Valab NOS DAC - subjective listening and modifications - Page 14

Off Topic

Wish I could join in on the fun here but I've got a litter of premature Bulldog pups that need to be fed on mom then tube fed every 2 hours around the clock for at least the first 10 days. If I picked up a soldering iron right now with no sleep I'd sure as hell burn everything up.

Cute! Good luck with them. I am a Beagle owner but have friends with Bulldogs. Lots of fun. It's like having a walking small wall in your house. It's a shame we don't use them to tenderize meat any longer.

Yeah, I remember there was a big stink over pet posts on the Vinyl Asylum, they were mostly the passing of a pet. Thanks Pat, when I get some sleep and some time I'll put em in.

mrdon

It's looks like there is mod wire between CM108 and 74xx157, What mod is that ? Thanks

May

Has anyone considered the LM334 constant current source mod (apparently standard now on Ver 2.1.1). The LM334 is utilized together with the 10-turn trimmer pot to provide a constant current source at the DAC chip Vref input. Together with the lowered (330R) I/V resistors this is part of their (Valab/TeraDak) solution to the "clipping/distortion" problem.

The LM334 may provide better temperature tracking and stability for the 1543 reference voltage (and thus DAC chip bias current).

I have just received a VALAB late version but unfortunately did not have the LM334 addition, but does use the 330R resistors. I think that I may be detecting a trace of the clipping/distortion issue on certain occasions/recordings - if so it is most noticeable on vocals. The recordings in question are "modern" but the overall recording level does not seem particularly "hot". Just wondering if the constant current source might provide additional reduction/elimination of this issue.

I'm running itunes on an iMac and connecting via USB. Also, noted that Steve Nugent has recommended feeding this DAC with a "low-jitter" source to realize it's full potential. Does the iMac/USB connection alone qualify as low-jitter or would adding his Off-Ramp USB/SPDIF converter kick things up appreciably?

Dave

Practiced soldering and desoldering earlier this week with an old PCB - that went very well.

Ordered Caddock MK-132's for I/V and 33 uF Black Gate N's for output caps plus 0.1 uF Mundorf ZN's for bypass.

I was wondering if there's a good way to stand the Black Gates off the board a bit to provide a little more lead to solder the Mundorfs to. I was thinking perhaps a ball of Blu Tac under each Black Gate would raise them off the board a few mm. They have radial leads whereas the stock caps are axial.

The other approach I was thinking is to bend the Black Gate leads into an upside-down "L" and solder the Mundorfs to the Black Gates first, then solder the Black Gates to the board. I was thinking that I might attach the Mundorfs to the top of the Black Gates with a little Blu Tac before soldering. I attached a crude diagram of this. I won't know if this will work until I get the parts, but does this seem feasible?

In general, is Blu Tac ok for attaching a cap to a board or stacking caps together? Not sure is the oil/solvent in the Blu Tac could damage the board or if the Blu Tac could melt with heat.

Voltage Regulators

Phill- nice puppies , do you miss the smell of solder??

Rich- can you elaborate on the sound of your 3 son's.

Dave- The OffRamp would offer much lower jitter than the USB interface built inside the Valab Dac. Change the Midi output on your Mac Mini to 48/20 for lower jitter.

As far as regulators go I really don't have any opinion at the moment. Yea I know, hard to believe. But for now I will spilt the DC power delivery and run the 317A regulator on one battery and the 1084 regulator on another battery. We will see what kind of benefit this discrete power path provides. Here is the response my friend Tony recieved from Paul Hynes.

