Hi,
I have two SqueezeBox Duett used as streamer only.
We know some of the simpliest tweaks popularysed with photographs by Lampizator from Poland like new caps, new powersupply or added lamp stage for the dac chip. Some others went further with hard tweaks : swap of the XO, chips, I2S mode, new LDO regulator.
I never wanted to try the hard way because of the rarety of the Duett. IMHO one of the best light tweak is the added cap of J9 in the pcb near the Xilinx chip and one of the antenna. each cap or bypass change the sound.
I spent many many time tweaking and try combination. The SEPC cap of Lampizator is fun, try too the red FP cap with less ESR, MKP, ELna, Black Gate...near hundred parts alone or in // : electrolytic, mylar, polystyrene, ceramic...all what I have and i have many !
For the one who have the chance to have genuine BlackGate BGNX bipolarized, try it with a panasonic SEPC (the purple), not a Nichicon FC (the red) which is better alone than the SEPC. Here red BG and purple SEPC give me a full color 😀.... The mate sound of BlagGate without the up harcheness. Just try is you use the Duett in streamer mode ... or with its Dac chip too ! Outstanding softness with details and dynamic : rare! Its good enough for me after o,ne month of listening to write to you this little tip ...but great improvement ! It's an another machine !
The value I use are for the Blackgate BGNX : 33 uF/16 V and for the SEPC : 100 uF /25V. Be carefull with the polarity : look at at the back of the pcb if it's a first tweak with J9 printing on the pcb .
On the potographs, the good mix of the two box is the one with the piggy back genuine red Black Gate on the SEPC : as you see there are twoo legs for swap rapidly the caps. It was for easier test, but it's better like that in relation to final soldering... strange ?! No scope but my ears !🙄. I put yellow blue tack to fix the caps.
The second box is two 390 uf/50V expensive Nichicon metalic can FP cap.
I have two SqueezeBox Duett used as streamer only.
We know some of the simpliest tweaks popularysed with photographs by Lampizator from Poland like new caps, new powersupply or added lamp stage for the dac chip. Some others went further with hard tweaks : swap of the XO, chips, I2S mode, new LDO regulator.
I never wanted to try the hard way because of the rarety of the Duett. IMHO one of the best light tweak is the added cap of J9 in the pcb near the Xilinx chip and one of the antenna. each cap or bypass change the sound.
I spent many many time tweaking and try combination. The SEPC cap of Lampizator is fun, try too the red FP cap with less ESR, MKP, ELna, Black Gate...near hundred parts alone or in // : electrolytic, mylar, polystyrene, ceramic...all what I have and i have many !
For the one who have the chance to have genuine BlackGate BGNX bipolarized, try it with a panasonic SEPC (the purple), not a Nichicon FC (the red) which is better alone than the SEPC. Here red BG and purple SEPC give me a full color 😀.... The mate sound of BlagGate without the up harcheness. Just try is you use the Duett in streamer mode ... or with its Dac chip too ! Outstanding softness with details and dynamic : rare! Its good enough for me after o,ne month of listening to write to you this little tip ...but great improvement ! It's an another machine !
The value I use are for the Blackgate BGNX : 33 uF/16 V and for the SEPC : 100 uF /25V. Be carefull with the polarity : look at at the back of the pcb if it's a first tweak with J9 printing on the pcb .
On the potographs, the good mix of the two box is the one with the piggy back genuine red Black Gate on the SEPC : as you see there are twoo legs for swap rapidly the caps. It was for easier test, but it's better like that in relation to final soldering... strange ?! No scope but my ears !🙄. I put yellow blue tack to fix the caps.
The second box is two 390 uf/50V expensive Nichicon metalic can FP cap.
Attachments
Last edited:
NoVoxonline not my philosophy here , this is just a post for a simple tip for streaming mode: Easy, no expensive.... and great in relation to that and in the absolute.
I try many powersupply, and the changes was few. The original is not so bad and look at your Duet : there is a resistor after the plug and before it a long ohmic line (just 1 ohm) then The 200 uf metalic can cap... then ceramic caps : that's a Pi filter. Without changing these my humble opinion is that changing the main powersupply has no sense. I always go back with the nervvous little original powersupply.
