[demos]

@b0bb Do you thing how big the difference would be between the AOCJY sc-cut OCXO and pulsar clock? Do you think the crystal they use has similar spec, Thanks in advance!

 

[Whitigir]

Minimum 0.1uF 10V or higher surface mount MLCC cap.
You have to measure the distance between the pads on the RHEA and look up the right part on Mouser

ESR is too high at 100MHz, MLCC caps are made in very thin stacked layers to reduce the inductance which lowers ESR.
MLCC caps are made for noise decoupling.

Thanks @b0bb , in is my Murata MLCC. I especially love murata Q_Q, and I don’t even know why. This little lad here is 0.1uf and 50V

https://www.mouser.com/ProductDetail/81-GRM40X104K50L

Though, I notice when at cold turn up, the lks004 displayed unlock into lock BW01 and 44khz, then unlock into lock at 48khz, then unlock into lock at 176khz. I quickly played music upsample to 176khz, and there was no Unlock or stuttering at all. Just beautiful music. Is this normal ? ***must be normal as I am connecting via I2S and it only becomes unlock when my PC is off, or powering on and goes through booting steps

 

[Whitigir]

Are you asking about the PLL phase lock range on the 9038?
(That kind of info is not public)

I do not fully understand the qestion.

Yes, that is what I am asking. I wonder if the ES9038Pro can be latched onto as high precision as those Modules by MSB themselves as their galaxy clock is 6k clock Q_Q

 

[xiamen]

Not fast enough !



So...what is better than that little abracon ? A larger abracon ! 2.5x more expensive ! Also 5X more in stability of 10PPB and not 50PPB.

Go in Frankenstein’s style !! Still Dropped-in !!!

This one is from digikey and it requires some works with the twisted pear audio adapter (sugru fix everything !). Terminated differently ! So you can see me running hard-silver all around X_X. This one started cold crank of 3W and is operating at 1W, which is well within the regulators capability !

https://www.digikey.com/product-detail/en/abracon-llc/AOCJY2-100.000MHZ-E/535-10666-ND/2441432

Yes! Once again, she upscaled with Pinkfaun I2S! Wow, I never would have thought the lks004 has so much more Juices to be squeezed out like this ! Totally craziness ! This is definitely the heart of LKS004 !

Not fast enough !



So...what is better than that little abracon ? A larger abracon ! 2.5x more expensive ! Also 5X more in stability of 10PPB and not 50PPB.

Go in Frankenstein’s style !! Still Dropped-in !!!

This one is from digikey and it requires some works with the twisted pear audio adapter (sugru fix everything !). Terminated differently ! So you can see me running hard-silver all around X_X. This one started cold crank of 3W and is operating at 1W, which is well within the regulators capability !

https://www.digikey.com/product-detail/en/abracon-llc/AOCJY2-100.000MHZ-E/535-10666-ND/2441432

Yes! Once again, she upscaled with Pinkfaun I2S! Wow, I never would have thought the lks004 has so much more Juices to be squeezed out like this ! Totally craziness ! This is definitely the heart of LKS004 !

Lol, is that a"dropped-in"? I think the abracon is too much work for me, I'll be favorite to the first to brick my lks 004. What I have in mind is to start by upgrading to the 575x but to do it in such a way that at a later stage, I can have a real dropped in upgrade to the Pulsar (to avoid having to upgrade vr and caps?). From what I read here, I will have to other than the XO, get the Rhea adapter as well as a dip 14? So I hot ironed off the original 575. From your image of the Rhea, it has 4 legs. Can I assume they fit into the pads for the x575? Do I solder the Rhea onto the pads? Then what do I do next. Can you give me simple intructions? Thx whitgir.

 

[Whitigir]

Lol, is that a"dropped-in"? I think the abracon is too much work for me, I'll be favorite to the first to brick my lks 004. What I have in mind is to start by upgrading to the 575x but to do it in such a way that at a later stage, I can have a real dropped in upgrade to the Pulsar (to avoid having to upgrade vr and caps?). From what I read here, I will have to other than the XO, get the Rhea adapter as well as a dip 14? So I hot ironed off the original 575. From your image of the Rhea, it has 4 legs. Can I assume they fit into the pads for the x575? Do I solder the Rhea onto the pads? Then what do I do next. Can you give me simple intructions? Thx whitgir.

Every lks004 can be fixed for the : Pulsar OCXO, without any changing of capacitors

What you need in parts ? DIL-14 Dip sockets, Rhea adapter, clocks of choices (575x, 950x, Pulsar)

1/ prepare dil-14 dip by pushing out the unnecessary pins by a hot iron.

