filling it with clean warmth. For now I will keep it this way because I first want to do the other upgrades step by step and maybe this works out very well together.

On my way are:
Vishay / Dale ERC55 resistors (couldn't find the desired values of RLR07 or PTF56)
Amtrans AMCH caps (bastards charged almost $10 per piece.. 100pf was hard to get)
Fairchild Zener Diodes 13V

and I am still deciding on Mundorf Caps. I can only find the 63V version for now which I've measured will just fit. I just wonder if they are not to big for the position and making my power supply slow or something.

Please comment on that.
 

Attachments

  • DSC05705_small.jpg
    DSC05705_small.jpg
    440.4 KB · Views: 2,036
  • DSC05700_small.jpg
    DSC05700_small.jpg
    482.7 KB · Views: 1,363
The RLR07 are not very well available here. But I can get the RLR20? It's 1/2w

No space for 1/2w resistors...

I was only thinking of removing the small .22nf caps since I just invested in the Mundorfs and I want to run the amp from different DAC's including the internal one from my laptop so DC blocking will be questionable probably. I want the system to be able to connect to any device.

Then you should stick with the input cap and its bypass: don't remove nothing.

Why did you choose Elna Silmic II caps for the powersupply? They are known for being slow and muddy there.. I tried them for the power supply of my DAC and the sound became slow and muddy. Replaced them for Panasonic FR's and everything opened up again. They where quite big though.

First of all you should distinguish between brown/white and black/gold Silmics II, they sound different.

Silmics are not muddy or slow, after some burn-in, instead I find them full, detailed and rich of harmonics that other caps masks.

Why do you use radial caps here anyway instead of the originals? Can you explain to me why you chose these changes?

According LM317's datasheet C Adj should be an high value (>= 10uF) capacitor, it's presence reduces ripple and noise, see:

Using 3-pin regulators off-piste: part 1

In the same article you'll find references about the zeners if I remeber correctly.

Ordered some of these: http://www.vishay.com/docs/31025/erc.pdf

Couldn't find the RLR07 and PTF56 in the values I want.

They should be pretty similar but never tried them.

I also put the Nippon Chemi-con LXZs back on advice of you guys. Not sure if I really like them as much as the Panasonics. The Panasonics are a little thinner but also are more detailed and open.

:confused: My impressions are quite the opposite...

Amtrans AMCH caps (bastards charged almost $10 per piece.. 100pf was hard to get)

10$ for AMCH?? :eek:

They're better but I wouldn't have spent so much on them... they costs only 2.30$ from HiFi Collective.

and I am still deciding on Mundorf Caps. I can only find the 63V version for now which I've measured will just fit. I just wonder if they are not to big for the position and making my power supply slow or something.

Don't bother about it for now... first try other mods for which you already bought parts.

If and when you decide to buy them use the 40V version, they're more than enough.
 
Thnx! I have stopped looking for the Mundorfs for now. Still finding out if the different specs of the 63V will have an influence for good or for worse on the audio signal in the end against the 40V.

Yeah the Amtrans were very expensive but I despised the ugly polystyrenes anyway!

And yes I cannot rule any else about the Panny and Nippon caps. Maybe they are fake? Not sure. I bought the Pannys from a respectable caps seller with test rapport. The LXZs came with the device. They're supposed to come from RS-Components.com. When I look on taobao.com I can buy LXZ's for about 0.35 yuan.. thats about 6 or 7 cents.. so could fakes?

And regarding the brown and gold black Elnas. I have both in 470uf 25V. Always used the black gold ones. They are manufactured last year while the browns are from 2012 or 2013 or something.
 
And yes I cannot rule any else about the Panny and Nippon caps. Maybe they are fake? Not sure. I bought the Pannys from a respectable caps seller with test rapport. The LXZs came with the device. They're supposed to come from RS-Components.com. When I look on taobao.com I can buy LXZ's for about 0.35 yuan.. thats about 6 or 7 cents.. so could fakes?

0.35 sounds about right for the Taobao price of the real deal. Fakes are normally cheaper than that and Nichicon is more often faked than NCC. They might not be particularly new though.
 
0.35 sounds about right for the Taobao price of the real deal. Fakes are normally cheaper than that and Nichicon is more often faked than NCC. They might not be particularly new though.

I ordered some ones I am pretty sure of they are real. The really cheap ones are shiny while the little bit more expensive ones are non-glossy. Also the cheap ones have a strange code on them where the originals have the build year and week. 2012 and week 26.

What is the life expectancy of a capacitor like this (long life brandend)?
 
Cap lifetime depends on the temperature its at, the warmer they're kept the quicker the electrolyte dries out. Cap lifetime roughly doubles for every 10oC reduction in temperature. So if the cap has a 3000hr/105oC rating and its kept below 45oC that's 192khrs or almost 22years.
 
Thank you for that info.