All the lampizatir tweaks are not good : for example the spidf tweak is worse than the original smd caps for me : tested.
This tip can be a begginning for choosing between dangerous picked I2S inside the Duett and Spidf mode that is very transparent, smooth, dynamic and details with these two caps... no harcheness heared, free enough of jitter maybe !
I try many powersupply, and the changes was few. The original is not so bad and look at your Duet : there is a resistor after the plug and before it a long ohmic line (just 1 ohm) then The 200 uf metalic can cap... then ceramic caps : that's a Pi filter. Without changing these my humble opinion is that changing the main powersupply has no sense. I always go back with the nervvous little original powersupply.
All the lampizatir tweaks are not good : for example the spidf tweak is worse than the original smd caps for me : tested.
This tip can be a begginning for choosing between dangerous picked I2S inside the Duett and Spidf mode that is very transparent, smooth, dynamic and details with these two caps... no harcheness heared, free enough of jitter maybe !
Last edited:
Yes I know and hesitate to post. But it doesn't work so good with another combinations and approximation with this tip give nothing !
But if some o have the chance two have both the Duett and a genuine BGN or GNX BlackGate he just has to try (no problem with sepc at Mouser for example) ! After all they are both industrials goods !
But believe, tried so many, so many type, old, new, orange drop, oil, mkc, mkt, mkp SN, old and new elna, Muse, big esr from 70's cap gave me incredible result... sometimes it's just not only an low esr question !!! But here don't no why the result is great enough for posting. Understand it's maybe a tonal balance with may entire system but that's cost nothing to try
But if some o have the chance two have both the Duett and a genuine BGN or GNX BlackGate he just has to try (no problem with sepc at Mouser for example) ! After all they are both industrials goods !
But believe, tried so many, so many type, old, new, orange drop, oil, mkc, mkt, mkp SN, old and new elna, Muse, big esr from 70's cap gave me incredible result... sometimes it's just not only an low esr question !!! But here don't no why the result is great enough for posting. Understand it's maybe a tonal balance with may entire system but that's cost nothing to try
In streaming mode...
I'm not said that changing the micro ldo (like the SUBBU 😉) and crystals give nothing, just say my tip is great and safe !
played a lot with powersupply and bypassing the main 220 uf caps : always come back !
I'm not said that changing the micro ldo (like the SUBBU 😉) and crystals give nothing, just say my tip is great and safe !
played a lot with powersupply and bypassing the main 220 uf caps : always come back !
I've got a Squeezebox Duet that I've done some mods on. I agree with Eldam that the SPDIF mods from Lampizator don't make sense but his other suggestions on adding capacitance around the various regulators does seem to work for me.
Eldam - can you tell us which power supply is affected by your mod? Is this local bypassing for the Xilinx chip?
Thanks,
---Gary
Eldam - can you tell us which power supply is affected by your mod? Is this local bypassing for the Xilinx chip?
Thanks,
---Gary
I have no shematics, but Lampi wrote this is for the Xilinx (think he read himself from john Swenton... not sure for the name). maybe its bypass is for anothers chips too, just don't know. The changes are incredible, checked by ears.
In fact, these tip give far more result than playing with the three power caps of my AD1865 DAC (feeded by the Duett)...
In fact, these tip give far more result than playing with the three power caps of my AD1865 DAC (feeded by the Duett)...
Interesting. So far there is no consensus of duet mods whatsoever as I remember from different forums
Yes, the harder is not be slave of your own discovery.
Here I gave the tip, because I think the combination is a big improvement...but maybe just because of the tonal balance... But my amp and speakers are very resolving.
As the improvement concern the SPIDF quality, I write because one year ago I hesitate to go with I2S peaking before the 3 resistors of the XILINK on the way of the embeded DAC to work with an external DAC chip (TDA1541)... I gave up because no result but no fried the Xilinx. I will be curious to test two DAC feeded by the duett one with shorter picked I2S and the other with my tip and the spidf and a good spidf input like the one of the Subbu (WM8804 in hardware mod with FOX crystal).
