... can not tell, since I've got only the ESSTECH recommended AD797, so I don't know about the diff towards the LMEs ... maybe worth a try ...
... sometimes I had the locking issue as well ... could be that my settings in Voodoo2000 were somewhat off ... played with the settings and found it to be cumbersome on the long run, so I changed to JRiver.
After using the windows one, it played quite nice, switch to OSX with JRiver as well as using JRemote on the iPad Air, that just played very nice ...
currently using the linux variant (still in beta, if not alfa) playing along very nice as well ...
since I used that setup the locking issues where kind of history give or take a few exceptions.
... sometimes I had locking issue in the middle of a song, or at the beginning, stopping the song and start it again mostly solved it. This is most happening with switching to DSD or from DSD.
What might be a cause can be, but must not be, a bandwidth issue if using wifi instead of cabled LAN. At least that's what happened sometimes as well. Having your house kind of full wifi users (almost a lan party going on) their might some bandwidth limitations. 3 Kids and the better half of me do take up some bandwidth if playing along ....
That all got away once using a cabled connection ... plays quite nice I must say for a beta piece of software (almost two weeks without a glitch at the moment)
BTW, does somebody have any idea/experience how the AD797 compare to an LMExxxxx theoretically and soundwise .... ???
Edit: Do you have the Amanero up to date concerning the firmware ?
... sometimes I had the locking issue as well ... could be that my settings in Voodoo2000 were somewhat off ... played with the settings and found it to be cumbersome on the long run, so I changed to JRiver.
After using the windows one, it played quite nice, switch to OSX with JRiver as well as using JRemote on the iPad Air, that just played very nice ...
currently using the linux variant (still in beta, if not alfa) playing along very nice as well ...
since I used that setup the locking issues where kind of history give or take a few exceptions.
... sometimes I had locking issue in the middle of a song, or at the beginning, stopping the song and start it again mostly solved it. This is most happening with switching to DSD or from DSD.
What might be a cause can be, but must not be, a bandwidth issue if using wifi instead of cabled LAN. At least that's what happened sometimes as well. Having your house kind of full wifi users (almost a lan party going on) their might some bandwidth limitations. 3 Kids and the better half of me do take up some bandwidth if playing along ....
That all got away once using a cabled connection ... plays quite nice I must say for a beta piece of software (almost two weeks without a glitch at the moment)
BTW, does somebody have any idea/experience how the AD797 compare to an LMExxxxx theoretically and soundwise .... ???
Edit: Do you have the Amanero up to date concerning the firmware ?
Last edited:
Yes my Amanero is up to date with regards to firmware.
I have attached some key data of the Op amps we have been talking about including the AD797.
The LME49990 appears to be best for noise,thd and pssr if you believe these factors are important for IV. On the other hand if you believe it is the slew rate, settling time and output current that are important, than the AD811 seems to have the best performance.
For this reason, I tried both the LME49990 and the AD811. To be fair, I found the tonality of the two to be radically different. The LME49990 sounds very clean and dynamic whilst the AD811 sounds a bit bassier but does not have the clarity and preciseness of the LME49990.
To make the AD811 stable you have to use a 1k resistor in series with its negative input. I am unsure as to whether this impedance is changing the mode of operation of the ES9018 from current to voltage mode and hence degrading the overall performance? A view from the experts would be very welcome.
Please post your own experiences with Opamps to see whether we can correlate this with the technical data.
Edit just spotted an error in the spreadsheet. All the noise figures are in nV/sqrHz including the LME's but the spreadsheet says V/sqHz for these.
I have attached some key data of the Op amps we have been talking about including the AD797.
The LME49990 appears to be best for noise,thd and pssr if you believe these factors are important for IV. On the other hand if you believe it is the slew rate, settling time and output current that are important, than the AD811 seems to have the best performance.
For this reason, I tried both the LME49990 and the AD811. To be fair, I found the tonality of the two to be radically different. The LME49990 sounds very clean and dynamic whilst the AD811 sounds a bit bassier but does not have the clarity and preciseness of the LME49990.
