[krisno]

So Jogi, you rather change the 3x1000uf(or whatever the size) from Rubicon to blackgates, and the 47ufx5 to blackgates as the first upgrade? Are these difficult to solder / desolder?

I allready changed alot of output caps.... still I have problem seing that chaning I/V resistor will improve sound when I actually pick up the sound signal to the Jupiter output caps on the input side of the resistor..

Thanks!

 

[Chris19]

Hey guys,

Just wanted to share the results of some experiments with you before I go off to work.

I've had a Valab since around January or February of this year, and I've enjoyed it, but I was getting a little bothered by some clipping on very loud tracks that I have. (I listen mainly to metal.) I had already tried lowering the I/V resistor down to about 300 ohms, but I couldn't get rid of all of the distortion. On top of that, I was getting a bit tired of the emphasized bass and harmonics on the lower end. It can really bloat the sound of some guitars.

Enter lampizator.eu and the FETishizator. I'm not a big fan of vacuum tubes, but his simple output stages seemed interesting, so I decided to give the FETishizator a try. Check it out here.

My theory is that the DAC chips are somewhat fighting each other when outputting the current in parallel. They need to make a voltage adjustment in order to flow current, and they're probably all competing and continually adjusting their output voltage based on what the other chips are doing. I'm guessing that's why the stock Valab gets so warm, and it's responsible for the harmonics too.

So, here's the short version of what I did:
1. Open up the case and remove all but 2 of the DAC chips. This is reversible; the chips are in sockets.

2. Lift the I/V resistors to steal the signal from the side closest to the DAC chips. There might be some DC bias there; I haven't measured it, but I put a cap inline just in case. The 300~400 ohm resistor in there right now is too small for proper operation; the TDA1543 datasheet shows a 1.2K I/V resistor, so you need to use 1K to 1.2K on the FETishizator board.

3. Now we need to steal power. On the bottom of the DAC, you can tap into the legs of the power supply filter caps to get the roughly 13.20 volts found there.

4. Build the circuit that either Lukasz drew on paper, or build the one that is in the PDF by Dietmar. Lukasz uses just the single JFET; Dietmar uses a source-follower MOSFET in addition. Lukasz's is just a simplified version of Dietmar's. Note: You don't need the resistor in series with the JFET gate after the I/V resistor; the JFET gate is in the megaohms, and the tiny resistor before it is like a drop in a lake in comparison.

If you follow Dietmar's JFET_CDOut schematic, skip his little power supply portion in the top right corner. Just use a single voltage regulator to provide for both channels; it's much simpler this way.

5. Connect back up to your output caps on the main Valab board, or use your own caps and wire straight to the RCA's on the case.

I can go into more detail when I get home, but the main advantage of this is to get rid of any remaining distortion and restore balance to the sound spectrum. There is some hiss from the TDA1543's, but I greatly prefer the Valab this way. Voices are amazingly clear in this configuration, and highs aren't drowned out anymore.

 

[Realrift]

Hi All,

Just wondering what DC voltage the Valab is supposed to run at. At the moment, I have connected the external power supply to the power supply caps and have switched the power supply to 9v DC input (i think it acutally supplies about 10V).

Cheers,
Rift.

 

[Ranchu]

Quote:

Originally Posted by Chris19 /img/forum/go_quote.gif I can go into more detail when I get home, but the main advantage of this is to get rid of any remaining distortion and restore balance to the sound spectrum. There is some hiss from the TDA1543's, but I greatly prefer the Valab this way. Voices are amazingly clear in this configuration, and highs aren't drowned out anymore.

Awesome, I'd love to hear it!

I was considering experimenting with different numbers of TDA's, is it as simple as pulling some out and changing the vref pot and i/v resistor?

And welcome, even though I'm a noob!

 

[Ranchu]

Realrift,

I think you need to look up the datasheet of the dac chip's regulator in your dac (they vary), and find the dropout voltage, usually 1 or 2 volts, add the 8V the Dac chips run on, and make sure you supply the reg more than that, and less than the 13.2V Chris19 mentions above.

Sounds like you're good.

That is my understanding, someone please correct me if it's bull.

 

[PatOMalley]

anybody try a tube buffer with this DAC yet?
some are trying it with multiple NOS DACs : Facelifted no oversampling dual Philips TDA1541A DAC and 6C45π tube outputbuffer

I think I will try the Yaqin tube buffer to see. CineMags are still on their way. I could A/B them.