Note that in my DAC there is no cap or coil before the SPIDF input chip. I just use the one for DC blocking of the Duett. I make the same with the Subbu and avoid the 100 nf cap (double DC blocker is a duet non-sense)
Here I gave the tip, because I think the combination is a big improvement...but maybe just because of the tonal balance... But my amp and speakers are very resolving.
As the improvement concern the SPIDF quality, I write because one year ago I hesitate to go with I2S peaking before the 3 resistors of the XILINK on the way of the embeded DAC to work with an external DAC chip (TDA1541)... I gave up because no result but no fried the Xilinx. I will be curious to test two DAC feeded by the duett one with shorter picked I2S and the other with my tip and the spidf and a good spidf input like the one of the Subbu (WM8804 in hardware mod with FOX crystal).
Note that in my DAC there is no cap or coil before the SPIDF input chip. I just use the one for DC blocking of the Duett. I make the same with the Subbu and avoid the 100 nf cap (double DC blocker is a duet non-sense)
Last edited:
My theory is that paralleling capacitors helps fix unwanted resonant behavior between various capacitors in the power supply. Below I've included a figure from page 25 of this document http://www.chemi-con.co.jp/e/catalog/pdf/Application_Note_NPCAP_090716e.pdf
It shows that a low ESR cap like the polymer caps that are used here can resonate with small ceramic caps. It suggests that using 1uf ceramic caps with polymer caps isn't a good idea because of this effect. Increasing the size of the ceramic cap to 10uf seems to greatly reduce this effect. So I think that bypassing the polymer cap with a 10uf ceramic might be worth trying and perhaps as effective as the BG bypass.
I just installed a 220uf Nichicon FP cap in the J9 position and bypassed it with a 10uf capacitor. I'll see how it sounds compared to the previous mods.
---Gary
p.s. I measured the voltage and this is a 3.3v supply at the J9 pins used here. So it probably is the Xilinx power supply.
Attachments
Yes, I'm sure it acts as a filter...
Yes the voltage is printed at the back of the PCB, sure it's for the Xilinq but maybe flow some others chips ?
I tried small and highs values, alone and together, ceramics too and all types like military BC class.... Nothing give the same results as these two caps. Ceramic, even dry and polarized dielectric are good for the scope and often bad to the ears (of course I talk about numeric, elswhere it's a no go in analog domain)
With sounds signature words : Blackgate mate sound signature but without its harchness in the high (alone). The SEPC add its signature but remove the harchness and add dynamics without avoid softness... with clarity in the mid and bass. It's no science, I found it by chance and hard work (many many hours).
Yes the voltage is printed at the back of the PCB, sure it's for the Xilinq but maybe flow some others chips ?
I tried small and highs values, alone and together, ceramics too and all types like military BC class.... Nothing give the same results as these two caps. Ceramic, even dry and polarized dielectric are good for the scope and often bad to the ears (of course I talk about numeric, elswhere it's a no go in analog domain)
With sounds signature words : Blackgate mate sound signature but without its harchness in the high (alone). The SEPC add its signature but remove the harchness and add dynamics without avoid softness... with clarity in the mid and bass. It's no science, I found it by chance and hard work (many many hours).
I am sure that this combination sounds good. It's too bad that you are so far away. It would be fun to hear your version to get a sense of what you are hearing.
Regards,
---Gary
Yes, teleportation à la Star Trek would be a waited next revolution ! Had a childhood french friend in Sillicon valley who works at chip design development as Ingeener and never went to saw her for 20 years!
I have a poor hifi in relation to some here, but i really think that a hifi is like a Formula one and a circuit (Hifi + room). You have to set up.
Speakers (have three but the main : ) Boston Acoustic Lynfield 400L wit a narow passive notch (after test and opinions of DIYAUDIO fellows and self developer Phil Jones opinion by mail... very busy man as PURE Sound CEO... thanks Phil for your Kindness!)
