Hi Peter,
Just ran through all the postings and photos in this thread and must say you are wonderful. Other than electronics, it looks to me that you are also a mechanic!
You are right that someone bought extra 38S upgrades mainboards from Zenn and thats me. I brough 4 of them while I was in Singapore. I replaced my original 38 and find that the 38S is having a quieter background with more details. Yet, I don't like the the high was not than bright (although I am already using a pair of JMLab Micro Utopia Be). Your last advice in my thread regarding the opamp in 380S makes me to explore my 38S again. I've mounted IC sockets onto the PCB and tried a number of opamp combinations in these days. The IC I used are already fully burnt in since I have been using them for other projects. They are: -
OPA2132 - The high is brighter but the low is being sacrified.
OPA2134 - More balanced but I feel the high is not enough.
AD8620 - Too bright
OPA627 + OPA2134 + AD823 - The AD823 are new and needs time to burn in. However, this combination seems to be very smooth to me and have the balance even at this stage. I have ordered some 2604 from RS and will replace the 2134 when I received them. (The original 2604 on the mother board were being "killed" by me when de-soldering ).
As regarding the remaining 38S boards, I have already replaced 2 of my friend's 38 to 38S and the last one will be reserved for my set if I go anything wrong when upgrading my current one. There will be no 38S left behind but I am now having 3 pcs of 38 mainboard sitting with me. I perhaps will try to build 3 x 38 at later days.
Oh yes, I want to mention that there is a filter circuit right after the IEC socket and before the mainboard. The older version of 38 is having a Mexico made filter built in the IEC socket while the newer version (with a higher serial number) is having a small piggy back PCB after the IEC socket. My digital cam is not with me right now and will post the piggy back in later days.
I saw one of your posting saying that you have changed the digital and analog rectifier. I can see there are two bridges on the two sides near the transformer and I believe they are both analog. Where is the digital bridge? I also want to give it a try.
By the way, thank you for your posting of 380S motherboard layout. Where did you get this? Is there any empty PCB available in the market?
Thanks regards,
Sunny
Just ran through all the postings and photos in this thread and must say you are wonderful. Other than electronics, it looks to me that you are also a mechanic!
You are right that someone bought extra 38S upgrades mainboards from Zenn and thats me. I brough 4 of them while I was in Singapore. I replaced my original 38 and find that the 38S is having a quieter background with more details. Yet, I don't like the the high was not than bright (although I am already using a pair of JMLab Micro Utopia Be). Your last advice in my thread regarding the opamp in 380S makes me to explore my 38S again. I've mounted IC sockets onto the PCB and tried a number of opamp combinations in these days. The IC I used are already fully burnt in since I have been using them for other projects. They are: -
OPA2132 - The high is brighter but the low is being sacrified.
OPA2134 - More balanced but I feel the high is not enough.
AD8620 - Too bright
OPA627 + OPA2134 + AD823 - The AD823 are new and needs time to burn in. However, this combination seems to be very smooth to me and have the balance even at this stage. I have ordered some 2604 from RS and will replace the 2134 when I received them. (The original 2604 on the mother board were being "killed" by me when de-soldering ).
As regarding the remaining 38S boards, I have already replaced 2 of my friend's 38 to 38S and the last one will be reserved for my set if I go anything wrong when upgrading my current one. There will be no 38S left behind but I am now having 3 pcs of 38 mainboard sitting with me. I perhaps will try to build 3 x 38 at later days.
Oh yes, I want to mention that there is a filter circuit right after the IEC socket and before the mainboard. The older version of 38 is having a Mexico made filter built in the IEC socket while the newer version (with a higher serial number) is having a small piggy back PCB after the IEC socket. My digital cam is not with me right now and will post the piggy back in later days.
I saw one of your posting saying that you have changed the digital and analog rectifier. I can see there are two bridges on the two sides near the transformer and I believe they are both analog. Where is the digital bridge? I also want to give it a try.