To make the AD811 stable you have to use a 1k resistor in series with its negative input. I am unsure as to whether this impedance is changing the mode of operation of the ES9018 from current to voltage mode and hence degrading the overall performance? A view from the experts would be very welcome.
Please post your own experiences with Opamps to see whether we can correlate this with the technical data.
Edit just spotted an error in the spreadsheet. All the noise figures are in nV/sqrHz including the LME's but the spreadsheet says V/sqHz for these.
Last edited:
I have used LME49990 (L49990MA on adaptor board) for a long time before switching to 49713HA. They were my favorite as well, the sound signatures are very close. To me LME49713 is more fuller with equal transparency, perhaps due to its higher drive capability. I had excessive heat buid up, killing some of the chips in a given time. I ended up with buying a big quantity from Mouser, adapter boards and pins to make it cheaper and keep up with the losses! ( later I used 600 ohm resistors to reduce the oscillation with some level of success ). LME49713 handles heat and I actually got to like them more then 49990s.
FLAC --- I did try AD797 just out of curiosity ( may be just like you, reading the ESS's recommendation, I don't remember much, though) but I can tell you that I replaced them more or less after an hour of testing, with 49990s. If I were you I wouldn't bother.
I've tried lm49710, lm6171, OPA627, 637 .. and many others. None of them are bad chips, but they have different tonal characteristics for sure.
FLAC --- I did try AD797 just out of curiosity ( may be just like you, reading the ESS's recommendation, I don't remember much, though) but I can tell you that I replaced them more or less after an hour of testing, with 49990s. If I were you I wouldn't bother.
I've tried lm49710, lm6171, OPA627, 637 .. and many others. None of them are bad chips, but they have different tonal characteristics for sure.
nah, too late, I just ordered a couple of lme49990s 
just to hear if there's difference ....
I've got a real crispy clean sound through my AD797s, maybe a bit too clinical but that could be caused by the other parts as well.
too bad I can't let you hear it, it's a 384 kbps single track, size 600+ MB, maybe upsampled though .... ?
If I close my eyes and just listen to the voice I almost think she sings just in front of me ... at the moment it's my reference ...
Don't think I could get it uploaded or
... well, actually you could try and load it from here in the 384 DXD version ...
http://designwsound.com/dwsblog/recordings/audiophile-jazz-prologue-3-free-hi-res-samples/
just to hear if there's difference ....
I've got a real crispy clean sound through my AD797s, maybe a bit too clinical but that could be caused by the other parts as well.
too bad I can't let you hear it, it's a 384 kbps single track, size 600+ MB, maybe upsampled though .... ?
If I close my eyes and just listen to the voice I almost think she sings just in front of me ... at the moment it's my reference ...
Don't think I could get it uploaded or
... well, actually you could try and load it from here in the 384 DXD version ...
http://designwsound.com/dwsblog/recordings/audiophile-jazz-prologue-3-free-hi-res-samples/
Last edited:
For pure CFB amps like LME49713 there are other tricks you can play which are outlined in this appnote from TI : http://www.ti.com/lit/an/slyt099/slyt099.pdf
Using the ferrite bead in series with the -ve input isn't something I've tried myself, but I sure plan to
For those using CFB, this post and article goes a long way towards explaining my subjective listening experiences with AD811 CFB Opamp. Previously using a 1k resistor to stabilise the opamp, I think it was a double whammy of the increased noise levels due the resistor and possible the resistance forcing the Sabre chip to operate in "voltage" mode.
I liked the idea in the article of using an inductor of which I had some 1uH ones in my part box. This would give an impedance of 628 Ohms at 100MHz. I tried this in the circuit but the impedance was too low and the op amp was unstable. I guess this circuit would need an impedance of around 1k at 10MHz to be on the safe side so possibly an inductance of around 16uH would do the job?