> Hello Paul,
> > I have a few questions about ordering a few voltage regulators from
> > you. Thanks in advance for your time.
> > 1. I would like to order a voltage regulator to replace an LT 1084
> > unit that supplies 7.8 volts to 8 paralleled TDA1543 DAC chips. Do
> > you have anything that fits the bill?
> > 2. I would like to order a Z1705 to replace an LM 317 in the same
> > DAC that supplies 5 volts to the rest of the DAC circuitry. It seems
> > that you have those available according to the website.
> > 3. What would be the approximate shipping the the United States? A
> > friend and I would like to order two sets of these so if shipping is
> > expensive we will have them shipped to one address.
> > 4. Am I correct in assuming a 3-4 week lead time on these units?
> > 5. Can I pay using Paypal?
> >
> > Regards,
> > Tony

> Hello Tony,
>
> Pleased to help.
>
> 1. Eight TDA1543 chips can draw up to 480ma max so the SR1-M module
> set to 600ma will work well with this arrangement. I supply two
> versions of this module. The DC input version is the regulator/current
> source section only which is what you will require if you already have
> a good DC supply to feed it. The input DC voltage headroom requirement
> for this module is 5 volts higher than the regulated output voltage.
> The SR1-M modules cost £80 each. The SR1-MR module includes Schottky
> barrier rectification, low ESR energy storage and additional bias
> circuitry to allow low dropout operation. You would need a 9 vac
> transformer with a minimum rating of 18 va to power this. The SR1-MR
> modules cost £100 each.
>
> 2. The Z1705 would work well with the additional dac circuitry
> provided the current draw is less than 80ma. This can be fed from the
> SR1-M if the current setting is increased to say 800ma and, if using
> the SR1-MR, the transformer rating is increased to 24 va minimum. The
> Z1705 modules cost £30 each.
>
> 3. The insured air carriage and packing for one SR1M (quite heavy)and
> one Z1705 would be £15 and for two sets would cost £20.
>
> 4. Lead time is currently 20 working days.
>
> 5. You can pay via paypal by sending funds to paul@paulhynesdesign.com
> but please ensure that you pay in advance for any currency conversion
> charges.
>
> Regards
> Paul

OK, amateur question here...

On the PCB I used for soldering practice, all the parts are on the side of the board without the pads and traces. So, when I was soldering and desoldering, the parts were on the opposite side.

On the VALAB, the IV resistors and output caps are on the side with the traces and pads, right? This means the the leads get pushed through the opposite side of the board vs. how I practiced above.

The overall soldering technique (below) is the same for both, right?

For the VALAB IV resistors and output caps:

1) Desolder the parts from the pads on the trace side of the board.
2) Push the leads through the holes.
3) Solder the leads to the pads.
4) Clip the leads on the opposite side of the board.

Valab DAC comparision & mod question

Hi Everybody,

I'm new here and decided to join the club as I also own the Valab DAC (newest version) since a few weeks.
I've started to tweak the Valab DAC this week and the first mod I did was the DIR9001 PIN 13/14 to +3.3v.
That was quite a nice step up.

I am a bit spoiled when it comes to DACs as I've heard quite a lot great sounding DACs at work (Prism/DCS/heavily modded Benchmark).

We've compared the Valab dac this week with the above mentioned DACs:
Great thing about the Valab DAC is that is that the width/depth and overal spaciousness is great. Compared to the expensive boys it's not really accurate but it's very pleasing when having this is your living room. I have to keep in mind that the Prism/DCS/Benchmark are designed for studio use and therefore very analytical, something you might not want in your livingroom.
The DIR9001 mod makes it more accurate by the way.

The Valab's presentaion of the mid & high frequencies are well balanced in my opinion and there's not much I would like to do about it except the highs, which sound a bit like they have some sort of distortion over them but that could be a jitter/caps problem. There are not too much highs compared to the other DACs, it's just that harsch glow around it.

The biggest problem I have with the Valab is the low section. This is where the valab stays behind. It's not tight, the sublow frequencies are simply not there, or hardly there, and the overal sound below, let's say 180-200Hz is just fluffy, boomy compared to the other Dacs.

Of course I have in mind that's not really fair to compare a $200 DAC against a $5000+ DAC but it is a great challenge trying to get the Valab closer by doing some tweaks.

I would like to start with the bass, that's personally my biggest issue.

I've done some testing with 2,2uF Mundorf Silver/Gold with a MKP1837 0,01uF bypass on the output this week. The mids got much better, soundstage bigger but the highs got too laidback and the bass got even worse than with the standard russian output caps.
It could be the low value of the output capacitor being too low but that's just a guess. I had these Mundorfs left so those where the first ones to try.