 

[Chris19]

Ranchu, I think you have it right on the power recommendation. On my board, the v-reg for the digital chips outputs 4.99 volts, and the TDA power supply v-reg is 7.8 volts. Should be no problem at 10, maybe 10.5 volts. More will just make the regulators run hotter, but seeing as the stock transformer is making 13.2, it seems they can take it.

When I yanked out my chips, I didn't change the v-ref pot. Actually, I'm not sure if mine has the v-ref pot. I might try lowering it to 5 volts to see if it makes the excess hiss go away. With different numbers of DAC chips, you would have to toy around with it to see what value I/V resistor makes the best sound with no distortion.

I chose to do this JFET output stage because Mr. Gerhold claims it sounds like a tube stage. Mostly 2nd and 4th harmonics, I believe he wrote. Tubes are cool and all, but I didn't want to jump right in and buy transformers, tubes, sockets, high-voltage regulators and caps, etc. and not know how it would sound. Lukasz claims he prefers tubes still. I'm hoping he'll do more experiments with the FETishizator, but he already updates his site with so many projects.

 

[Realrift]

Ranchu and Chris19, thanks for the info.

I have another question.

I am using USB myself but I wonder if anyone using coax has tried to output a word clock signal from the Valab. This would be to sync to a transport or soundcard via coax to try and reduce jitter even more by slaving your source to the clock in the dac. (Probably won't make much difference in the reclocking version but perhaps for the earlier versions).

I know this worked well with an EMU 1212M soundcard which was able to lock to an external clock signal from an ART DI/O dac.

Cheers,
Rift.

 

[Realrift]

Hi All,

Out of curiosity, I checked the output voltage of my cheapie ebay power supply while the dac was on and found it reading 13.82V! (With the dac disconnected it reads 15.2V.)

This is some sort of unregulated variable power supply. I was wondering if I could add a resistor to the positive lead going to the dac to reduce the voltage down to around the suggested 10.5V or should I be looking for a new power supply before I fry the dac? (I had noticed it was running a bit hotter than before).

I have seen a power supply in a local electronics store which is a regulated variable supply 9v, 12v etc. But I wonder if 9v is too little and 12v is too much (assuming this supply outputs what it is rated at!).

Thanks for any advice (sorry for flooding, I am asking a lot of questions).

Cheers,
Rift.

 

[Ranchu]

Power supplies are kind of voodoo in the way they work, a resistor would slow down your supply, but would cool down your regs.

This guy (a trade member here[Thanks, man!]) has a 'Power Supply Parameter Estimator' that I've a lot learned from, plugging different values in, reading the helpful 'question marks'.

Also has nice looking kits.

Tangentsoft

 

[krisno]

Mother of god...

The 1.0uf Jupiters have been looping white-noise the past 60hours. They have opened up some more, resolution is higher, and they seem somewhat brighter. BUT, now the 'sharpness', fatiguing sound is back. The 'abit etchy' top.. This was even more pronounced on the obbligato's but was not noticable in the jupiters at the beginning(probably masked)?

Is this because i use USB? Can somebody verify is COAX and USB from the same computer sounds the same?? Maybe I should get a new computer and switch to coax? I tried the latest driver, heard no difference....

Maybe I need a amp with a OPA 2134 on the input or something, just to filter out those nasty's. The AKG K 701 are sensitive.

Will it be fixed by replacing that resistor?? Is it clipping / distortion I hear???

Maybe all NOS have this problem, and my only solution is the USB Ayre

K

 

[Realrift]

Ranchu: Thanks for the link.
I remember some1 saying the stock transformer puts out 13.2V so I am guessing 13.82V is going to be manageable for the voltage regulators on the board and leave it at that for now.

I think I will spend my time doing some other mods like power caps, resistors, grounding and modifying the noise filter network instead of worrying about 0.62V extra before the regulators

Krisno: Have you tried bypassing the output caps. It has been suggested on here before that smaller caps like russian teflon or MKP 1837 may inprove the top end eg. 1uf then bypass with 0.1 and 0.01uf caps. I am going to try this myself as sometimes I feel the Mundorf Silver/Foil I stole from another dac can be a bit bright.

Cheers,
Rift.

