Thank you

Hi B0bb,
 
I have some XpressO Ultra XO's - FXU-HC73x series which you maybe could add to the list?
Those are not the best specs, but maybe nice to have them in the list.
 
http://www.foxonline.com/pdfs/FXU-HC73.pdf
 
I have:
 
FXU-HC738 (20ppm) 22.5792mHz
FXU-HC736 (25ppm) 24.576 mHz
 
Mounted them on Ian Canada's XO boards:
 

 
Used some 3M AB5100S EMI/RFI absorbing material:
 

 
Finally put Vishay 1837 0.01uF MKP's on top of them:
 

 
 
They are burnin-in in Ian's FIFO II - Isolator - DUALXO II reclocker, and, so far so good
 
Use them, for testing, with a low-cost Chinese AK4495SEQ I2S-Only dac, together with Bursun V5i-D opamp (for reviewing)
 

 
 

 
RPI2 using piCorePlayer 2.06, I2S output from GPIO thru Ian Canada's I2S output adapter fed into FIFO II.
 
Sound so far: Astonishing!
This set almost sounds like my "MonsterDAC" which, normally is a HUGE difference with this low-cost chinese dac, when fed I2S thru RPI+HB DAC+Pro
 
Btw, SS V5i-D does a very good job too Will review them when burned-in, next weekend.
 
Finally Ian's FIFO II + Isolator + DualXO II board will be built-in in MonsterDAC. Can't wait, that's for sure
 
 
Regards,
Alex

abartels said:

  Hi B0bb,
 
I have some XpressO Ultra XO's - FXU-HC73x series which you maybe could add to the list?
Those are not the best specs, but maybe nice to have them in the list.
 
http://www.foxonline.com/pdfs/FXU-HC73.pdf
 

Click to expand...

Here are the numbers.
 
 
The r.m.s. phase jitter is the area under the phase noise plot averaged over the frequency span.
The FXU-HC73 number is for 156.25MHz as described in the datasheet.
 
 
*The calculated control numbers have an integration bandwidth of 10kHz-40MHz, this gives a marginally higher r.m.s jitter of 268ps vs the datasheet 260ps which used a bandwidth of 12kHz-20MHz

Hi B0bb,
 
Thanks for adding them!
 
 
Regards,
Alex

Hi
 
First post on here, so go easy on me!
 
I've just acquired an LKS Audion MH-DA003, which has had certain modifications performed on it (the details of which I don't have yet).
 
On listening to this DAC for the first time, I was a little disappointed in how it sounded compared to the modified Eastern Electric Minimax Supreme which I have owned and enjoyed for the last two years.
 
With the EE Supreme, I had replaced the original opamps (NE5532 and NE5534 IC's) with single and dual discrete op amps from Dexa. The following link shows very similar (or possibly the same) dual op amps:
http://www.newclassd.com/index.php?page=122&hv=1
 
My MH-DA003 arrived with two pairs of Texas Instraments IC op amps installed. The print is very small, but a magnifying glass shows TI 1612A, so I suppose they are similar to those in the link below:
http://www.ti.com/product/opa1612Hi
 
Today I replaced the pair of 1612s located nearest to the unbalanced outputs with the Dexa's that I used in the Supreme. The sound is now less gritty and less thin compared to how it was with those TI IC op amps inline.
 
As a result of this exercise, I have come up with some questions:
 
1) I'm not very good at all at matching op amp specs to real world performance. The link I posted to the Dexa Op Amp describes supply current as: "typ. 14mA". Does this mean that this op amp drops out of Class A operation after 14mA?
 
2) After replacing the first pair of TI 1612 IC op amps, I decided as a test to remove - and not replace - the remaining pair of 1612's (those located nearest to the 9018 chips). The DAC still operated ok through the unbalanced outputs. So, what is the function of the op amps I removed? Do they simply service the balanced outputs?
 
3) The Dexas sound good to me. I also have a pair of Burson discrete dual Op Amps, which I will try later. I know it is a subjective thing, but from a technical perspective would any of you recommend against using the Dexas with the MH-DA003 based on their spec?
 
4) I don't understand why my EE Minimax Supreme uses a pair of single op amps and a pair of dual op amps, whilst the MH-DA003 I acquired has two pairs of dual op amps. Is there a simple explanation about when to use single and when to use dual?
 
Sorry if my questions are a bit ignorant, I'm not an Electronics Engineer, I'm more a hifi buff with an interest in and rudimentary understanding of electronics.
 
Thanks!

penguin140369 said:

   
1) I'm not very good at all at matching op amp specs to real world performance. The link I posted to the Dexa Op Amp describes supply current as: "typ. 14mA". Does this mean that this op amp drops out of Class A operation after 14mA?
 
