Experience with this DIY DAC ?

Jump to Latest
#2,742
fred_com said:
Hi guys, wanted to share my implementation of 2.5V mod for CS8421. All you need here is 1117 fixed voltage regulator (I went with LM1117-2.5) and possibly some kind of mounting pad for SOT-23 casing (I cut mounting pad from main DAC board, where 3.3V regulator used to be, it was replaced by Salas regulator), or you can mount it on the wires, or some other way. Now, if you'll look at the CS8421 daughterboard layout, you'll see that 3.3V comes out of AMS1117 regulator and goes into pin 17(VL) of the chip, and then _under_ the chip this trace goes to pin 3(VD), which is the pin we need to provide with 2.5V. So, it is not possible to cut the trace without removing the chip, which can be done with the help of air gun of course, but I went with carefully lifting pin 3 with a needle and connecting 2.5V output from new regulator. You'll need soldering iron with fine tip to do that. Here are the pictures, sorry fo my ugly soldering 🙂
An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.


As for the wiring of new 2.5V reg - connect it's input with 3.3V reg input and ground with 3.3V reg ground. Output goes to the lifted pin 3. The capacitor on top of the new reg is 1uF (connected to the output and ground), don't know if it is really needed there, the CS8421 datasheet says that it should be there (although it's 0,1uF in the datasheet).
Don't know if this mod intended to change anything in sound, seems that it is more about safety of the chip.
That's all, use it on your own risk 🙂

Thanks,
Fedor
Click to expand...

Hi Fedor

Don't you think it would be easier if we remove resistor next to VD and tap to the resistor hole there...???

An externally hosted image should be here but it was not working when we last tested it.
 
#2,744
Bill Fuss said:
Thanks Fedor,

Great pics, I will have to muster up the courage to do it.
Click to expand...

You're welcome 🙂

gabanyayaya said:
Forget it...3.3V is fed from behind to leg no.3 directly...

Is there any easier solution?
Click to expand...

Hi, yes, that's the problem. I don't think there's an easier solution to this one, sorry. VD and VL are connected under the chip, and there's only one way to break this connection without removing the chip - by lifting the pin 3. But that's just my humble opinion, maybe someone will come up with better solution.

Thanks,
Fedor
 
#2,745
3.3V issue

fred_com said:
You're welcome 🙂



Hi, yes, that's the problem. I don't think there's an easier solution to this one, sorry. VD and VL are connected under the chip, and there's only one way to break this connection without removing the chip - by lifting the pin 3. But that's just my humble opinion, maybe someone will come up with better solution.

Thanks,
Fedor
Click to expand...

Fedor

while using this dac before this mod did you observe hi temperature on the cs8421?

it seams that this over voltage is not an issue at all for those with cs8421 in their dac. i don't see any complains so far nor any mishap during operation of this dac.....should we take this seriously or just a lousy mistake by the Chinese engineer...??? 😕
 
#2,746
gabanyayaya said:
Fedor

while using this dac before this mod did you observe hi temperature on the cs8421?

it seams that this over voltage is not an issue at all for those with cs8421 in their dac. i don't see any complains so far nor any mishap during operation of this dac.....should we take this seriously or just a lousy mistake by the Chinese engineer...??? 😕
Click to expand...

I think, our Chinese friends made it this way to cheapen the production. 🙂

Well, actually, I didn't measure the temperature of the chip before, so it's hard to say if it was sufferring from overvoltage. But look at the datasheet: "Normal operation is not guaranteed at these extremes." (p.10), of course, maximum allowed voltage is 3.5, but it is really close to 3.3 which is supplied.
It was fairly easy mod for me, so I decided why not? We have to always follow the datasheet directions 🙂 If you decide to do this mod, you have to weigh pro's and con's, because now it's working fine, with potential voltage peak, which may never happen. But while doing this mod, you can break the leg of the chip and it wouldn't work at all. So it's your choice 🙂

Thanks,
Fedor
 
#2,748
CS8421 mod Idea

How about replacing the stock AMS1117 3.3v reg to 1117 2.5v reg. This way Vd will get 2.5v as in the sheet. Then for VL which need 3.3v or 5v it gets 2.5v since in the data sheet the absolute Max ratings for VL is -0.3V min and 6v max.