By the way, thank you for your posting of 380S motherboard layout. Where did you get this? Is there any empty PCB available in the market?
Thanks regards,
Sunny
Hi Sunny,
The ML380S layout was posted by somebody from China, the link doesn't seem to work anymore, but I have saved the file.
Regarding the bridges, while the rect in digital section shouldn't have any effect on the sound, Madrigal still claims they have (maybe less pollution to analog section?).
I changed analog rect for MUR860 and rect in digital section for MSR860. Whenever I use those diodes I'm always getting great results.
So far the biggest improvement in sound (38S board) was achieved by changing regulators, diodes and main filter caps. I'm using BG STD 1000/50 at bridges and BG N 1000/50 right before regulators. I place some 4,7 BG N after regulators and they work very good (in place of regular 68u caps). I plan to put there 47U or 10u when I get them.
I reported previously that I didn't like AD823 for I/Vconversion after volume DAC. Recently I changed input op amps for OPA627 (in place of LT chips), and francly I am not impressed that much. It seems like resolution has improved, midrange is maybe slightly better too and more clarity overall, but somehow the magic I was experiencing before seem to be gone, I mean less musical presentation. I changed for AD823 (instead of OPA2604) and it seems like improvement this time, but I'm still if it's better when what I had originally. Please note that four 1k resistors (around I/V chip) should be replaced to 100ohm when switching to AD823.
I didn't notice any improvement when changing for BG411 switch, so I don't really recommend it. I also think that OPA2604 works very well in this preamp, and while new chips may have better specs, for some reason 2604 sounds very good here.
I also changed those green Wima bypass caps to blue Vishay/ROE and they bring further improvement.
Overall it is a very good preamp, well worth the effort.
While you bough 4 of those 38S boards, I bought 5 of them (sold as no functioning, however I managed to get 2 of them working). I also have 2 of the regular 38 boards
The ML380S layout was posted by somebody from China, the link doesn't seem to work anymore, but I have saved the file.
Regarding the bridges, while the rect in digital section shouldn't have any effect on the sound, Madrigal still claims they have (maybe less pollution to analog section?).
I changed analog rect for MUR860 and rect in digital section for MSR860. Whenever I use those diodes I'm always getting great results.
So far the biggest improvement in sound (38S board) was achieved by changing regulators, diodes and main filter caps. I'm using BG STD 1000/50 at bridges and BG N 1000/50 right before regulators. I place some 4,7 BG N after regulators and they work very good (in place of regular 68u caps). I plan to put there 47U or 10u when I get them.
I reported previously that I didn't like AD823 for I/Vconversion after volume DAC. Recently I changed input op amps for OPA627 (in place of LT chips), and francly I am not impressed that much. It seems like resolution has improved, midrange is maybe slightly better too and more clarity overall, but somehow the magic I was experiencing before seem to be gone, I mean less musical presentation. I changed for AD823 (instead of OPA2604) and it seems like improvement this time, but I'm still if it's better when what I had originally. Please note that four 1k resistors (around I/V chip) should be replaced to 100ohm when switching to AD823.
I didn't notice any improvement when changing for BG411 switch, so I don't really recommend it. I also think that OPA2604 works very well in this preamp, and while new chips may have better specs, for some reason 2604 sounds very good here.
I also changed those green Wima bypass caps to blue Vishay/ROE and they bring further improvement.
Overall it is a very good preamp, well worth the effort.
While you bough 4 of those 38S boards, I bought 5 of them (sold as no functioning, however I managed to get 2 of them working). I also have 2 of the regular 38 boards
Thank you Algar in pointing to me about the digital rectifier.
Hi Peter, do you mean you only get 2 of them working out of 5? Affirmatively, my last 38S mainboard may not be working! Touch wood.
Besides 2604, I have also ordered AD797 too. They will arrive in the next few days. I am having more fun with the combination of so many opamps.
