Looking at the schematics
would tell you, 1200pf is the value of the capacitor in the analog RC filter (after the DAC ladder) and 470pf of the capacitor in the feedback loop of the last op amp
My calculations (which may well be wayward) put the cut-off frequency (-3dB point) of your mod at 25kHz, not 32kHz. 625R output impedance, and 470pF of stock gives a 212kHz cut-off frequency, rather than the 270kHz shown in some places. Please correct me if I'm wrong.
Lowering the cut-off frequency to that point or even 32kHz means that rather than being the highest filter curve in the FIR1-FIR2 chain it will be lower than all but the 44.1/48kHz and will "apodize" all other filters. Rather than the final "clean-up" filter having virtually no influence on the audio at higher rates, it become the dominant filter and either replace or fundamentally alter the response of any custom filter designed for the higher rates.
To my mind it's a mod that will cripple the customisable filters, and degrade performance at higher sampling rates.
After listening to the new Muse album yet again today, make no mistake there is a big difference in low-end performance. At the same volume the songs have more power and punch to them. I've been using the NewNOS filter for the last few weeks again.
Heres the pic with diode bridge removed, highly recommended mod when using shunt regulators, as otherwise the diode bridge doesn't allow any reverse current back into the shunt FET. It certainly allows the dam to shutdown quicker at poweroff due to the onboard filter caps now able to throw any charge back into the shunt. (I think?). Also WARNING! With the diode bridge bypassed as shown the + power lead HAS to be the one nearest the screw hole.
Heres the pic with diode bridge removed, highly recommended mod when using shunt regulators, as otherwise the diode bridge doesn't allow any reverse current back into the shunt FET. It certainly allows the dam to shutdown quicker at poweroff due to the onboard filter caps now able to throw any charge back into the shunt. (I think?). Also WARNING! With the diode bridge bypassed as shown the + power lead HAS to be the one nearest the screw hole.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Thanks.
Since the cap is in an audio circuit the impedance won't be very low (as for a digital power supply) and hence the physical position won't matter much. Some inductance in series with the cap won't be a problem so you can run a wire from a quiet ground instead of the via into the groundplane.
Is there a schematic and layout depiction somewhere so I could identify a suitable quiet ground node? I'd not want to use an output ground node as they tend to be noisy due to common-mode currents.
I'd highly recommend trying the EQHQv2r1 if you haven't already. Its very similar in sound to the NewNOS but looses the grain and mid-range haze I hear in the NewNOS.
I also agree with your comments on the cap across the 3.3V reg output. I couldn't notice any real benefit from adding a 56uF panasonic FC there.
I'm also dubious about the gains from swapping out the existing PSU caps - I've upgraded to panasonic SMD OS-Cons and I'd be hard pressed to say there was any noticeable benefit from doing so.
I've only swapped out the +/-5V rail caps, left the main supply caps alone. Did you change the same caps as well? Upgrade to a larger value? I'm also using the single ended output which most certainly has a better sound to it.
In other words, too many variables between setups.
That is an interesting observation.
I decided to remove the 470pf NPO's from the output buffer feedback loop just to see if it made any difference. I think there might a subtle improvement in terms of detail and air, but it's pretty minor. It's possibly a bit smoother too but as vuki observes, without an unmodified DAC to compare with I could be just imagining things.
Lots of variables... The balanced output is far less sensitive to power supply issues as we saw with zfe's tests on the influence of USB -> rs232 noise break through. That is mainly down to common-mode rejection. It's very likely that improvements to the power supplies are going to have relatively less audibility via the buffered outs cf the unbuffered outs.
I got the Per-Anders SSR-04 board in the mail today so I'll build that up over the next week or so and see how the DAM sounds with a good 12V supply vs AC.
FWIW the main PSU caps have been replaced with 1000uF/16V Panasonic SPVC caps, which I think are the ones Søren suggested or close to it. I haven't touched the 5V supplies at this stage.
You'll die waiting.
From the filter thread one of my withering "insults" to you that seems to have provoked a lingering resentment...
And a post from you directed at me later in the thread, after I'd commented that I could hear no difference with the USB -RS232 hooked up. Oddly enough, I'd measured this prior to hooking up permanently and found very little difference via balanced outs - just some very low levels of added noise and a couple of spurs at 30-40kHz. This has since been backed up by zfe's measurements.
I've consistently maintained that the filters needed to be sorted out to an acceptable standard on a stock DAC, not that mods were verboten. I believe the filters reached that point about a month ago with the EQHQ filters and I'm not alone in that opinion.