Amp: Chord SPM 1000B
CD player : a Philips with Lundhal transformer
planar : Rega 3
wires : signal : Oyaide, speakers : Cardas
DAC : Raindrop Hui AD 1865 but very tweaked
Synology Nas for the SB Duetts streamer
More or less ten tweaked cd players for pleasure, experiment of bad beginner in diying 🙂 and collection as fanatic myself !
When have a doubt : swap with beloved Kef 104/2 ref or Proac D15.
Gary, did you hear the difference with one or two same FP caps ? : the result is totaly different !
I have a poor hifi in relation to some here, but i really think that a hifi is like a Formula one and a circuit (Hifi + room). You have to set up.
Speakers (have three but the main : ) Boston Acoustic Lynfield 400L wit a narow passive notch (after test and opinions of DIYAUDIO fellows and self developer Phil Jones opinion by mail... very busy man as PURE Sound CEO... thanks Phil for your Kindness!)
Amp: Chord SPM 1000B
CD player : a Philips with Lundhal transformer
planar : Rega 3
wires : signal : Oyaide, speakers : Cardas
DAC : Raindrop Hui AD 1865 but very tweaked
Synology Nas for the SB Duetts streamer
More or less ten tweaked cd players for pleasure, experiment of bad beginner in diying 🙂 and collection as fanatic myself !
When have a doubt : swap with beloved Kef 104/2 ref or Proac D15.
Gary, did you hear the difference with one or two same FP caps ? : the result is totaly different !
Last edited:
I agree that tweaking and tuning the system in one's room will give much bigger benefit compared to just throwing money at the system. Being able to build and tweak and measure one's components gives us a huge advantage over those who can only write a check.
Your system has some similarities to mine but also some big differences. I have 2 stereos - one in the living room and one in the family room. The one in the family room gets listened to more often. Somehow in the U.S., living rooms have become formal rooms that don't get used very often.
I also use NAS for my music - a QNAP NAS
I stream using an SB Duet right now but also own a Touch and an SB3 and an SB2. With a bit of tweaking, they all make very good streamers.
Speakers (family room) - Reference 3A mm de capo i
6V6 push pull integrated amp
DAC - Twisted Pear Buffalo III with transformer output stage and FET buffer
No phono player in this system
I'm not sure I understand the question. Can you ask it a different way. Which 2 FP caps are you asking about?
---Gary
My fault, have to read some books of english grammar🙁.
In fact with the second picture which is a poor alternativ to the main tip you can see two metalic FP can caps in //. My question was : did you try with one FP cap then with two in // and notice difference between the two listening ?
I did, it's totaly different on my system. With one : soft but the sound is behind the speakers with a big lake of dynamic, maybe a dip in the mid bass: lake of softness is a no-go.
With two // (2 x 390 uf/50 v) : the sound signature is different : voice are in front of the speaker, it's very clearer and more dynamic in a good way but the main quality (softness) diseapper. In my system which is very clear and transparent with aluminium cone for mid and tweeter (125 hz to 2600 hz then tweeter).
I don't know how a designer do with a scope to measure that, I can't and don't want to learn it but in my system : with a cap or other cap all change : width, height, stage... strange.
As there is no universal system I can understand why the measurement is only way : not too much time for audio tweaking details or changing suppliers for caps, etc... but sometimes it can be great !
I'm waiting parts from Mouser , 2O uf sepc caps and ldo and crystal for the subbu DAC. I will try with this DAC if the tip is always ok !
In fact with the second picture which is a poor alternativ to the main tip you can see two metalic FP can caps in //. My question was : did you try with one FP cap then with two in // and notice difference between the two listening ?
I did, it's totaly different on my system. With one : soft but the sound is behind the speakers with a big lake of dynamic, maybe a dip in the mid bass: lake of softness is a no-go.
With two // (2 x 390 uf/50 v) : the sound signature is different : voice are in front of the speaker, it's very clearer and more dynamic in a good way but the main quality (softness) diseapper. In my system which is very clear and transparent with aluminium cone for mid and tweeter (125 hz to 2600 hz then tweeter).