Sunny
Hi Peter, do you mean you only get 2 of them working out of 5? Affirmatively, my last 38S mainboard may not be working! Touch wood.
Besides 2604, I have also ordered AD797 too. They will arrive in the next few days. I am having more fun with the combination of so many opamps.
Sunny
From some pics that I've seen of this pre, layout is not good around de op-amps.
Changing op-amps here may be very misleading, the best ones may actually sound worse...
The OPA2132 sounds (and IS) better than the OPA2604, but here it will sound bad.
Yes Elso, I also wonder what's so extraordinary about this pre?
Changing op-amps here may be very misleading, the best ones may actually sound worse...
The OPA2132 sounds (and IS) better than the OPA2604, but here it will sound bad.
Yes Elso, I also wonder what's so extraordinary about this pre?
carlosfm said:From some pics that I've seen of this pre, layout is not good around de op-amps.
Changing op-amps here may be very misleading, the best ones may actually sound worse...
The OPA2132 sounds (and IS) better than the OPA2604, but here it will sound bad.
Yes Elso, I also wonder what's so extraordinary about this pre?
Carlos
Is it OPA2132 that sounds better or architecture in which OPA2132 sound better than 2604. Right now I am burning AD823AN which has a different sound stage than OPA.
Trigon
Re: Schematic
Don't ask me as don't really know
When I got it in stock form (38s) it sounded so average that I wanted to abondon the whole project.
After some changes in PS, mostly diodes, caps and regulators, and further the caps on board after regulators, the sound improved dramatically. Chnging ICs makes actually for much smaller difference (almost negligable, comparing gains with passive parts).
Check here what the guy from empiricalaudio.com says about this preamp and the mods:
If I didn't do that myself, I wouldn't believe it. You might still not believe me though
I could provide simplified schematic, but not the whole version, as I was sworn to keep secrecy.
Elso Kwak said:
Any idea what makes the good sound? Or would that be the stinking normal regulators? Or the opamps?
carlosfm said:
Yes Elso, I also wonder what's so extraordinary about this pre?
Don't ask me as don't really know
When I got it in stock form (38s) it sounded so average that I wanted to abondon the whole project.
After some changes in PS, mostly diodes, caps and regulators, and further the caps on board after regulators, the sound improved dramatically. Chnging ICs makes actually for much smaller difference (almost negligable, comparing gains with passive parts).
Check here what the guy from empiricalaudio.com says about this preamp and the mods:
If I didn't do that myself, I wouldn't believe it. You might still not believe me though
I could provide simplified schematic, but not the whole version, as I was sworn to keep secrecy.
Sunsun22 said:Thank you Algar in pointing to me about the digital rectifier.
Hi Peter, do you mean you only get 2 of them working out of 5? Affirmatively, my last 38S mainboard may not be working! Touch wood.
Besides 2604, I have also ordered AD797 too. They will arrive in the next few days. I am having more fun with the combination of so many opamps.
Sunny
I will be also testing AD797 and AD8610. BTW, I'm not using any AC filters, they never worked in my house.
I was saying that the boards that I bought were sold as not working, I'm still happy that two of them worked right away
I don't see the other would be hard to fix.Regarding removing/desoldering parts from multilayer boards, unless you have proper equipment, it's probably best to cut off all the pins first, from ICs, and then remove each pin separately. I can sacrify few components, but it would be really upsetting to break the through hole plating on those multilayer boards.
Keep us posted about your listening impressions.
carlosfm said:
While this is only your opinion, the 380S version, comparing to 38s, does not change parts layout. Just the ICs are swapped (and some decoupling caps upgraded), and some people, includding John Atkison, couldn't believe how this can bring so much improvement to the sound. The more I change from original configuration, the more confusion I have if the changes are indeed for the better. While sound is maybe more high-end, it's also more mechanical and distant, without "flesh and blood" being present anymore
trigon said:
Everywhere I compare an OPA2132 with an OPA2604 the 2132 sounds better.