I the tried some AG grade Murata Ferrite beads with an impedance of 600 Ohms at 100MHz which I had in my parts box. These worked and the Op amp is stable. These are a big improvement on the 1k resistor in subjective SQ terms. For the first time using this opamp, I felt I was hearing the benefits of the high output current, the massive slew rate and extremely fast settling time. The presentation to my ears is "weightier" than the LME49990 and the bass response/dynamics are in a different league.
I need to listen for an extended period to make up my mind as to which presentation I prefer but if you are already using a CFB opamp, the ferrite bead is a big improvement over a stabilisation resistor in this application.
Thanks Abraxalito for posting that article.
Last edited:
I have experimented with the LME49713 CFB opamp. Unfortunately, this is not stable with a Murata AG grade 600 ohm Ferrite Bead. It appears to be stable with a 680 Ohm resistor but because of the issues stated in the TI article on the above post I am going to try a HG grade ferrite with a higher impedance.
I have also tried overclocking the chip with a 125Mhz SAW Epson clock as strongly recommended by a couple of people in other threads. Funnily enough, there is no crackling with the single ended output but a lot when I use the balanced output (through the evaluation board recommended LPF circuit). I can only assume that overclocking increases noise breakthrough so not recommended and possibly the recommended passive LPF filter in the balanced mode is not as aggressive as the single ended Op amp based one.
I have also tried overclocking the chip with a 125Mhz SAW Epson clock as strongly recommended by a couple of people in other threads. Funnily enough, there is no crackling with the single ended output but a lot when I use the balanced output (through the evaluation board recommended LPF circuit). I can only assume that overclocking increases noise breakthrough so not recommended and possibly the recommended passive LPF filter in the balanced mode is not as aggressive as the single ended Op amp based one.
I drive Cambridge Audio 851W power amplifier directly with this DAC. I tried OPA134 and OPA827 for I/V and LME49710 and LME49990 for output.
Despite I feel I have little balance problem with left channel, the sound is clean and detailed but little bit dry. I tried Eli Duttman's 12B4 line amplifier as preamplifier and the sound became more alive and punchy but it was not as clean as before.
I use 1 meter gold plated interconnect cable between the DAC and 851w. I was thinking to add a preamplifier between DAC and 851w with using one of CFB video opamps like AD815 which should drive capacitive load of the cable much better until I saw your trial with AD811. I am not so experienced to try AD811 for IV stage but it would be good idea to try AD811 for output stage instead of using separate preamplifier. Is there anyone who tried it? Any suggestions?
Despite I feel I have little balance problem with left channel, the sound is clean and detailed but little bit dry. I tried Eli Duttman's 12B4 line amplifier as preamplifier and the sound became more alive and punchy but it was not as clean as before.
I use 1 meter gold plated interconnect cable between the DAC and 851w. I was thinking to add a preamplifier between DAC and 851w with using one of CFB video opamps like AD815 which should drive capacitive load of the cable much better until I saw your trial with AD811. I am not so experienced to try AD811 for IV stage but it would be good idea to try AD811 for output stage instead of using separate preamplifier. Is there anyone who tried it? Any suggestions?
Hello Skyfighter
For your channel imbalance problem, try to measure all of the resistors in the signal path in the left channel and right channel and make sure that they are reasonably well matched between the channels, say less than 1%. Paricularly important for the IV resistors (680 Ohms). Also I found in my setup that if for any reason, one of the four IV opamps was not functioning properly (eg. dry joint, op amp not seated properly, faulty opamp etc. than it manifested itself as a channel imbalance problem. I suspect you have an issue with one of your opamps rather than a resistor mismatch but worth checking both avenues out.
The output stage has capacitance in its feedback path so I am pretty certain that CFB opamps such as AD811/LME49713 will be unstable as a direct replacement. Lets hope an OP amp Guru on this forum is able to advice you on what circuit changes to make to make it stable. The output current capability of these Op amps is brilliant at 100mA so it should have no problem at all driving the capacitive load of your cables.