I don't know how a designer do with a scope to measure that, I can't and don't want to learn it but in my system : with a cap or other cap all change : width, height, stage... strange.
As there is no universal system I can understand why the measurement is only way : not too much time for audio tweaking details or changing suppliers for caps, etc... but sometimes it can be great !
I'm waiting parts from Mouser , 2O uf sepc caps and ldo and crystal for the subbu DAC. I will try with this DAC if the tip is always ok !
Last edited:
Hi, thanks for posting your findings.
I too am planning on using Subbu DAC with a Squeezebox Duet. I was wondering just yesterday about best way to feed the SPDIF. I had seen the Lamp’s suggestions about a more direct SPDIF route, but note that you don’t approve.
Please could you describe more about where you take the SPDIF signal from on the Duet and any modifications suggested on the Subbu DAC board?
I will be installing the Duet in the same chassis as the DAC so would be particularly easy to wire direct, but people often talk of the impedance of RCA/BNC plugs for SPDIF, I was wondering if any changes are needed if a coax or twisted pair is soldered directly from Duet to Subbu without any plugs?
I too am planning on using Subbu DAC with a Squeezebox Duet. I was wondering just yesterday about best way to feed the SPDIF. I had seen the Lamp’s suggestions about a more direct SPDIF route, but note that you don’t approve.
Please could you describe more about where you take the SPDIF signal from on the Duet and any modifications suggested on the Subbu DAC board?
I will be installing the Duet in the same chassis as the DAC so would be particularly easy to wire direct, but people often talk of the impedance of RCA/BNC plugs for SPDIF, I was wondering if any changes are needed if a coax or twisted pair is soldered directly from Duet to Subbu without any plugs?
OK - I understand now. I did not try two polymer caps in parallel. After your first post, I did a quick installation without any experimentation. I took a 220uf Nichicon polymer cap and paralleled it with a 10uf MLCC ceramic cap and installed that in J9. My Duet already sounds much better than stock because of my previous mods and I wanted to see if this change made an improvement or at least a change. My experience has been that the caps need to run for at least 24 hours before making any judgment. So far the change seems very subtle if it exists at all, but that's because it already sounded good with the other changes.
Yes, you can wire directly from the Duet to the Subbu V3 DAC. Just remove the SPDIF RCA connector and use a good 75 ohm coax to make the connection. I like RG174, a 75 ohm miniature coax cable but any 75 ohm cable would be OK. In my Duet, I also added an SPDIF transformer since I think it's a good idea to isolate grounds between the digital source and the DAC. I used a Newava S22083, digikey # 470-1003-ND. I also increase the size of the blocking cap feeding the SPDIF transformer.
---Gary
Garry,
Do you use a FP cap (red mark for negative) ? did you try without the MLCC (don't like it...). The blue nichicon polymer sound awful here in my system, but same value with FP or SEPC are far better than the blue nichicon polymer.
Bud,
I know that lampizator use lamps only after the dac chips (in fact after the coil, the resistor and the caps in serie after the dac chip...). No difference for me between rca plugs and soldering just on the legs of it... for my ears I mean (have no scope).
My Subbu DAC is not working because of a fried LDO and maybe Fox crystal (Q2)... I'm wating spares from Mouser... very long because added in my BOM some 20 uf SEPC caps out of stock
Do you use a FP cap (red mark for negative) ? did you try without the MLCC (don't like it...). The blue nichicon polymer sound awful here in my system, but same value with FP or SEPC are far better than the blue nichicon polymer.
Bud,
I know that lampizator use lamps only after the dac chips (in fact after the coil, the resistor and the caps in serie after the dac chip...). No difference for me between rca plugs and soldering just on the legs of it... for my ears I mean (have no scope).
My Subbu DAC is not working because of a fried LDO and maybe Fox crystal (Q2)... I'm wating spares from Mouser... very long because added in my BOM some 20 uf SEPC caps out of stock
Last edited:
- Status
- Not open for further replies.