But the OPA2132 is more picky about the layout and the PSU bypassing.
Put everything right and the 2132 really sings.
In fact, some parameters of the 2604 are better, others are much worse.
THD is night and day comparing the 2604 to the 2132, check the datasheets.
Anyway, I really don't give so much importance to specs, the 2132 just sounds much better to me.
Yes, I do believe OPA2132 is better than OPA2604 in most situations. I've been using OPA2132 replacing OPA2604, NE555 etc. in a few of my hi fi gears and OPA2132 always being the better choice. However, this is not the case in the 38S since the sound becomes too bright.
2604 is an old design and has been in the market for many years. OPA2132 is newer and an instrument grade opamp with extended high frequency response. I think the circuit or layout in the 38S is the crucial element that makes 2604 sounds better (in my opinion and current setup) than 2132. I am not an electronic guru and is not able to understand why. My ears pick up the difference.
I've been using two other pre-amps before - Aragon 24 KPS and Krell KSL but the ML38S has the most details. I can hear many missing sounds that the former two were not able to reproduce. I'll sure give it a trial to modify other components to make it sounds better and in any case if I fail, I still have one 38S with me so that I can switch it back to ground zero.
Sunny
2604 is an old design and has been in the market for many years. OPA2132 is newer and an instrument grade opamp with extended high frequency response. I think the circuit or layout in the 38S is the crucial element that makes 2604 sounds better (in my opinion and current setup) than 2132. I am not an electronic guru and is not able to understand why. My ears pick up the difference.
I've been using two other pre-amps before - Aragon 24 KPS and Krell KSL but the ML38S has the most details. I can hear many missing sounds that the former two were not able to reproduce. I'll sure give it a trial to modify other components to make it sounds better and in any case if I fail, I still have one 38S with me so that I can switch it back to ground zero.
Sunny
As stated earlier, I am posting the picture of the filter circuit board of later ML38X herein.
I did not change the LT but instead, I changed the second 2604 into OPA627 and the third into AD797 while the 4th to 2134. After listening for a few days, I feel it is smooth but a bit bright in the high therefore I have changed the 4th opamp back to 2604 that I recently bought. I think it needs to sit in for a few more days before I can comment.
I have also changed the cap next to the bridge into RS's 3300 uF low ESR type. I can immediately feel the punch.
Peter, you mentioned that you changed the diode on the PCB, please adv what is the new part number of your diode? You also mentioned that you changed the decoupling caps to Vishay/ROE caps but I cannot find them in Hong Kong (not even with RS). Do you have a pic of what does they look like? Can you also tell me which part-number you have changed on the PCB?
I also wish to change the cap after the bridge into BG N/NX but they are very expensive in Hong Kong. Besides, their value is lower than 6800 uF. Do you think its worth to stick to BG N/NX or ELNA audio grade is comparable?
Sunny
An externally hosted image should be here but it was not working when we last tested it.
I did not change the LT but instead, I changed the second 2604 into OPA627 and the third into AD797 while the 4th to 2134. After listening for a few days, I feel it is smooth but a bit bright in the high therefore I have changed the 4th opamp back to 2604 that I recently bought. I think it needs to sit in for a few more days before I can comment.
An externally hosted image should be here but it was not working when we last tested it.
I have also changed the cap next to the bridge into RS's 3300 uF low ESR type. I can immediately feel the punch.
An externally hosted image should be here but it was not working when we last tested it.
Peter, you mentioned that you changed the diode on the PCB, please adv what is the new part number of your diode? You also mentioned that you changed the decoupling caps to Vishay/ROE caps but I cannot find them in Hong Kong (not even with RS). Do you have a pic of what does they look like? Can you also tell me which part-number you have changed on the PCB?
I also wish to change the cap after the bridge into BG N/NX but they are very expensive in Hong Kong. Besides, their value is lower than 6800 uF. Do you think its worth to stick to BG N/NX or ELNA audio grade is comparable?
Sunny
