Reference DAC Module - Discrete R-2R Sign Magnitude 24 bit 384 KHz

[TNT]

Well the preparation for Munich should be possible to delay so that could help.

//

 

[Peterma]

I don't really know what you are talking about. But I don't consider it good business practice not to quickly move to fix an obvious flaw in a product. And yes, part of the frequency spectrum (11025 Hz) jumping around 3dB in amplitude due to massive clock drift is an actual flaw.

is there a flaw? maybe he doesn't think there is one.

 

[soekris]

is there a flaw? maybe he doesn't think there is one.

Exactly. There are no flaws, but as most other designs, things can always be improved, especially when it's now over 6 years old....

I'm about to launch dam1021 rev 7, which have several improvements, like the clock is now driving the shift registers via 'LVC FFs, not though the FPGA, improved vref (like dac2541), improved R-2R networks....

 

[avr300]

 

[PavZ]

I'm about to launch dam1021 rev 7, which have several improvements, like the clock is now driving the shift registers via 'LVC FFs, not though the FPGA, improved vref (like dac2541), improved R-2R networks....

That's great news, is there any ETA?

 

[Spartacus]

Soren, could you describe what you've done to Vref? Just curious.

 

[baten]

I'm about to launch dam1021 rev 7, which have several improvements, like the clock is now driving the shift registers via 'LVC FFs, not though the FPGA, improved vref (like dac2541), improved R-2R networks....

Awesome!

 

[TNT]

Are you planning also a similar sort of upgrade to 1121 or is it finished?

//

 

[cab]

Definition of reference (Entry 2 of 3 Webster)
: used or usable for reference especially : constituting a standard for measuring or constructing

Hardly. More like a work in progress.

 

[derekr]

Are you planning also a similar sort of upgrade to 1121 or is it finished?

//

I, too, am anxious to hear the answer to the above, Søren. I love my 1121’s

 

[soekris]

I, too, am anxious to hear the answer to the above, Søren. I love my 1121’s

The dam1121 don't need as much upgrade as the dam1021, the clock is driving the shift registers directly, and the R-2R network is also 3 bit thermometer.... But at some point the vref will also be upgraded on the dam1121.

 

[tiggerkater55]

I'm about to launch dam1021 rev 7, which have several improvements, like the clock is now driving the shift registers via 'LVC FFs, not though the FPGA, improved vref (like dac2541), improved R-2R networks....

thats great news! can you say, when it will be ready?


best!

 

[kronzilla]

Exactly. There are no flaws, but as most other designs, things can always be improved, especially when it's now over 6 years old....

I'm about to launch dam1021 rev 7, which have several improvements, like the clock is now driving the shift registers via 'LVC FFs, not though the FPGA, improved vref (like dac2541), improved R-2R networks....

Hello Sören, i have a new dam1021 board, still unused version 4or 5, could i trade that board in for the version 7?

Thank you

 

[MvonB]

Can you swap an unused tube TV for an 8K?

 

[ripson]

The last time I read about the upgrade of DAM1021 it was about improving Vref - adding plenty of capacitance (low ESR caps) and adding 2 x BC550C and 2 x BC560C between opamps and capacitors. I haven't done it.

But today I desoldered the output opamps, connected it directly to FW F2J and the difference is pretty obvious.

After years, is there any generally accepted/recommended list of modifications that is worth implementing?

EDIT: my DAM1021 has a serial number: 428, but not sure which rev it is

 

[stephengrenfell]

After years of listening to my DAM1021 unmodified with the great sound that it provides.. since I finally have some time on my hands I was about to embark on whatever improvements could be done to Vref etc...

However in reading …..

Exactly. There are no flaws, but as most other designs, things can always be improved, especially when it's now over 6 years old....

I'm about to launch dam1021 rev 7, which have several improvements, like the clock is now driving the shift registers via 'LVC FFs, not though the FPGA, improved vref (like dac2541), improved R-2R networks....

+1 for the ETA of dam1021 rev 7, please.

Thanks Sören for your great designs.

 

[roderickvd]

Can you swap an unused tube TV for an 8K?

I hear tubes are pretty great!

 

[ghitus]

Can you swap an unused tube TV for an 8K?

I think this comparison is a bit too much...

 

[danny_66]

After years, is there any generally accepted/recommended list of modifications that is worth implementing?

An easy mod is adding some capacitance to the clock and the vrefs.

Use a very good regulated power supply for the DAM and also for the USB interface.
If it's a V1 you definitely need to do the lowresmod or the transistormod for the vrefs.

 

[Berny]

After years, is there any generally accepted/recommended list of modifications that is worth implementing?

Totally agree with Danny ;-)
I will say that at least on my dac, there was an issue with the 3.3V regulator, generating some kind of a 80mV peak-peak saw-tooth pumping at around 40MHz.... this gave still 20mVpp ripple on the oscillator powersupply and a nasty 200mVpp swing on the positive raw supply rail. Buy a bag of 15 polymer caps and that will give you the most and simples improvement according to me.

Edit: Pics from left to right: pic1: 3.3v regulator output, pic2: impact on the supply rail, pic3: 3.3V reg output after adding caps on the two locations on the 3.3V(only 2mVpp remaining!), pic4: bottom of my board after mods. I also added 1000uF cap at the input of the 3.3V regulator (on top of the regulator, directly to its legs)