What do you mean by "On Semi ones"?
a second question:
What VA rating of the transformer should I use? Is 5 VA enough (that is 2 x 0.138A at 18V)?
The 9V section, is 3 VA enough?
Here are mods I have made:
- added 3x 4700uF more smoothing capacitance
- cut 7812 trace to feed only opamp (burned my finger on 7812 before this).
- added extra 7812 (with heatsink) to feed +5v regulator
- changed opamp to LT1364 (also with heatsink)
- replaced 100nF opamp powersupply bypass caps with parallelled 1nF SMD ceramic, 100nF MKP and 15uF SMD tantalum.
- added toslink receiver module
Sounds at least as good as my modded CS4398 super pro dac. But
for somereason it can only play 96khz from toslink. As this 4393 plays 192khz.
- added 3x 4700uF more smoothing capacitance
- cut 7812 trace to feed only opamp (burned my finger on 7812 before this).
- added extra 7812 (with heatsink) to feed +5v regulator
- changed opamp to LT1364 (also with heatsink)
- replaced 100nF opamp powersupply bypass caps with parallelled 1nF SMD ceramic, 100nF MKP and 15uF SMD tantalum.
- added toslink receiver module
Sounds at least as good as my modded CS4398 super pro dac. But
for somereason it can only play 96khz from toslink. As this 4393 plays 192khz.
Last edited:
Regulators made by On Semiconductors
I never did calculation, the seller specify a 30VA one.
Lots of fake opa627 on ebay. If its too cheap to be true then it probably is.
A genuine opa627bp gives very nice sound. Sound is subjective. The best one is the one that sounds best in your dac and your system. Only experimentation will tell you.
I think a valve output stage would sound "nicer".
A genuine opa627bp gives very nice sound. Sound is subjective. The best one is the one that sounds best in your dac and your system. Only experimentation will tell you.
I think a valve output stage would sound "nicer".
Last edited:
IMHO, obviously, OPA827...
BTW OPA 627 sound very very good too.
But genuine OPA627 cost much more than OPA827... why bother?
Guys,
I measured the current consumption on the AC input side:
9V AC: Measured 9,91V and maximum 26 mA current with only one led lit on. When more will lite it will go more, but it is actually not applicable after the signal start to go through the DAC, they all go out, except the power ON led.
15V: measured 2 x 17.9V input, and 2 x 26 mA maximum (that is with a 2 x OPA627, in stead of the NE5532).
So for a safe side:
9V: 100 mA transformer
2 x 15V: 2 x 90 mA transformer
Farnell numbers: 9531467 and 9531297
RS numbers: 2786891 or 2239092 (PCB mount) and 2786762 or 2238998 (PCB mount)
Haik
I measured the current consumption on the AC input side:
9V AC: Measured 9,91V and maximum 26 mA current with only one led lit on. When more will lite it will go more, but it is actually not applicable after the signal start to go through the DAC, they all go out, except the power ON led.
15V: measured 2 x 17.9V input, and 2 x 26 mA maximum (that is with a 2 x OPA627, in stead of the NE5532).
So for a safe side:
9V: 100 mA transformer
2 x 15V: 2 x 90 mA transformer
Farnell numbers: 9531467 and 9531297
RS numbers: 2786891 or 2239092 (PCB mount) and 2786762 or 2238998 (PCB mount)
Haik
This may have been covered earlier, but how does one determine a good 75ohm cable for SPDIF transfer. Everything that has that printed on the outside reads something else on the meter. I have three possible applications -
1. Normal 3-6 ft interconnect.
2. 12-16 ft remote feeder from computer to DAC
3. Short internal link from input socket to DAC PCB.
I'm using a 16 ft Monster Interlink 300 MK II Audio Cable (measures 0.0 ohms) with RCA ends that get the job done. I'm hoping to play perfectionist by converting to both optimal cable and connectors. My motherboards have output pins so I can bypass the standard MB RCA jack if that would be considered an improvement.