Which of the tweaks would you advise me to do first? Some tips would be great!
I have to say that the Valab only played +/- 50 hours so far and I've read that it needs at least 200 before it is sounds like it should. hopefully the bass & highs get better the upcoming weeks.

One last thing to mention is what I wrote above might sound a little negative but don't get me wrong, this Valab DAC sounds great for the money and I love it!

bottom of case

So, I'm taking the case apart in preparation for replacing IV resistors and output caps.

Does the bottom of the case need to be slid off the PCB to do this work?

I assumed yes, so removed the 3 screws near the DAC chips.

It seems like the PCB is pretty tight in the groove on the bottom case, so I didn't want to force it.

I know these are newbie questions, but hopefully they will also help other first-timers.

Hi Bill,

All three dac's sound amazing.

Dac 1 - Mundorf supreme bypassed with .22 SIO , Riken resistors at I/V bypassing ferrite bead and smd. Capacitor end soldered directly to Rca's...rubycon ZA for decoupling and Panasonic FM for power filtering..Sound is very analogue sounding with a deep and wide soundstage. Vocals have a spooky 3d image..very transparent and musicians have a lot of air and spaciousness around them.

Dac 2 - Obligattos bypassed with Vacuum tube Valley PIO's and caddock resistors at I/V.rubycon ZA and panny caps...I find the OB's a touch more dynamic than Mundorfs and not as smooth, but it could be it needs more break in time

Dac 3 Obligattos no bypass ..Ruby za's, panny caps and stock I/V...sounds better than stock unit.. Cap upgrade makes dac sound faster and more dynamic..doesn't have the refinement which the bypass caps add..overall best bang for the buck. rubycon ZAs are a cheaper alternative to Black gates

Now if any of you can try a Monarchy Dip ( non upsampling model) . What I heard was very realistic cymbals in jazz recordings. The tap when the stick hits the cymbal is more detailed and less smeared sound..very nice ..used DIPs go for about 125.00


Rich

Tough Love

Adam- I recommend that you cut the I/V Resistor and Tantalum Cap out leaving as long as legs as possible. Solder your new parts to these original legs. The board is to delicate for repeated soldering attempts.

Rhodes- yes please refrain from "sound comments" before 200 hours or during any cap break-in period. Patience Please. I have owned even higher end Dac's than the ones you mention and the Valab Dac smokes them all if feed ultra low jitter data. Understand almost all Dac’s use powered analog output sections, oversampling Dac chips, digital filters, ECT, another words, far too much complexity to let the sound of "real music" to flow. The Valab is not just a $200 Dac, it’s your ticket to ride.

On this forum we talk about bang for the buck options that you can do yourself on the Valab Dac. However we all still end up with wildly different results, not because of the parts changes but mainly to do to jitter. Steve Nugent says it best in his White Paper:

Q: Why re-clock the digital data?

"Because jitter at the clock inputs of the D/A converter causes modulation of output analog signal from the D/A converter. This is distortion. This modulation is a function of both the magnitude of the jitter and the spectra (frequency) of the jitter. This is one of the things that makes digital audio sound "digital" and not analog, along with sample rates that are not high enough. The evidence of this is really obvious when you compare several DAC's to one another. With a high-jitter input signal, they all tend to sound radically different. With a low-jitter digital input signal, they all start sound very similar. Each DAC behaves a bit differently in the face of jitter, the simplest ones tending to sound the worst with high-jitter input and the best with low-jitter input."

With the Valab being perhaps "the" simplest Dac on the planet, jitter has a huge impact on the sound, but it bucks the trend and manages to sound great with any source. Wildly different sounding yes, but still great sounding! While I encourage everyone to tweak the Dac to sound great in "their" system, understand this is back asswards, we should really first fix the input jitter first ... then move on to the "flavor mods". Only by lowering the jitter to Off Ramp 3 or Pace Car levels will you see what the Valab Dac is "really" capable of. From this vantage point everyone will find what I have, most hi-end Dac's are simply well tweaked good sounding toys, they will never convey the sound of "real" music.