 

[hifi2001]

Quote:

Originally Posted by Realrift /img/forum/go_quote.gif Have you tried bypassing the output caps. It has been suggested on here before that smaller caps like russian teflon or MKP 1837 may inprove the top end eg. 1uf then bypass with 0.1 and 0.01uf caps. I am going to try this myself as sometimes I feel the Mundorf Silver/Foil I stole from another dac can be a bit Rift.

Here is something to consider:

I bypassed the output cap of my 2009 version with 0.47uF Russian K40Y PIO initially and thought it sounded better than stock... Last night I decided to remove the stock 10uF output cap and replaced it with 0.22uF Mundorf Supreme I had on hand. I noticed right away my mid was much smoother, and little was lost in the bass area. I then removed the 10uF cap from the other side and bypassed it with0.47uF K40Y and did notice smoother mids as well. Between Mundorf Supreme and K40Y, Mundorf sounded slightly better... lesson learned here is that stock 10uF should come out and replaced with better film caps. I will give those Jupiters a try next and bypassing with teflons...

My comment about indifferent bass response above could be that my 300B SET amp and my 15" JBL pro drivers don't go below 40Hz thus I couldn't hear any difference...

Your mileage may vary.

 

[Bill Allen]

HiFi2000,

Thanks for sharing your experience.

The Valab DAC is without a doubt the best tool I have ever used for determining what a Cap "actually" sounds like. Every coupling cap gave a different result with the Valab. I tried a bunch and the results were a can of “mixed nuts”.

I really do believe most people could get away with using a 1uf cap and hear little or no bass loss. This would be especially true if you use ported box speakers. Most have a built in bass boost in the 60-80hz region and roll at 24 db below tuned frequency. With sealed boxes you are rolling off at 12db below the box tuning. One could even fine tune their "in room" bass roll off by proper selection of the Valab coupling cap. Perhaps a 2.2uf maybe a 3.0, who knows maybe a .47 may work for your situation. Kind of unorthodox, but the key is … you have to try ‘em for yourself!

My speakers have response down to 20Hz and I find that 4.7uf is my minimum size. I just hooked up some 5.5uf Jupiter's to my unmodified 2008 Valab and found maybe even a touch more bass air than the 4.7uf's. I directly compared the 5.5uf to some 1.0uf Jupiter's in my system. I listened to a well recorded acoustic jazz recording and if I focused on say the bass player, they sounded the same, drums the same, but if I just kicked back and took it all in, it’s easy to feel the hair on entire body moving with the big 5.5uf caps. That effect is completely absent with the 1.0's.

On the other hand there is more resolution of the finest details with the 1.0. I'm talking the rarified upper air here. To get that back with the 5.5uf I tried various bypasses. In the end the perfect balance was achieved with .22uf bypasses made by the same manufacturer. I tried this experiment 5 months ago and ended up with the same results. What can I say; Daddy likes the Jupiter’s bypassed with Jupiter’s in the Valab. (I know what you’re thinking, how about a triple by-pass with some tiny Teflon’s. Yea maybe but not on my Radar Screen right now).

I just received a Fed EX notice that my “Next Generation” Jupiter Caps are on their way to my Lab. For those of you in the know, NG Jupiter’s now come Cryo’d from the factory. Nice. I had Chris roll me some .22, .47, 1.0 and 4.7uf’s from his latest batch, fresh from the Cryo Lab. The 4.7uf are special order at this point and time. It will probably be a couple weeks, but I will report back on what the NG Jupiter’s sound like. Yea I’m a little excited!

 

[PatOMalley]

Re the 4.7 vs 1uF thing.

I did notice some small tightening up /w the 1uF vs the array of caps I had in previous but I thought was just the deal of using one cap vs an array /w the BG on the bottom. And my speakers are ported and are bass plentiful. And I did notice the air thing, and opening up, as well. Again, assuming it was just the change from the cap array.

The change is there and like any audiophile worth his salt the change is 'worth noting.' It is the "Ahh" factor. But it is a trade off. Or a sideways move in some cases, as you describe.

I think I am happy to make this trade off.

Now I don my flameproof vest. I have the Yaqin tube buffer on it's way. With these cap changes the change we may be seeing is in impedance differences not in signal across various mass. If that be so then impedance matching should be where the target is. Who knows? The tube buffer is on it's way as well as the slow delivery of the CineMag transformers.