2) After replacing the first pair of TI 1612 IC op amps, I decided as a test to remove - and not replace - the remaining pair of 1612's (those located nearest to the 9018 chips). The DAC still operated ok through the unbalanced outputs. So, what is the function of the op amps I removed? Do they simply service the balanced outputs?
 
3) The Dexas sound good to me. I also have a pair of Burson discrete dual Op Amps, which I will try later. I know it is a subjective thing, but from a technical perspective would any of you recommend against using the Dexas with the MH-DA003 based on their spec?
 
4) I don't understand why my EE Minimax Supreme uses a pair of single op amps and a pair of dual op amps, whilst the MH-DA003 I acquired has two pairs of dual op amps. Is there a simple explanation about when to use single and when to use dual?
 

Click to expand...

Comments to your questions
1) I normally estimate the output stage bias current at 80% of the supply current. This gives a bias current of about 5.5 mA, the opamp falls out of ClassA at 2X this value or about 11mA of load current.
(The DEXA you mentioned is a dual opamp so the supply current is 7mA for each half).
To put things is perspective, the ES9018 has 8 DACs each producing 4mA current output at full scale.
On the Dual Mono (8 DACs/ch) LKS, this is 32mA, on the EE (4 DACs/ch) this is 16mA.
The DEXA does not appear to run in ClassA in any of the situations, the special edition that runs at 20mA supply current will work as the I/V converter on the EE but not the LKS.
 
2)You removed the Current to Voltage (I/V) converter, the DAC is still connected via the I/V resistor so it continues to work in Voltage output mode instead of current output mode.
 
3)The DEXAs will do ok in the unbalanced stage, less so for the I/V role for the reasons explained in (1)
 
4)The EE uses a single pair of opamps, a pair of dual opamps, a pair of triodes and a bunch of capacitors that appear to be in the signal path. The LKS is a conventional DC coupled active filter with no caps directly in the signal path, can't say more without seeing the schematic of the EE

Brilliant, thanks b0bb. Appreciate the response.
 
The DAC actually sounds better in voltage output mode with only the discrete Dexa opamps in place than it does in current mode with the two TI 1612s in place.
 
I will put one pair of TI chips back in for the I/V coversion, as the datasheet shows 55mA as the typical output current which, with my limited knowledge, would appear to be up to the job. Later this week I will look back through the thread and find something better to use - I think I saw Sonic Imagery Ticha 994's recommended for their high idling current.
 
This should all be a good experiment. I thought the LKS would beat the EE Supreme hands down out of the box due to it employing twin ESS9018's vs the Supreme's single chip ... but this was obviously a simplistic notion as it doesn't get close, even with the mods built in to the LKS (which I am not familiar with yet but will learn about and then compare to the recommendations here).

penguin140369 said:

 
Brilliant, thanks b0bb. Appreciate the response.
 
The DAC actually sounds better in voltage output mode with only the discrete Dexa opamps in place than it does in current mode with the two TI 1612s in place.
 
I will put one pair of TI chips back in for the I/V coversion, as the datasheet shows 55mA as the typical output current which, with my limited knowledge, would appear to be up to the job. Later this week I will look back through the thread and find something better to use - I think I saw Sonic Imagery Ticha 994's recommended for their high idling current.
 
This should all be a good experiment. I thought the LKS would beat the EE Supreme hands down out of the box due to it employing twin ESS9018's vs the Supreme's single chip ... but this was obviously a simplistic notion as it doesn't get close, even with the mods built in to the LKS (which I am not familiar with yet but will learn about and then compare to the recommendations here).

Click to expand...

The I/V converter is also the first stage of the analog filter, it gets rid of the Delta-Sigma conversion artifacts above 1MHz.
 
Over the long term this can do a lot of damage to specific types of speakers especially things like beryllium tweeters if the first stage is removed.

Thanks again b0bb. I think the tweeters in my Kudos speakers are about $500 a piece for replacements so that is very valuable information.
 
I restored the 1612's late yesterday and it compromised the SQ. I was told by the seller of DAC that these op amps were good quality and up to the job, however my ears are telling me otherwise. With the 1612's in place, the sound has that slightly brittle digital signature, making the music unrelaxing, and - oddly - there is something missing in the mid-range: vocals sound thin and unnatural. My hearing isn't great so I am surprised that I am picking these things up.
 
I looked at one of the pics earlier in this thread and it seems, b0bb, that you have settled for Bursons on the unbalanced stage, with (what look like) the Sonic Imagery discrete op amps for the I/V conversion. I will put in an order for a pair of the latter. (I also have some discrete dual Burson op amps at home but they are older, although they look somewhat like the v4's.)
 