2/ take out the PCB board

3/ desolder stock 575 clock, no need to clean up, you can if you are neatly person

4/ turn around and fix in to fit the DIL-14, solder it up

5/ prepare the Rhea and the clock. Make sure you are watching and following the pins out arrangement

*both 575x and 950x has the same direction to put in and replace the stock one*

6/ solder paste onto the pads of the clocks (enough to cover the pads neatly and not over the pads).

7/ use tweezers to drop the clock onto the Rhea board and hot air solder it on

***if you care, the back of the Rhea board has a place for noises decoupling capacitors as @b0bb suggested, I used 0.1uf and 50V MLCC for it***. It brought out better upper treble details in my experiences

8/ make sure you are watching the pins out arrangement on #5 and the picture I provided

9/ Dip-in, and done!



The Abracon is only possible if your version has 1A version of regulators. The smaller one does not need capacitor upgrades. The biggest Abracon will need capacitors upgrade

 

[b0bb]

@b0bb Do you thing how big the difference would be between the AOCJY sc-cut OCXO and pulsar clock? Do you think the crystal they use has similar spec, Thanks in advance!

Pulsar has 25x less jitter.
SC is the most common type of crystals used in an OCXO.

OCXOs that perform at the Pulsar's level are considered dual use devices governed by the Wassenaar Arrangement (aka weapons grade) so commercial outlets like Mouser will not stock it.
I had to fill out the Wassenaar paperwork before I could buy the Pulsar.

Abracon 503fs r.m.s jitter

Pulsar 20fs r.m.s jitter

 

[b0bb]

Yes, that is what I am asking. I wonder if the ES9038Pro can be latched onto as high precision as those Modules by MSB themselves as their galaxy clock is 6k clock Q_Q

Yes it can, the phase lock range is programmable.

The DPLL is an error correction device, the wider the range the lower the quality of the signal and this is why LKS recommends the lowest BW setting be used where possible.

 

[b0bb]

Thanks @b0bb
Though, I notice when at cold turn up, the lks004 displayed unlock into lock BW01 and 44khz, then unlock into lock at 48khz, then unlock into lock at 176khz. I quickly played music upsample to 176khz, and there was no Unlock or stuttering at all. Just beautiful music. Is this normal ? ***must be normal as I am connecting via I2S and it only becomes unlock when my PC is off, or powering on and goes through booting steps

That is normal I see this on my I2S connection

 

[Whitigir]

Wow, and here I thought AT cut is common where SC is less common. Thanks for the graphs above @b0bb. Was that option standard or E, F that you calculated above for Abracon 2 ?

How is real application number calculated ? I saw this post here

The Abracon 2 is 5x better than Abracon 1 on specs, But the rms is only 3x or so ?

 

[b0bb]

Wow, and here I thought AT cut is common where SC is less common. Thanks for the graphs above @b0bb. Was that option standard or E, F that you calculated above for Abracon 2 ?

How is real application number calculated ? I saw this post here

The Abracon 2 is 5x better than Abracon 1 on specs, But the rms is only 3x or so ?

The graphs I posted above is the phase noise (PN) profile if temperature is kept constant.
The posts you linked is why OCXOs exist, it addressed the temperature dependence issues of frequency stability.

The PN plots are 2D slice taken at a specific temp typically 25 ºC, along the temp axis much like a CAT scan.

Below is an example on how the performance varies with temp from http://www.quartzpro.com/crystalsfacts.html

XO datasheets usually quote performance at the inflection points so it minimizes the differences between the different XO options (E,F etc)

 

[demos]

The graphs I posted above is the phase noise (PN) profile if temperature is kept constant.
The posts you linked is why OCXOs exist, it addressed the temperature dependence issues of frequency stability.

The PN plots are 2D slice taken at a specific temp typically 25 ºC, along the temp axis much like a CAT scan.

Below is an example on how the performance varies with temp.


XO datasheets usually quote performance at the inflection points so it minimizes the differences between the different XO options (E,F etc)

XO options are more like how the manufacture rate the Xtal, The SC Oven are hot BTW

 

[bibo01]

@Whitigir
For the Pulsar is it better to have its own PS?

 

[Whitigir]

@Whitigir
For the Pulsar is it better to have its own PS?

I dont see why or how so. It just complicate the stuff further

 

[littlexx26]

help! i was tring to remove the 575 using hot air gun, now the lid of 575 detached. but the board still stick on the main board. how to remove it?

 

[Whitigir]

X_X sorry to hear that

Use flux apply around the 575, turn the air gun at the right temperature, and use a tweezers to grab it out while applying hot air around it. Then toss the 575 away