Not sure I understand that. If you disconnect/remove the stage following, you have a 499 Ohm resistor cap-coupled off the previous op amp and driving into a 100K Ohm pot. For unbalanced home use, you don't need the gain of the following stage. 499 Ohm in series with 100K Ohm should be unnoticable and also consistent, as the buffer input after the pot will be very high Z. The only thing the 499 Ohm should do is be a load for when the mute transistor is on.
Am I missing something here?
Thanks,
Paul
Single ended
All of my amps except the BPA100 are not bridged/ balanced so unless I need to run 20 meter interconnects some day, balanced is not needed. The BPA100 can take full advantage of a balanced source but doesn't sound quite as good as the modified Sure 2X100 so I won't bother any more with BPA chip amps.
Yes - "everything goes" in audio - but sadly not always 
Just as a remark - balanced isn't exactly to be recommended with long lines *only* - see my rambling about doing balanced signal transfer within a a single piece of gear.
It all depends on radiating environment *and* the difference of mass potentials.
Seems you are a lucky guy in both respects...
Stay clear about bridged (and complimentary) versus floating balanced tho
Michael
Just as a remark - balanced isn't exactly to be recommended with long lines *only* - see my rambling about doing balanced signal transfer within a a single piece of gear.
It all depends on radiating environment *and* the difference of mass potentials.
Seems you are a lucky guy in both respects...
Stay clear about bridged (and complimentary) versus floating balanced tho
Michael
Last edited:
I got my 4396's from Nu Horizons: Nu Horizons Electronics Corp. - Electronic Components Distributor.
Not sure how long it would take them to get 4395's, though, as they're not regularly stocked. The 4396's are.
Not sure how long it would take them to get 4395's, though, as they're not regularly stocked. The 4396's are.
Last edited:
I woud say stay balanced any time you can. Single ended is there just because it is cheaper not because its better. It has no advantage just disadvantage vs. balanced.
I have plenty AK4395
They are sold in quantity from Nu Horizons or I can sell small quantities.Wow! Lots of things I want to comment on, but don't have time today.
Quick question. Does anyone know of a vendor that is selling the AKM4395 DAC chips? I'd like to get a few to try out in place of the AKM4393.
Gary
Note that the minimum quanitities from Nu Horizons are not that high though. I ordered 12 4396's, which came to about $54 with shipping. Of course, you only need three of them for a DCX....
IIRC, I could have ordered less, but since the shipping was about the price of 6 DAC's, regardless, I figured I'd get a few extra.
IIRC, I could have ordered less, but since the shipping was about the price of 6 DAC's, regardless, I figured I'd get a few extra.
Thanks for the input!
So, is there any advantage/disadvantage to using the 4396? With just a cursory look at the data sheets the only thing that seemed to jump out is lack of DSD on the 4395. I could be totally full of it though...
If the 4396 works I'll just order them as they are the stocked item. The first thing I would be installing them in is the Echo Layla 96/24, 2 DACs (4 channels worth) first so the 93s could be compaired against the 96s, then the other 4 channels to finish the project (4 DAC chips total). I've got a second Layla and a couple of Behringers that may get the treatment also if I like the results...
Gary
So, is there any advantage/disadvantage to using the 4396? With just a cursory look at the data sheets the only thing that seemed to jump out is lack of DSD on the 4395. I could be totally full of it though...
If the 4396 works I'll just order them as they are the stocked item. The first thing I would be installing them in is the Echo Layla 96/24, 2 DACs (4 channels worth) first so the 93s could be compaired against the 96s, then the other 4 channels to finish the project (4 DAC chips total). I've got a second Layla and a couple of Behringers that may get the treatment also if I like the results...
Gary
The 4396's are drop-in compatible with the 4393's. The 4395's don't work off of 3.3V, so to install them in a DCX you'd need to lift their PS legs and wire them up to 5V.
Either one gives you an improvement in internal filtering, which, more-or-less, at least in the DCX application, would allow you to get by without an external LP filter.
Either one gives you an improvement in internal filtering, which, more-or-less, at least in the DCX application, would allow you to get by without an external LP filter.
I'd certainly be interested in seeing if you can reliably hear a difference between the 3 and 6. And any reasonable measured difference too if you feel like making the effort.
THAT1200 balanced line receiver
Ever considered THAT1200 balanced line receiver?
THAT Corporation 1200-series InGenius High CMRR Balanced Line Receiver ICs
The specs look promising:
CMRR: typ. 90 dB at 60 Hz
Extremely high common-mode input impedance
Maintains balance under real-world conditions
Transformer-like performance in an IC
Excellent audio performance
Wide bandwidth: > 22 MHz
High slew rate: 12 V/us
Low distortion: 0.0005 % THD
Low noise: -107 dBu
Several gains: 0 dB, -3 dB, & -6 dB
Applications:
Balanced Audio Line Receivers
Instrumentation Amplifiers
Differential Amplifiers
Transformer Front-End Replacements
ADC Front-Ends
Basic Specifications (typical)
Input Voltage Range (Differential - equal & opposite swing)
1200: 21.5 dBu
1203: 24.5 dBu
1206: 24.5 dBu
Output Noise (BW = 20 kHz)
1200: -106 dBu
1203: -105 dBu
1206: -107 dBu
CMRR (matched source impedances) (Vcm = ±10V)
DC: 90 dB
60 Hz: 90 dB
20 kHz: 85 dB
Total Harmonic Distortion: 0.0005%
Positive Supply Voltage: +20 V
Negative Supply Voltage: -20 V
Input Voltage: ± 25 V
Output Short Circuit Duration: Continuous
Operating Temperature Range: 0 to +85 °C
I´ve ordered a sixpack and will try them. I´ve swapped the 4393s for 4396s, so I´ll hook them right on the outputs as no LPF is needed.
Ever considered THAT1200 balanced line receiver?
THAT Corporation 1200-series InGenius High CMRR Balanced Line Receiver ICs
The specs look promising:
CMRR: typ. 90 dB at 60 Hz
Extremely high common-mode input impedance
Maintains balance under real-world conditions
Transformer-like performance in an IC
Excellent audio performance
Wide bandwidth: > 22 MHz
High slew rate: 12 V/us
Low distortion: 0.0005 % THD
Low noise: -107 dBu
Several gains: 0 dB, -3 dB, & -6 dB
Applications:
Balanced Audio Line Receivers
Instrumentation Amplifiers
Differential Amplifiers
Transformer Front-End Replacements
ADC Front-Ends
Basic Specifications (typical)
Input Voltage Range (Differential - equal & opposite swing)
1200: 21.5 dBu
1203: 24.5 dBu
1206: 24.5 dBu
Output Noise (BW = 20 kHz)
1200: -106 dBu
1203: -105 dBu
1206: -107 dBu
CMRR (matched source impedances) (Vcm = ±10V)
DC: 90 dB
60 Hz: 90 dB
20 kHz: 85 dB
Total Harmonic Distortion: 0.0005%
Positive Supply Voltage: +20 V
Negative Supply Voltage: -20 V
Input Voltage: ± 25 V
Output Short Circuit Duration: Continuous
Operating Temperature Range: 0 to +85 °C
I´ve ordered a sixpack and will try them. I´ve swapped the 4393s for 4396s, so I´ll hook them right on the outputs as no LPF is needed.
AK4395 vs. 4396
.
http://www.diyaudio.com/forums/digi...ak4396-listening-comparisons.html#post1722952
.
This graphic of the ultrasonic noise spectrum is also interesting.
.
http://www.diyaudio.com/forums/digi...5-4396-ultrasonic-noise-spectrum-graphic.html
I posted a listening comparison between the AK4395 and 4396 running direct out into a foil cap here. My new amps are even more revealing now so it will be fun to check them again but I like the superior low level resolution of the 4395 even though it requires the addition of a 5v supply to pin 2.
.
http://www.diyaudio.com/forums/digi...ak4396-listening-comparisons.html#post1722952
.
This graphic of the ultrasonic noise spectrum is also interesting.
.
http://www.diyaudio.com/forums/digi...5-4396-ultrasonic-noise-spectrum-graphic.html
Good description I'd rely on if I'm ever on modding a DCX again - thanks.
Never changed a DAC but was surprised that the decoupling caps for the *digital* supply of DAC actually make for quite a difference if changed - and surprise - also audible benefit from break in
Well "quite a difference" meant in the sublime range such changes occur of course
- crazy audio world !
Michael
Last edited:
I've used it in several circuits and tried it briefly in the DCX output. It's a good chip, but I didn't pursue it enough in the DCX to have an opinion there.
Another chip I've used and like that might work is the TPA6120 headphone driver. It has differential inputs and 2 channels. Very clean and with plenty of drive. Layout is highly critical, tho.
If I get a chance, I'll try it for the output section.
For the input device for mine (going into the DEQ, since that then feeds digital to the DCX), I'm using the OPA1632. It automatically handles balanced or unbalanced input, and allows for the elimination of input coupling capacitors. It also simplifies the addition of the 2.5V DC offset to the ADC input.
Last edited:
Last minute modifications to order(s) took place.
I decided to go with Elna cerafine capasitors in analog board and digital bords caps that are nars DAC's and ADCs insted of Silmic IIs. I don't know why, I just did, cerafines were also little cheaper. Then after ordering them I decided to change every cap inside DCX. So I would also have to replace the 1µf anf 47µf caps at the PSU and all the capasitors near the wire on the digital board that comes from PSU. To replace these I got Silmic IIs.
I decided also to order everything for two DCXs since I will mod also the other if the difference is clearly audible in A/B comparison.
Btw Does it matter that 4393 has load resistance of 600 ohms and 4396 has 1k ohms?
(FYI, I called to local tv and radio repair shop and they are willing to change my DACs if I hesitate to change them myself.)
I decided to go with Elna cerafine capasitors in analog board and digital bords caps that are nars DAC's and ADCs insted of Silmic IIs. I don't know why, I just did, cerafines were also little cheaper. Then after ordering them I decided to change every cap inside DCX. So I would also have to replace the 1µf anf 47µf caps at the PSU and all the capasitors near the wire on the digital board that comes from PSU. To replace these I got Silmic IIs.
I decided also to order everything for two DCXs since I will mod also the other if the difference is clearly audible in A/B comparison.
Btw Does it matter that 4393 has load resistance of 600 ohms and 4396 has 1k ohms?
(FYI, I called to local tv and radio repair shop and they are willing to change my DACs if I hesitate to change them myself.
)
Might be, I didn't get the point of that OPA1632 but I see no more "differential" operation as from any other standard opamp - plus an additional inverted out
They not even spec the intrinsic CMRR !
Michael