1. Normal 3-6 ft interconnect.
2. 12-16 ft remote feeder from computer to DAC
3. Short internal link from input socket to DAC PCB.
I'm using a 16 ft Monster Interlink 300 MK II Audio Cable (measures 0.0 ohms) with RCA ends that get the job done. I'm hoping to play perfectionist by converting to both optimal cable and connectors. My motherboards have output pins so I can bypass the standard MB RCA jack if that would be considered an improvement.
It is not the "ohm" (resistive) which is the "bad part". The S/PDIF has a 75 ohm output, and input. So the interconnect cable should also be 75 ohm. If it is for example 50 ohm, you get reflection at the end of the cable, and these reflections distort the signal VERY BADLY.
Second: The "losses". The S/P DIF signal is 2-3 MHz. A "good" RG59 cable (Farnell 3711869) gives an attenuation of 2 - 2,3 dB/100 meters (303 feet). So if you have 1 or 2 meters of interconnect (3-6 ft), the losses are neglegible (+/- 0,3 dB). That is nothing for a signal. Probably the connector losses are as high as the cable losses.
And besides, high loss will lower the level of the "1" signal, but if it is still "seen" by the decoder or buffer as a HIGH signal, it will accept it anyway. Did you ever thought about the fact, that the distance between a mobile station and you mobile phone, does not affect quality? Even if you are close, or far, it is still good (do not go too far, then of course it is degraded, but the first mile or 2 it does not change).
The base station and the phone, still see the "1" and "0" and can decode them. It is just a rough comparising, but just to explain in "human language".
So it is useless to use a Monster cable for S/P DIF. RCA jacks on your MB does not change anything.
Use the Farnell cable, or any 75 ohm (video) cable with RCA jacks).
Second: The "losses". The S/P DIF signal is 2-3 MHz. A "good" RG59 cable (Farnell 3711869) gives an attenuation of 2 - 2,3 dB/100 meters (303 feet). So if you have 1 or 2 meters of interconnect (3-6 ft), the losses are neglegible (+/- 0,3 dB). That is nothing for a signal. Probably the connector losses are as high as the cable losses.
And besides, high loss will lower the level of the "1" signal, but if it is still "seen" by the decoder or buffer as a HIGH signal, it will accept it anyway. Did you ever thought about the fact, that the distance between a mobile station and you mobile phone, does not affect quality? Even if you are close, or far, it is still good (do not go too far, then of course it is degraded, but the first mile or 2 it does not change).
The base station and the phone, still see the "1" and "0" and can decode them. It is just a rough comparising, but just to explain in "human language".
So it is useless to use a Monster cable for S/P DIF. RCA jacks on your MB does not change anything.
Use the Farnell cable, or any 75 ohm (video) cable with RCA jacks).
Last edited:
Hi haik, so you don't think using BNC connectors are worth the effort? I have bunches of leftover male and female ends from some analog video workstations I built before digital took over.
Just to be a bit of a pest - are you saying there is really no practical way to test/confirm the "75 ohm" rating (particularly with short segments) with a simple DMM - just trust the ratings?
Just to be a bit of a pest - are you saying there is really no practical way to test/confirm the "75 ohm" rating (particularly with short segments) with a simple DMM - just trust the ratings?
BNC is effective in the "high" MHZ. Like 50-100 or higher Mhz. The SPDIF with it's 2-3 Mhz is way below that.
A simple DMM cannot measure anything in the RF part. For that you need a tone generator, oscilloscope (digital or analog), and a frequency counter.
And actually, what are you going to test? How the signal propagates? It does, other way's you would not hear anything.
Just buy a simple video cable with RCA plugs, that will be sufficient. Do not bother to test it. See the Farnell numbers in my previous post. Or go to any store that sells VIDEO cable (usually yellow RCA plugs). NOT audio (white and red).
A simple DMM cannot measure anything in the RF part. For that you need a tone generator, oscilloscope (digital or analog), and a frequency counter.
And actually, what are you going to test? How the signal propagates? It does, other way's you would not hear anything.
Just buy a simple video cable with RCA plugs, that will be sufficient. Do not bother to test it. See the Farnell numbers in my previous post. Or go to any store that sells VIDEO cable (usually yellow RCA plugs). NOT audio (white and red).
Just to be a bit of a pest - are you saying there is really no practical way to test/confirm the "75 ohm" rating (particularly with short segments) with a simple DMM - just trust the ratings?
It's an impedance measure, no way to do it with a multimeter, maybe with an LCR bridge you'll have more luck.
BTW regarding sound quality the digital cable could make a difference.
I was using an high quality HDTV 75 Ohm Belden video cable with crimped Canare 75 Ohm RCA (made by Blue Jeans Cables).
In theory, there's no better cable...
Well when I've swapped with a 75 Ohm Neotech OCC cable with CMC USA connectors (hand soldered by myself) the difference was clear...
The new cable was better for refinement, fullness of timbre, soundstage.
Before that Neotech I was fully satisfied with the Belden/Canare.
Bob, are you writing about short analog video cables? or are you only concerned about the connectors? If it's the cables, the manufactures cable number will have all the information and I think that most video cables were 75 Ohms. If it's old BNC connectors, then things get tricky fast. A few old 75 Ohm BNC's had different diameter pins.
Hi Dario,
Did you do a double blind test with the Neotech and the Belden cable?
It would be interesting to do a test, by sending an x amount of samples through the cable and count them with a digital counter.
Then replace the cable with a different one, and repeat the test.
And compare the difference.
Did you do a double blind test with the Neotech and the Belden cable?
It would be interesting to do a test, by sending an x amount of samples through the cable and count them with a digital counter.
Then replace the cable with a different one, and repeat the test.
And compare the difference.
Hi Haik,
No, I don't have the necessary hardware to do one and BTW I don't trust much on blind tests in audio.
When you try to discriminate between two 'samples' in a DBT, only obvious (and BIG) differences can be detected thus the test itself is often useless because the differences are so obvious at listening that no test is necessary... strictly IMHO.
This is a diagnostic test the tell us if the cable works, not how it (if) affects sound.
A cable that don't passes this test is simply not suitable.
No, I don't have the necessary hardware to do one and BTW I don't trust much on blind tests in audio.
When you try to discriminate between two 'samples' in a DBT, only obvious (and BIG) differences can be detected thus the test itself is often useless because the differences are so obvious at listening that no test is necessary... strictly IMHO.
This is a diagnostic test the tell us if the cable works, not how it (if) affects sound.
A cable that don't passes this test is simply not suitable.
Please no one mention DBT in this thread. This thread has always been about subjective listening reports, never about DBT. Please don't go there. It always results in a horrendous argument and closure of the thread. Thank you.
If you want to see if there is difference to the sound, try different cables in your own system - if you hear a difference there is a difference. If you don't, then there is no difference -in your system.
"Cable Wars" ?? No I don't want to go there either. All the information I needed was obtained from the posts answering the "can you measure..." question.
This supplier has provided good products/service, and I'm sure I will be able to pick a few options to try. I'll probably try Dario's recipe at some point mainly because his track record has been so good that I'm dying to get a chance to "thump" him.
The Monster audio cable I've been using was never intended to be permanent (used 2 for dual discrete sub amps in an old HT system). Now that I have the DAC upgraded and sounding great, it is time to revisit the connection options. The decision to use long or short, external or a one chassis integrated build, are based on decor/placement as much as the technical stuff on the cable itself. Knowing the basics gathered in the last day clears the options.
BTW, I got the other DAC kit to try a comparrison to this design. Only problem - the supplier neglected to include the 45 resistors in the shipment.
They're on the way.I won't dilute this thread with the results from this package, but it will be fun to hear what the WM8741 has to offer.