Again I don't really understand where you get the idea I was against modding, my point has been that I wanted a stable platform while I was focusing on the filters. There were enough small changes to listen for without introducing a raft of variables through component changes.
Last edited:
Have to say I'm frustrated – at my end - as I want to have a go at building filters myself, albeit a purely academic exercise as the work documented in the Filters Forum I doubt can be bettered. Also, I want to look at power supply options as I think likely this is the area where there is the most opportunity for improvement, most especially the 4V analogue supply. I wonder if we can make improvements with a dedicated, high-quality 4V supply? I’m led to believe the 5V op-amp input is too low, but I could be misinformed. The only mod on my board currently, I have fitted a couple of 2200uF Black Gate caps across the 5V supplies. I feed output from the ladder into a dedicated op-amp/valve amp combination. Not noticed any real noise pollution with the usb permanently connected.
As we all know, trying new options takes time to do correctly – having ideas, implementing, testing, evaluating, etc. I’ll have to wait for now until the garden gets sorted (be a first) and the holiday season draws to a close. No worries, I am enjoying listening when I can and reading all the posts in the meantime. Thanks to all.
As we all know, trying new options takes time to do correctly – having ideas, implementing, testing, evaluating, etc. I’ll have to wait for now until the garden gets sorted (be a first) and the holiday season draws to a close. No worries, I am enjoying listening when I can and reading all the posts in the meantime. Thanks to all.
FWIW, I can verify that the board works with +-5V (all functions, except for the output opamps which have been removed here - obviously the onboard 5v regulators are removed).
I'm running SSR03 set for +-5V and a simple ADM7150 supplying the 3.3v... only the dirty side of my isolators are powered by Amanero at the moment. Everything else is from those two supplies.
Just finished the mods and won't have time with the kit until next week - so no sound impressions.
I'm running SSR03 set for +-5V and a simple ADM7150 supplying the 3.3v... only the dirty side of my isolators are powered by Amanero at the moment. Everything else is from those two supplies.
Just finished the mods and won't have time with the kit until next week - so no sound impressions.
Last edited:
Actually I've been through several hundred of my posts and I yet to find a single instance of me insulting someone for modding a DAM.
You might want to back up your claim with some evidence, otherwise I think it is safe to assume you are simply engaged in a petty attempt at character assassination.
Last edited:
BRAVO!!!
You beat me to it since I have been working with the battery/floater power supply for the computer.
I think you owe a little gratitude to nige2000 since he is the only person I am aware of to have removed the 5 volts regulators. A correction would be appreciated if I have made an assumption.
to tnt
I do not remember spzzzzkt ever giving anyone a hard time about supply mods. There were plenty of others whose names I never bothered to remember but spzzzzkt has been active on this forum, from the beginning? Really close if not then. I feel sure he was involved before I became aware of the DAC. If he had made such a comment I would have taken notice.
Unfortunately for me he took a comment I made at TIR NA which was aimed at the folks saying, here, snake oil and such with the assumption that with moderately priced components SOEKRIS had made the perfect supply - improvement was impossible! Of course, only GOD could make such a thing happen. Only a miracle could effect a transmogrification of that magnitude.
There is no question that spzzzzkt had made this DAC a pleasure to listen to. The DAC's architecture is fine - the details were rough and spzzzzkt is a first class finish carpenter of the audio variety.
So, I have to say, you are wasting your time with these kinds of posts. EVEN if he had said such a thing it would pale in consideration of what he has added.
Nigel is not the only one to have removed the 5v regs... See :
http://www.diyaudio.com/forums/vend...magnitude-24-bit-384-khz-285.html#post4353563
Last edited:
@Soren. Further investigation reveals that the reason for this is the 4V rail not coming up. -4V is ok though. Please note that this issue has been present since first receiving the DAC. However as I have now done modifications I understand all bets are now off regarding possible warranty coverage.
However I was hoping you could shed some light on this please? I recall Jung regulators being notorious for start issues in the early days.
Sounds more like a simple manufacturing defect, probably a bad connection somewhere....
As you have done modifications I assume you can do a little troubleshooting, before sending the board back for warranty repair....
1) Is +5V ok ?
2) Is the problem with both channels ?
If so then the problem is around the the -4V to 4V inverter, those parts are at the edge next to J8, look closely for bad solderings....
If you're unable to fix it yourself, send an email to info@soekris.dk for RMA instructions.