Tomorrow I will look again at the board to figure out the path for the balanced signal, which I would like to test out also.
 
I am still a little bit flummoxed as to why my single chip MiniMax DAC is outperforming the MH-DA003 DAC with its two chip architecture and seemingly clean design. Let's hope persistence and further tweaking pays off...

Out of curiosity, has anyone compared this to something like the gustard x20u?
I have a x20u, but am considering this unit.
 
what are some dacs that this has been compared against?

penguin140369 said:

 
Thanks again b0bb. I think the tweeters in my Kudos speakers are about $500 a piece for replacements so that is very valuable information.
 
I restored the 1612's late yesterday and it compromised the SQ. I was told by the seller of DAC that these op amps were good quality and up to the job, however my ears are telling me otherwise. With the 1612's in place, the sound has that slightly brittle digital signature, making the music unrelaxing, and - oddly - there is something missing in the mid-range: vocals sound thin and unnatural. My hearing isn't great so I am surprised that I am picking these things up.
 
I looked at one of the pics earlier in this thread and it seems, b0bb, that you have settled for Bursons on the unbalanced stage, with (what look like) the Sonic Imagery discrete op amps for the I/V conversion. I will put in an order for a pair of the latter. (I also have some discrete dual Burson op amps at home but they are older, although they look somewhat like the v4's.)
 
Tomorrow I will look again at the board to figure out the path for the balanced signal, which I would like to test out also.
 
I am still a little bit flummoxed as to why my single chip MiniMax DAC is outperforming the MH-DA003 DAC with its two chip architecture and seemingly clean design. Let's hope persistence and further tweaking pays off...

Click to expand...

 
The brittle sound quality comes from the CCHD-575 XO used in the LKS, putting a better quality opamp will bring this to the foreground.
 
Opamp replacement is easy but does not return much on that investment until the XO is replaced with something better like the CCHD-950X, this is something to keep in mind before putting down the cash for a  Ticha 994.

b0bb said:

   
The brittle sound quality comes from the CCHD-575 XO used in the LKS, putting a better quality opamp will bring this to the foreground.
 
Opamp replacement is easy but does not return much on that investment until the XO is replaced with something better like the CCHD-950X, this is something to keep in mind before putting down the cash for a  Ticha 994.

Click to expand...

Thanks as usual b0bb. The LKS I have acquired already has an upgraded oscillator amongst its mods - I'll check which one it is and report back, I expect it is on the list that exists in this thread.

penguin140369 said:

 
Thanks as usual b0bb. The LKS I have acquired already has an upgraded oscillator amongst its mods - I'll check which one it is and report back, I expect it is on the list that exists in this thread.

Click to expand...

Back again.
 
So, the oscillator mod appears to be as follows: the stock XO (CCHD575) has been lifted out from the PCB and moved to a small separate board providing dedicated power regulation.
 
So good call on the stock sound b0bb, you spotted the XO sound just from my description.
 
Looking through this thread, it looks like the Pulsar XO is the preferred replacement. I will see how I can source it.
 
Access to the XO and the small pins means I cannot perform this mod with my current, rather basic soldering iron. Any recommendations for an iron I can use for work with these kind of smaller components?
 
Also, and on a separate note, does anyone have the pinout for the ESS9018? It seems to be a closely guarded secret. I've looked before and not had much luck. it would help me understand better what is going on.

Ok, is the low jitter pulsar clock in the XO table on page 20 of this thread referencing this OCXO below:
 
http://www.pulsarclock.com/Clock.html
 
Or is it referencing another one?
 
(Sorry for all the questions)

penguin140369 said:

 
Back again.
 
So, the oscillator mod appears to be as follows: the stock XO (CCHD575) has been lifted out from the PCB and moved to a small separate board providing dedicated power regulation.
 
So good call on the stock sound b0bb, you spotted the XO sound just from my description.
 
Looking through this thread, it looks like the Pulsar XO is the preferred replacement. I will see how I can source it.
 
Access to the XO and the small pins means I cannot perform this mod with my current, rather basic soldering iron. Any recommendations for an iron I can use for work with these kind of smaller components?
 
Also, and on a separate note, does anyone have the pinout for the ESS9018? It seems to be a closely guarded secret. I've looked before and not had much luck. it would help me understand better what is going on.

Click to expand...

 
Post pictures of your LKS and I can take a closer look.
 
If the XO has been put on a separate board thinks are easier, just drop in  a 14pin dip socket and solder it in.
 
Here is mine.

 
 
The ES9018 datasheet is around, google for it with the keyword "ES9018 datasheet"