But I am quite confused about the LXZ caps. The ones that came with the machine are non-glossy. And the one that are on rs-components.com for example are glossy.. different types? or is it a fake after all? Can barely imagine why you would fake something like that in a machine but maybe they weren't aware.. I ordered new ones from 4 different shop with prices varying from 0.40 yuan to 2.5 yuan per piece. I also ordered 25v and 35v versions of the 470uf capacity.

Also ordered small 10uf 35V Cerafine caps. the seem more suitable for the power supply than the ELNA SILMIC IIs that were advised here. Cerafines are definitly faster than the Silmics II for the power supply. Ordered some SILMIC IIs too ofcourse hehe. To test the difference.

Next to that I am still looking to improve the signal path combination of the 470uf caps with the 22nf caps in the middle. I was told that the capacity difference is pretty big and this could cause troubles but I am not sure what to replace them with. The OSCONS are not good? I can order those. Check them out.. But prefer a good option that is a solid choice to begin with.
 
Also I keep bumping into people saying that using the smallest voltage value with your caps is preferable because this will give the fastest sound in the end.

The transformer is 2x 15v. Most caps that are chosen by default are 25V. I see ClaveFremen uses the 35V variant of the LZX caps. I would like to know why to choose 35V over the smaller 25V? I ordered both.
 
Higher voltage caps normally (for the same capacitance) have lower ESR. In this case the LZX 25V, 470uF has 68mohm and the 35V, 52mohm. The 35V cap is also taller, 20mm, vs 16mm but has the same 10mm diameter.

<edit> The Samyoung NXA 470uF, 50V has even lower ESR at 43mohm but is 30mm tall.
 
Last edited:
I know about the lower ESR but some people told me I shouldn't stare blind on only low ESR. If I would do that I would intermediately put my Panasonic FR's back with their 28mohm... and 8000 hours vs 5000 hours endurance from the LXZ (I find the Panasonic FR's sounding better than the questionable LXZs I have soldered in now) But people here disagree and say the LXZs are supposedly better so I tried them with negative result so far.

So I am a bit confused now.

To make a fair comparison I ordered some 35V LXZs to compare directly to the 25V LXZ and 35V Panasonic FRs.

I also have to wait to get my Mad Dog Pro's back this week to use them to do the comparison with the DT1770 and DT770. I normally use the Mad Dogs for daily listening and for more analytical music I pick the DT1770 and sometimes for fun I do a round of EDM on the DT770's. So the sound mostly needs to be tuned for the Mad Dogs since I use them for everything including gaming and watching series.
 
Last edited:
of course. I will not cling to a less sounding device because people that haven't heard the system advise differently because it did work out for them.

This is a forum for sharing experiences. And if they never told me that the LXZs were possibly the better choice I would have never put them back for a listening session to find out which ones I prefer. The next time I am doing a project I probably know which capacitors to pick a lot quicker from the start after comparing these.. thats the whole point of being here.. learning.
 
It's funny how little changes can make a big impact on the sound. At least to my ears.

I changed the LCR polystyrenes for the Amtrans AMCHs (which I bought way to expensive!) BUT the effect is quite there. Lifted a bit of the veil and made everything a bit more detailed some way. Maybe because also the sound is a little bit more harsh. Does this small 100pf cap need some burn in time?

I mostly use the Nirvana MTV unplugged album which has a wide dynamic range and a lot of details in it. The clapping seems a bit more harsh with the Amtrans, but the voice seems less distant and more natural. So thats a good thing I guess. Guitars have a little bit more metallic sound.. hope this will smoothen out a bit The Amtrans countered the change to the LXZ caps a bit I guess by lifting a bit of that veil again and restoring soundstage a bit. I also use my active studio monitors to compare the pre-amp outs. So I make sure that the modifications I do are well suited for all my attached devices.

  • Next mod will be replacing the resistors around opamp followed by listening session with all my headphones with the same album.
  • Mod after that will be replacing the resistors of the power supply with the Vishay/Dales and the other two with the Zener diodes, also placing the 10uf caps for the LM317 C ADj. Whatever that means. Will do A/B comparison with cerafines and silmics on this position.
  • Next mod after that will be replacing the LXZs I have with LXZs that I ordered (25V and 35V) Do A/B comparison also with Panasonics FR.
  • Possible next mod after that will be comparing the Panasonic Pureisms I use right now with some smaller ELNA Silmics and some smaller Pureisms.
 
Thnx! I have stopped looking for the Mundorfs for now. Still finding out if the different specs of the 63V will have an influence for good or for worse on the audio signal in the end against the 40V.

Usually capacitors with higher voltage rating sounds better and have less distortion and lower dissipation factor.

If space and cost are not a problem the higher voltage rating caps have to be preferred but in this case I would stick with the 40V ones.

But I am quite confused about the LXZ caps. The ones that came with the machine are non-glossy. And the one that are on rs-components.com for example are glossy.. different types? or is it a fake after all?