For your channel imbalance problem, try to measure all of the resistors in the signal path in the left channel and right channel and make sure that they are reasonably well matched between the channels, say less than 1%. Paricularly important for the IV resistors (680 Ohms). Also I found in my setup that if for any reason, one of the four IV opamps was not functioning properly (eg. dry joint, op amp not seated properly, faulty opamp etc. than it manifested itself as a channel imbalance problem. I suspect you have an issue with one of your opamps rather than a resistor mismatch but worth checking both avenues out.
The output stage has capacitance in its feedback path so I am pretty certain that CFB opamps such as AD811/LME49713 will be unstable as a direct replacement. Lets hope an OP amp Guru on this forum is able to advice you on what circuit changes to make to make it stable. The output current capability of these Op amps is brilliant at 100mA so it should have no problem at all driving the capacitive load of your cables.
Thank you themystical, you are right, I will check the resistor values.
Yes, it would be very nice to implement CFB opamp for the output stage. I hope someone will have idea.
I will also try using preamplifier option so I ordered some AD815s to try as line amplifier between DAC and power amplifier. I am curious if it can drive the power amp as 12B4 can do but with less distortion.
Yes, it would be very nice to implement CFB opamp for the output stage. I hope someone will have idea.
I will also try using preamplifier option so I ordered some AD815s to try as line amplifier between DAC and power amplifier. I am curious if it can drive the power amp as 12B4 can do but with less distortion.
Can someone attach photos of both side of the working board, finished with all components? Trying to figure out wich smd capacitors to paralell with electros. I've already done with chip AVCC bypass, onboard 1,2V regulator but is worth anywhere else?
Beginner's work in progress going on so any photos should be useful..
Glt, did you finish your board? Thanks for the huge resources on HIFIDUINO!
KlipschKid, thank goes to you for the info on your blog! Maybe you could put more photos of finished board
And BIG thanks to all of you for making my first diy DAC happening!
Thanks!
Beginner's work in progress going on so any photos should be useful..
Glt, did you finish your board? Thanks for the huge resources on HIFIDUINO!
KlipschKid, thank goes to you for the info on your blog! Maybe you could put more photos of finished board
And BIG thanks to all of you for making my first diy DAC happening!
Thanks!
Hi Guys
I am still trying to find a solution for utilising CBB opamps in the IV stage in this design. Both the AD811 and the LME47913 can be stabilised by using the correct ferrite bead or resistor in the negative input as I posted before (in that the opamps do not get too hot). The only problem I have found is that the DC offset after the opamp seems to rise greatly and varies from opamp to opamp. Normally the DC offset before and after the opamp is 1/2 AVCC but with the CFB Opamps, it can be as high as 12-13 Volts.
Is this because the Opamps are oscillating or is it that I need to put the same resistance or ferrite bead in the positive input to the Op amp to make sure that the offset nulling circuitry is effective?
Any help appreciated as the CFB Opamps provide the best sound but I can't use them because the high DC offset is incompatible with my Ncores which run balanced.
I am still trying to find a solution for utilising CBB opamps in the IV stage in this design. Both the AD811 and the LME47913 can be stabilised by using the correct ferrite bead or resistor in the negative input as I posted before (in that the opamps do not get too hot). The only problem I have found is that the DC offset after the opamp seems to rise greatly and varies from opamp to opamp. Normally the DC offset before and after the opamp is 1/2 AVCC but with the CFB Opamps, it can be as high as 12-13 Volts.
Is this because the Opamps are oscillating or is it that I need to put the same resistance or ferrite bead in the positive input to the Op amp to make sure that the offset nulling circuitry is effective?
Any help appreciated as the CFB Opamps provide the best sound but I can't use them because the high DC offset is incompatible with my Ncores which run balanced.
I have got to the bottom of my DC offset issues. Unreliable connections between the Op amp adapter boards and the DIL sockets that I have fitted.
