I/V

[multibit]

Just came across the testing posts, I also thought the output of my Ezdac was a little low tbh, I've been using a Burson buffer on the output which has 6db of gain. The op-amps I'm using in the Ezdac are THS4031's, if I want to increase the gain without hurting sonics do I change R23,R24,R29 and R30 from 25R to 50R? R26,R27,R33 and R34 from 1k5 to 2k7?   Do I keep  R25,R26,R31 and R32 left at 200R?   Cheers, Leo

[Evan Zamir]

Leo, you have three choices:

1) Raise 25R->50R
2)Raise 1k5 to 3K (or 2k7 or whatever)
3)Lower 200R to 100R

or some combination thereof. I put up the simulation so you (or anyone else) could play around with the values and see the net effects. I would actually try #2 or #3 before trying #1. It's probably best to keep the load on the I/V as low as possible. Good luck! Does everything sound o.k. otherwise?
-evan

[mult...@ntlworld.com]

On Apr 11, 3:54 pm, "Evan Zamir" <zamir.e...@gmail.com> wrote:
> Leo, you have three choices:
>
> 1) Raise 25R->50R
> 2)Raise 1k5 to 3K (or 2k7 or whatever)
> 3)Lower 200R to 100R
>
> or some combination thereof. I put up the simulation so you (or anyone else)
> could play around with the values and see the net effects. I would actually
> try #2 or #3 before trying #1. It's probably best to keep the load on the
> I/V as low as possible. Good luck! Does everything sound o.k. otherwise?
> -evan
>

> On 4/11/07, multibit <multi...@ntlworld.com> wrote:
>
>
>
> > Just came across the testing posts, I also thought the output of my
> > Ezdac was a little low tbh, I've been using a Burson buffer on the output
> > which has 6db of gain.
> > The op-amps I'm using in the Ezdac are THS4031's, if I want to increase
> > the gain without hurting sonics do I change
> > R23,R24,R29 and R30 from 25R to 50R?
> > R26,R27,R33 and R34 from 1k5 to 2k7?
>
> > Do I keep R25,R26,R31 and R32 left at 200R?
>
> > Cheers,
> > Leo
>

> > [image: Free Animations for your email - By IncrediMail! Click Here!]<http://www.incredimail.com/index.asp?id=101274>
>
>
>
> imstp_pets_cat1_en.gif
> 47KViewDownload

Thanks Evan,
I'll try 2 and 3 then:-) Once I've replaced those I'll let you know
the results.
The Dac is sounding fine, I posted the latest pic on Diyaudio but
here it is again just incase
[IMG]http://pic20.picturetrail.com/
VOL1293/4634025/12749957/244973386.jpg[/IMG]

[mult...@ntlworld.com]

http://pic20.picturetrail.com/VOL1293/4634025/12749957/244973386.jpg

[mult...@ntlworld.com]

I've replaced 1k5's to 2k7 and and lowered the 200R's to 100R.
Its MUCH louder now, only thing is the bass seems a little boomy,
not sure what may be causing that, midrange is a bit euphonic.
Could raising the gain be causing this? I have several different
amplifiers here and tried it with my main ones, modified UCD, Aksa,
Avondale NCC200

Cheers.
Leo

[Evan Zamir]

Well, maybe the output is too high now (like >> 2 V). Can you measure it? I think maybe just one of those changes would have been good enough. You've raised the gain 4X.
-evan

[ezkcdude]

One more thing. According to my simulation, which I just ran, the
changes you made would produce about 4V at the output. That's too
much.
-evan

[mult...@ntlworld.com]

Ok;-) I'll try changing the 2k7's back to 1k5 and leave the 100R's,
the I/V resistors was left at 25R
before changing the 2k7's I'll measure the voltage just to see

Cheers,
Leo

[Evan Zamir]

Cool. I'll stay tuned. It will be interesting to see if your measurements come close to my sim. You're using PCM1794, right? Just wanted to be sure.
-evan

On 4/11/07, mult...@ntlworld.com <mult...@ntlworld.com> wrote:

[mult...@ntlworld.com]

Good timing Evan;-)
Yes, dac is PCM1794, with 25R,100R and 2k7 I measure 3.96v, this was
with a 1k Sinewave.
Now we have 25R,100R and 1k5, I measure 2.1v using same 1k Simewave.

Resulting sound is better, bass boom gone, midrange back to being
delicate;-)) I still like the sound much better using the Burson
buffer on the output connecting to the amp.
If you design a new dac board its deffinately worth considering some
nice regulation, since fitting a Flea for the XO and one for the 5v,
the pair of discrete Bursons for the op-amps the sound has really
improved IMHO! its a nice dac using the standard regulation but the
better regulation really boosts it up.
Well done Evan! have you now seen the RAKK dac mkII? notice it now
uses SRC4192 and costs a lot more http://www.kandkaudio.com/digitalaudio.html

Cheers,
Leo

On Apr 11, 8:09 pm, "Evan Zamir" <zamir.e...@gmail.com> wrote:
> Cool. I'll stay tuned. It will be interesting to see if your measurements
> come close to my sim. You're using PCM1794, right? Just wanted to be sure.
> -evan
>

[Hans Nielsen]

Hi, now I am started on the dac projekt. A friend of me soldered the IC`s for me and another smd components. I have a question u6 it is a transistor or what? I will use lt 1084 instead of lm 317. I use ad 1896 as upsampling dac and pcm 1794, I use ad 8610 as output amps. I have a modified squezzebox3 as source and no preamp. I use my sbox directly in my new classd amp. I will let you now, when I am finish, I am 66 years old, so it is not so easy for me.      -------Original Message-------

[Evan Zamir]

Good luck, Hans! U6 is a reset generator. It's not really important for sound, but it keeps the power turning on and off "nicely". Please use the exact part number from the BOM.
-evan

[Evan Zamir]

Leo, this is neat! It confirms that my simulation makes sense.

As for the better regulation, that is obviously one of the ways that the DAC could be improved. Of course, the downside is cost, not just components, but increasing board size. With the current board size, it is very easy to prototype. If I make the board any bigger, then I must spend much more to prototype, and since I'm not really running a business, that's too much for me.

That said, I think one way to go about addressing this is 1) the way you did or 2) using the Jung super-reg design with remote sensing, and simply bypass the local regulation. The good part is that I think the current PCB could be used in this manner right now :)

Anyway, for the cost, it seems to me like the basic EZDAC or even the fully souped-up "Fancy" EZDAC is a pretty good deal in audiophile terms, right?
-evan

On 4/11/07, mult...@ntlworld.com <mult...@ntlworld.com> wrote:

[Evan Zamir]

Oh, I forgot, about the RAKK DAC - it costs a bloody fortune!
-evan

[mult...@ntlworld.com]

Hi Evan,

yes, for keeping costs down its better to stick with the current pcb,
it already sounds very good and so far as can be seen the other
PCM1794 kits cost much more;-)
I have done quite a lot of mods to my SB3 too, upgraded the output
stage with THS4032, LPF etc, internal regulation fitted with small
Audiocom super regs, built a modified VBE ALW super reg set for 5v for
the external supply, replaced the internal capacitors to Rubycon ZA/
ZL, changed the coupling caps to films, removed the inductors in the
SPDIF out and fitted a pulse transformer so the SB3's analogue out
sounds much better than the standard SB3 and the SPDIF is improved
too.
My Ezdac now beats the modified SB3's analogue out, its better than my
diy TDA1543 dac which uses BG caps and Asynchronous reclocking, its
not quite as good as my TDA1541S2 based dac but this one cost much
more! I've still a few things I want to try with the Ezdac.
Main thing being a Pass D1 I/V output stage but I'm not sure how to
fit one of these yet, I need to find out if it can do balanced in to
line out.
I'll post any updates on here, main thing is the Ezdac provides a
simple and cheap way to try the PCM1794 and gives good sound which can
be used as a building block to take it further if needed

Cheers,
Leo

On Apr 11, 9:55 pm, "Evan Zamir" <zamir.e...@gmail.com> wrote:
> Leo, this is neat! It confirms that my simulation makes sense.
>
> As for the better regulation, that is obviously one of the ways that the DAC
> could be improved. Of course, the downside is cost, not just components, but
> increasing board size. With the current board size, it is very easy to
> prototype. If I make the board any bigger, then I must spend much more to
> prototype, and since I'm not really running a business, that's too much for
> me.
>
> That said, I think one way to go about addressing this is 1) the way you did
> or 2) using the Jung super-reg design with remote sensing, and simply bypass
> the local regulation. The good part is that I think the current PCB could be
> used in this manner right now :)
>
> Anyway, for the cost, it seems to me like the basic EZDAC or even the fully
> souped-up "Fancy" EZDAC is a pretty good deal in audiophile terms, right?
> -evan
>

[mult...@ntlworld.com]

Good luck with it Hans:-) let us know how it turns out for you.
Do you have the Bolder modded SB3?

Cheers,
Leo

On Apr 11, 9:31 pm, "Hans Nielsen" <h...@tingbjergnet.dk> wrote:
> Hi, now I am started on the dac projekt. A friend of me soldered the IC`s
> for me and another smd components. I have a question u6 it is a transistor
> or what? I will use lt 1084 instead of lm 317. I use ad 1896 as upsampling
> dac and pcm 1794, I use ad 8610 as output amps. I have a modified
> squezzebox3 as source and no preamp. I use my sbox directly in my new classd
> amp. I will let you now, when I am finish, I am 66 years old, so it is not
> so easy for me.
>
> -------Original Message-------
>
> From: Evan Zamir
> Date: 11-04-2007 20:04:40
> To: ezdac-b...@googlegroups.com
> Subject: {EZDAC Builders} Re: I/V
>
> Well, maybe the output is too high now (like >> 2 V). Can you measure it? I
> think maybe just one of those changes would have been good enough. You've
> raised the gain 4X.
> -evan
>

> imstp_pets_cat1_en.gif
> 47KViewDownload

[Hans Nielsen]

No my friend do it for me, I use the selected dac, and the output amp is ad 8620. The 100 mf , and the 220 mf are now blackgate fk, and the 3300 mf is also blackgate fk. I use an analog powersupply with lt 1084 with 10000 mf on the input, and 47000 mf 4 pole on the output.     With kindly regards Hans Nielsen 38800206 30643809 -------Original Message-------

[mult...@ntlworld.com]

Sounds nice Hans, let us know how the Ezdac compares:-)
BTW I love the 4-pole caps, the 4-pole T-networks are my favourite psu
capacitor! the 4 pole Jensens are also very good

Cheers,
Leo

On Apr 12, 10:42 am, "Hans Nielsen" <h...@tingbjergnet.dk> wrote:
> No my friend do it for me, I use the selected dac, and the output amp is ad
> 8620. The 100 mf , and the 220 mf are now blackgate fk, and the 3300 mf is
> also blackgate fk. I use an analog powersupply with lt 1084 with 10000 mf on
> the input, and 47000 mf 4 pole on the output.
>
> With kindly regards
> Hans Nielsen

> 38800206
> 30643809
>
> -------Original Message-------

> 16_f_phone.gif
> 1KViewDownload
>
> 23_c_cell.gif
> 1KViewDownload
>
> kitty.gif
> 28KViewDownload
>
> imstp_pets_cat1_en.gif
> 47KViewDownload

[Hans Nielsen]

The output capacitors after the ad 8620, is 33 mf bipolar blackgate.    -------Original Message-------

[cyclebrain]

Like everything in life, changes are a trade off. It's always
something. Increasing the R value of the I/V resistor provides a
better drive signal to the gain stage. At the same time a lower R
value provides a better load to the DAC.
A comprimise.
Looking at the 1794 data sheet shows that having a 4 v output improves
the S/N ratio and dynamic range by 2dB from
127dB to 129dB. Insignificant and a 4 V output would be to high for a
preamp input. But a 4 V output should be OK as far as sonic value. The
max output is limited by the supply voltage. +/- 12 V.

What are you guys using for a reference 0dB digital input?
800000(-FS) and 7FFFFF(+FS)

[mult...@ntlworld.com]

I run the Burson regs at +/-14v for the output, the op-amp I use is
THS4031, maybe try fitting a CRD on the output to op-amps supply with
worst PSRR next.
For digital source I'm using a modified SB3 with the variable output
attenuation disabled and set to fixed

[Evan Zamir]

Looking at the 1794 data sheet shows that having a 4 v output improves
the S/N ratio and dynamic range by 2dB from
127dB to 129dB. Insignificant and a 4 V output would be to high for a
preamp input. But a 4 V output should be OK as far as sonic value. The
max output is limited by the supply voltage. +/- 12 V.

In this case (i.e. the PCM1794 datasheet), though, the I/V is done with the op amp. In my DAC, it is done with the resistor, so I don't think you can make a direct comparison. At any rate, if someone wants to play around with all the parameters, they can simply use the simulation I posted on my website.

[cyclebrain]

Just being anal, but the I/V conversion methods may give different
absolute values, but the gain in S/N and dynamic range would still
apply. Noise levels usually remain constant and increasing output
level improves S/N and dynamic range. Still 4V out is too high.

[Hans Nielsen]

Hi guys is the THS4031 better than opa 627.

-------Original Message-------   From: Evan Zamir Date: 13-04-2007 15:10:00 To: ezdac-b...@googlegroups.com Subject: {EZDAC Builders} Re: I/V

[mult...@ntlworld.com]

In my opinion the THS4031 is much better than the OPA627, of course
its all down to personnel taste:-) you may like the OPA627 better!

> imstp_chubbi_by_im_en.gif
> 27KViewDownload

[Hans Nielsen]

Can you tell why it is better.  -------Original Message-------

[mult...@ntlworld.com]

I've never liked OPA627, I find it to sound too artificial, added
glare in the mids and round bass, also this opamp works better at
+/-15v and the Ezdac uses +/-12v
Some people like the OPA627's signature, to my ears its too Euphonic

On Apr 14, 1:40 pm, "Hans Nielsen" <h...@tingbjergnet.dk> wrote:
> Can you tell why it is better.
>
> -------Original Message-------
>

> imstp_chubbi_by_im_en.gif
> 27KViewDownload

[Evan Zamir]

Leo, not that you'd want to, but the ezdac can be configured to operate at +/-15V. It just involves changing the adjustment resistors on the regulators. If anyone wants to do this, and doesn't know how, just drop me a line. Of course, you'd also have to feed the ezdac a higher supply voltage, but that shouldn't be a problem either. The only catch is that the 3.3V and 5.0V will be dropping quite a bit more voltage, so they would run hotter.
-evan

[mult...@ntlworld.com]

Thanks Evan, well its easier for me as the Burson regs have trimmer
pots to adjust the voltage, I now run the THS4031 at +/-14v.
I may next use separate transformers for each regulator or at least
use a seperate one for the 3.3v, its the main reason I fitted a din
socket to the case :)
Has anybody tried the AD825 in the Ezdac yet? I was thinking of
trying some other op-amps but because I've already tried a couple I
don't want to risk damaging the track

Leo

On Apr 18, 10:20 pm, "Evan Zamir" <zamir.e...@gmail.com> wrote:
> Leo, not that you'd want to, but the ezdac can be configured to operate at
> +/-15V. It just involves changing the adjustment resistors on the
> regulators. If anyone wants to do this, and doesn't know how, just drop me a
> line. Of course, you'd also have to feed the ezdac a higher supply voltage,
> but that shouldn't be a problem either. The only catch is that the 3.3V and
> 5.0V will be dropping quite a bit more voltage, so they would run hotter.
> -evan
>

[Hans Nielsen]

Hi Evan I use ad 8620 in my newclassd amplifier, it has a superb middle and top, and I use ad8610 in the ezdac so I think it will be very god, for the moment I use my sbox 3 directly analog output in to my new classd amplifier, and it sounds very god.

[mult...@ntlworld.com]

Yesterday I replaced the pair of 47uf tants on the Vcom L/R pins with
47uf 6.3v BG NX Hi-Q's, the dacs been running since yesterday, I know
these caps take a while to run in but first impressions are already
very good:-)

Does anybody know the current draw on the 3.3v ? I'd really like to
get a nice regulator in there too but want to know the max current
first.

I wish you could hear my dac now Evan, its performance is very
impressive, the better regulation has really transformed it.

Because I'm not sure how to use the Pass D1 I/V stage with this dac I
have my eye on something else, if it works out I'll post the details
later:-) This won't be until the 3.3v reg is sorted, I think I'm
going to be needing a bigger case:-)) space is getting limited in the
little Hammond

Cheers,
Leo

[Evan Zamir]

Leo, keep us posted. It's good to know that the same DAC can be improved so much just with better regulation. And I certainly don't mind if someone just uses my board as a starting point. And...yes...I wish I could hear the DAC, too! Maybe next time I'm in the UK...
-evan

> > > > > In this case ( i.e. the PCM1794 datasheet), though, the I/V is done