Will this works...?
 
#2,749
Bill Fuss said:
This is the problem.
Click to expand...

Thanks Bill, I was positive that i saw post about this issue. Was unable to find it though.

gabanyayaya said:
How about replacing the stock AMS1117 3.3v reg to 1117 2.5v reg. This way Vd will get 2.5v as in the sheet. Then for VL which need 3.3v or 5v it gets 2.5v since in the data sheet the absolute Max ratings for VL is -0.3V min and 6v max.

Will this works...?
Click to expand...

Um, don't forget about the clock generator, which is also supplied from 3.3V (but maybe it can be fed from 2.5V, don't know). And about 2.5V for VL - correct me if I'm wrong, but those min-max rating voltages are stating that chip would not die if supplied with those for some small period of time. Here's datasheet quote "Typical
performance characteristics and specifications are derived from measurements taken at nominal supply voltages" (p.10). Which means that it will probably work with as low as 2.5V, but performance may be low also (don't know how it will influence the sound though). But it is definetely safer mod to try, as you are replacing only one component (of course, with usual precautions, not to short something).

Thanks,
Fedor
 
#2,751
alpertandogan said:
Hi everybody

I saw two kit in ebay.
Which one is your favorite.

Thank you

Upsampling 24bit/192 DAC KIT,TOROID TRAN,USB, PCM1793 - eBay (item 330442739859 end time Jun-20-10 09:20:43 PDT)

Gigawork 24/192 Up-sampling DAC DA CONVERTER w/ USB kit - eBay (item 160367399418 end time Jul-02-10 01:00:00 PDT)
Click to expand...

The PCM1793 is an older chip, I've never heard one but it is V out, as is the CS4398.

I got my Ebay Special trafos this morning. They are indeed Jensen JE 11-dm, one was marked as this and the other was marked RE. I ran a couple sweeps with the generator and scope @10vp-p, they are flat from 15hz to well over 200khz. No sign of saturation. They sound great, the best I have now.
 
#2,755
The guy has built them to match the caracteristics of voltage out chips. He matches up the inductance and impedance to the values ideally needed. I think they are around 3k, dont remember the inductance. I believe there was a few posts by him on jkennys transformer thread. Pretty much all math.
 
#2,759
Hi Guys,

Didn't know my name was being taken in vain..... I do build Dac Buffer transformers. So do a number of other folks.

Major difference between my and Lundahl parts are in information retained. To use amorphous core material as Per does, you must eliminate capacitive coupling as much as possible. If you don't you get enormous FR spikes around 60 to 80 K.

The amorphous core is quite good enough for the purpose and the Lundahl products are very clear, linear, with excellent tone and transient response. They are however slightly lacking in internal note and transient gradient information. Not a bad thing.

In comparison, ours have a bit more distortion and so are slightly softer in character, but the trade off is in information density, in those small amplitude signals that provide coherent gradient information.

For those who prefer solid state sound qualities, I would strongly recommend Lundahls products. Clear, clean and sweet. For those of you who just can't get enough information, I would recommend O-Netics transformers. Clear, clean, sweet and with a fire hose of of music attached.

At these levels of performance it really is a matter of taste and I hope I have fairly portrayed the differences. Jensen is sort of in the middle and no negative connotations should be drawn about their performance either.

Bud
 
#2,760
some guidance in learning about audio trafos

Bill Fuss said:
The PCM1793 is an older chip, I've never heard one but it is V out, as is the CS4398.

I got my Ebay Special trafos this morning. They are indeed Jensen JE 11-dm, one was marked as this and the other was marked RE. I ran a couple sweeps with the generator and scope @10vp-p, they are flat from 15hz to well over 200khz. No sign of saturation. They sound great, the best I have now.
Click to expand...
...about this testing and all that is needed to be known in matching trafos, is there, to the best of your knowledge, a comprehensive collection of or even an old fashioned book, that I could buy and read?
as I have told Bill, my original training is far from this, but now and at my 50ies, I feel very much inclined in learning as much as I can about it, and more..in the future
regards
angelo
 
You must log in or register to reply here.
Top Bottom