There are no issues with either stability or DC offsets with all of the above CFB opamps. In terms of SQ, both the AD811 and the LME49713 CFB opamps sound much better than any of the VFB opamps that I have tried. They don't sound "bright" or "clinical"as LME49990 or LME49710 but have a very full sound without sacrificing any detail. The bass performance (which is not much noticeable with headphones but very apparent in my speaker system) is the best I have ever heard. With these CFB opamps you hear as well as feel the bass whereas with the VFB ones, you just hear it. Down to the 100mA output current or is it the toplogy, I dont know.
There are no issues with either stability or DC offsets with all of the above CFB opamps. In terms of SQ, both the AD811 and the LME49713 CFB opamps sound much better than any of the VFB opamps that I have tried. They don't sound "bright" or "clinical"as LME49990 or LME49710 but have a very full sound without sacrificing any detail. The bass performance (which is not much noticeable with headphones but very apparent in my speaker system) is the best I have ever heard. With these CFB opamps you hear as well as feel the bass whereas with the VFB ones, you just hear it. Down to the 100mA output current or is it the toplogy, I dont know.
I have noticed contact problems due to the test setup using sockets, I can confirm that. Somehow small surface oxidization causing bad contact, or heat related bad contact is the culprit in my system in my experience. I twist opamps little bit, heat dissipation and imbalance go away, awhile though! I'm going to solder those in the end but I can't stop using it as it is, it is so good!
I have been following your posts about the compensation beads at -Vcc with interest.
What is your conclusion? Are they needed, or do they improve heat buildup and stability?
Thanks for patiently experimenting on this and sharing your experiences!
I have been following your posts about the compensation beads at -Vcc with interest.
What is your conclusion? Are they needed, or do they improve heat buildup and stability?
Thanks for patiently experimenting on this and sharing your experiences!
I think the DIL sockets I am using are not really designed to cope with both Opamp pins and adapter pins (which are bigger and thicker). I have found that after changing opamps a few times, the clamping in the contacts reduce. I will try some contact cleaner in the first instance.
With regard to the ferrite bead, this goes in Pin 2, not the negative supply. Once you do this, there is no oscillation.....the opamp runs not even warm.
If you are not using a ferrite bead or a resistor in series with PIn 2, your opamp is definitely oscillating and running very hot. The oscillation will also cause some distortion in the SQ so your SQ can be improved by fixing this.
Have you ever tried using beads on both input pins? On the other hand, are you using the beads on pin2(inv. input) of the last stage opamp only?
I did use 2 600 ohm resistors on another diyinhk ES9018 board's last opamps which is the earliest version, I did not notice any sonic improvements. I have dismantled it to built a transformer based IV stage using a triode amp stage or buffer.
Lundahl has some input/inter-stage transformers, I tried one(LL1540) in a line stage to input SE inputs as balanced inputs with success. It came out good, so I intend to try same transformer or anther one, preferably 1:2 into a tube amp. and abandon the whole ss based IV and outputs, but in near future, I'm busy at the moment.
Apart from this, I believe you like 49713s probably you will go for them eventually... That is the same over here, I like 49990s too. Both opamps are very close in sonic signature to me.
Thanks for the posts again.
I did use 2 600 ohm resistors on another diyinhk ES9018 board's last opamps which is the earliest version, I did not notice any sonic improvements. I have dismantled it to built a transformer based IV stage using a triode amp stage or buffer.
Lundahl has some input/inter-stage transformers, I tried one(LL1540) in a line stage to input SE inputs as balanced inputs with success. It came out good, so I intend to try same transformer or anther one, preferably 1:2 into a tube amp. and abandon the whole ss based IV and outputs, but in near future, I'm busy at the moment.
Apart from this, I believe you like 49713s probably you will go for them eventually... That is the same over here, I like 49990s too. Both opamps are very close in sonic signature to me.
Thanks for the posts again.
- Status
- Not open for further replies.