Mine are all glossy (bought from Mouser and RS), see pictures to compare yours:
attachment.php


attachment.php



Also ordered small 10uf 35V Cerafine caps. the seem more suitable for the power supply than the ELNA SILMIC IIs that were advised here. Cerafines are definitly faster than the Silmics II for the power supply. Ordered some SILMIC IIs too ofcourse hehe. To test the difference.

Cerafines are good too, 10/35 are a bit too much low IMHO...

I suggest you to mount sockets on components positions under test so you can swap them quickly and avoid destroing pads with soldering/desoldering work.

SIL sockets are perfect for the job.

Also I keep bumping into people saying that using the smallest voltage value with your caps is preferable because this will give the fastest sound in the end.

This is nonsense, the higher voltage versions are always better.

The transformer is 2x 15v. Most caps that are chosen by default are 25V. I see ClaveFremen uses the 35V variant of the LZX caps. I would like to know why to choose 35V over the smaller 25V? I ordered both.

Same reason, higher voltage the better and they are same diameter and with similar cost.

(I find the Panasonic FR's sounding better than the questionable LXZs I have soldered in now) But people here disagree and say the LXZs are supposedly better so I tried them with negative result so far.

You should decide what is your goal:


  • a pleasant sound
  • a transparent sound
Without a control to compare to it's not trivial to distinguish but in this case you can use the elco capacitors decoupling the opamp.


With OPA2134 they're not strictly needed so you can omit them to have a control.


Which cap sound less distant from no cap?


When I've done the same test I've compared:



  • Silmics II
  • Cerafine
  • FR
  • FM
  • LZX
  • FW
  • FG
To my ears the most trasparent are the LZX, nevertheless I've choosen the Silmics II since the amp needed a fuller sound that Silmics II give.


The FR sounded full, pleasant but with rolled off highs, restrictred soundstage and lower harmonic content Vs the control.


If you replicate my test setup (you'll need only the SIL sockets) I will read with interest your results.


In general where current is needed (like in the diamond buffer of the lehmann) a low ESR is preferred, where there is voltage gain (like for the opamp of the lehmann) a higher ESR type usually sounds better, IMHO.

I changed the LCR polystyrenes for the Amtrans AMCHs (which I bought way to expensive!) BUT the effect is quite there. Lifted a bit of the veil and made everything a bit more detailed some way. Maybe because also the sound is a little bit more harsh. Does this small 100pf cap need some burn in time?

Yes they smooth a bit with use but possibly you have mounted them with the outer foil in the non optimal position.

For Amtrans the direction determined by markings should go to ground (as in this case) or following the signal path.
 

Attachments

  • IMG_20160121_010703.jpg
    IMG_20160121_010703.jpg
    140.4 KB · Views: 1,683
  • IMG_20160121_005702.jpg
    IMG_20160121_005702.jpg
    133 KB · Views: 1,688
Is see. The LXZs I have are definitely different. They are non-glossy. So I ordered the glossy ones and I will give them a run.

I already saw the test sockets you used and I was already looking for them on Taobao.com. Thnx.

What do you mean about the Am trans? I used the pin holes in the board closest together.. should I used the ones furthest away? and what about direct? I didn't see direction markings on them. Just some white letters on one side.


Cerafines are good too, 10/35 are a bit too much low IMHO...
What do you mean to low? you said to use 10/35V instead of the Wimas next to the LM317
 
Last edited:
I replaced the red wima caps at the LM317 position with the small elna's.. Seemed together with 2 Zener diodes..

Seemed to have introduced a lot of hissing in every song.. the background is now full of hissing. Is this the elna caps that need to run in?

The sound definitely didn't improve any other way except for annoying hissing.. When I pause the music it's gone but its very loud when playing songs..

I can remember I had this too when using the Elna Silmic IIs in the Aune T1.. but I think it disappeared after some run in time I gues.. when I think about it... I will let time tell me. Let the machine play for a few days an keeping it on.

Also the sound is to harsh for me now after replacing with the Amtrans caps. The LCR's were somwhat softer.. It's difficult to compare now. I should make test sockets so I can compare more directly.

I must say that the funny thing is that the DT770 totally opened up and became a more pleasant headset of most of the mods but the Mad Dog is not really profiting yet. When I picked it up today from a friend and tried it again it sounded to harsh. The DT1770 stayed mostly neutral but everything just a bit better with most mods.

Since I'm upgrading quite fast, I can not compare very well so I think I should give components more time burning in but the longer the burn in takes the further away my impressions of the previous setup are haha.

Also Since I gave my Mad Dog headphones to a friend for borrowing I have grown accustomed a bit to the DT1770 and DT770 sound. I actually did a lot comparing with the DT770 headphones since this was the more flawed set of cans anyway so there was more to improve. Maybe met hearing adjusted a bit more to them than the Mad Dog, but right now they are harsh and less dynamic. I need to tune this harshness down.
 
Last edited: