• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Amanero Isolator/Reclocker GB

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
made the comparison with what? connected to what? waste money? the thing costs less than a decent meal. if you are building one, the cost to get the rest of the parts is only a few dollars plus the $10 for the potato chips.

Based on the context of what RollE2k was asking about -- whether S01 or S02 was better. Or a comparison of the different implementations from Acko's App Notes.

nobody will EVER be able to truthfully tell you what is the best option, there is no best option, both are very good options. if you have a very well laid out dac, use the potato chips, if you are unsure, build both, or just build the Ti version.
This is the best answer thus far. And I was eluding to this as well, but in a different way.

the potato chips put pressure on dac layout and design, as well as your ability to route wiring well and solder neatly, so if you have to ask and you can only build one, just build the ti version; because the chances of you building the potato chip version into a dac of a standard where you can really hear a difference is very slim.
So in the end S01 is probably fine, but having S02 as an option is even better.

i'm going to do something that I hate other people doing, i'm going to use a car analogy. the potato chip version is like a formula 1 car, very high performance, very high revving and with lightning acceleration (slew rate) but its twitchy/unforgiving and if you cant drive it properly, you'll spin off the start line and likely off the road.

the ti version is like an Audi A8 quattro, its got high performance that is more than enough for any lead-foot daily driver on the autobahn, but it wont send you skidding off the road, unless you do something really stupid.
This is perfect for those who are newbs like myself and appropriate in this context. Clear and to the point. I find I ask questions twice just to double check I'm doing the right thing. When I teach someone something new at work I always give analogies. Video engineering is not easy for most to grasp.

i'm not against people asking questions that have meaningful answers, nobody can give you an honest answer to this question that actually means something to you in your build, especially since we dont even know what that build is...
I feel like I might have mentioned my build ideas a few posts up. I didn't want to repeat myself. I will for expediency and so others can see what is being done with Acko's board.

the whole point of the thing was to make it possible for people to try both, so that Acko could get feedback.
I will be glad to give feedback when I'm done with both projects that I'll mention again so I'm more clear.

1) Psync -- Amanero -> Acko Isolator S01 -> Curryman 9023 MCLK (49.152Mhz) -> (RCK) Re-clocker S01 -> i2s -> Curryman v1.08 9023 w/ JG Buffer

2) Async with re-clocking -- Amanero -> Acko Isolator S02 -> AKX302 (1x 98.304Mhz) -> Re-clocker -> i2s -> Subbu v3 (on-board 49.152Mhz XO) -> JG Buffer.

I AM against filling this thread up with meaningless questions and answers. even what i've written above has been written in here numerous times.
I'm glad it's been mentioned several times. Sometimes things that are complex need several explanations in several ways in order to be ingrained in ones head.
 
Is it worth it to go the faster potato chip route for re-clocking if going with an on-board 49.152Mhz XO in PSync mode? If I'm reading correctly it seems like it fits better when you're running a faster than 98.304Mhz XO.

The faster Potatochip FF can also be used for lower speed applications, after all it is rated from DC to 600MHz. The claim from Potsemi is that at higher speeds there is less noise (overshoot/undershoot) during switching but others like NXP, TI are also saying the same. So the reason for rolling out two different types is to see for ourselves the sonic difference, if any. Personally I would prefer the flow through design of the SOT23-5 packages. It makes the layout more compact and optimised I/O arrangement. The Potatochip uses the traditional CD4013 pinouts and for this re-clocker design, I must admit was bit of a struggle to get things aligned. You can see the results on the board layout. Anyway, as more of us get these built we will be able to get a clearer picture of the preferred board.


what's the difference between using a 50Mhz XO vs. 49.152Mhz XO? I know the latter is a direct multiple of 48Khz, 96Khz, etc. Not sure about the former. Does it only matter when dividing/multiplying clocking?

It only matters if you using synchronous clocking and yes, the divided clock. But there is no harm if you use 49.152MHz or 98.304Mhz for the appropriate DACs in async mode (single XO 50Mhz/100Mhz). Besides, with this dual clock board it would better to start with the synchronous frequency XOs so that there is no need to change XOs in case you wish to do synchronous clocking later on:

50MHz system:
X1 = 45.1584 MHz XO
X2 = 49.152 MHz XO (not needed for single XO setup)

100MHz system:
X1 = 90.3168 MHz XO
X2 = 98.304 MHz XO (not needed for single XO setup)

*check typo please!
 
On the BoM for the clock there is a typo - two U5s -- for the potato chips.

Thanks! One of them should be U3. There is an new update for download- registered users of AKX302 only. Please use the link given before or here.
Changes:
BOM update

Also, what is RG1? and what is K1-K3? Where do you "link" one of them? Does that mean you either use RG1, the LDO, or the LTC6655?

Apart from different types of XOs this board also allows you to experiment with different types of power supplies and see which is the best: 3-Term Shunts, ADP150/151 LDO (9uV) or LTC6655 (1uV) regulator. You can populate all of them but only one should be connected to the XO cct using the link pads K1, K2 or K3 on the board.

The XOs themselves present fairly constant load so I am not sure if there is any advantage in using shunt regs?
 
Acko,
Sorry for the late response, Here are a few pics of the finished boards with temp interconnects, I tried using a shorter cable to connect mclk to the dac, a few dropouts were still evident with 44K,

20130602_190443.jpg 20130602_190229.jpg
 
Ian, sorry but looking at that you shouldnt be surprised you are getting dropouts. I cant see where you have connected everything. MCLK should be the same length as all the others, you seem to have all the others connected by a huge ribbon cable, but where does that connect?
 
Ian, sorry but looking at that you shouldnt be surprised you are getting dropouts. I cant see where you have connected everything. MCLK should be the same length as all the others, you seem to have all the others connected by a huge ribbon cable, but where does that connect?

The dropouts are only when dpl is on the low setting in full sync mode, I am using the diyinhk $99 dac, it connects via headers, thats where the ribbon cable is connected. The MCLK is the only ufl connector on the dac, its 50mm in length. Since all cables should be the same length I will shorten the ribbon cable to match the ufl and shorten the /4 ufl from the AKX302.
 
OK yes, the ribbon cable is far too long and honestly the $99 dac board is an example of something I wouldnt be using the potato logic parts on, especially with a ribbon cable and connector like that. so you havent got the clock fitted on the dac board, the only clocks are that on the AKX302 correct?

also what are you using for the controller for the dac board register settings?
 
I understand it may not be optimal...but its what I have. There is a 100MHz clock on the dac, but it can be disabled for sync operation, Im using an Arduino uno with Hifiduino code for the dac board register settings. Is there a ufl to header adapter board that can be purchased?
 
Last edited:
its cool, not being critical, I think i've already made myself clear on that thread. just being practical ie. trying to save you possible trouble and several of us on this thread have already mentioned the potato option may not be best suited to that board, as their advantage can become a disadvantage.

can the onboard clock actually be disabled, or just turned off? ie are you actually using its output enable pin to disable it, or just not applying power? because even if its off its output stage is still in parallel with the input, so will negatively effect the clock wave shape due to the diode effect. that alone could cause dropouts.

as for a u.fl adapter, unless its made specifically for your dac board, wiring it in with flying wires or bending pins will I think do more harm than good. I would just try and get some good quality ribbon cable and make it as short and neat as possible (50mm is good) then only split the ribbon into individual wires for as short a length as possible and hard-wire it after testing, lose the connectors.
 
..... This is my first project, learning a lot, hopefully I can transition al this knowledge to a highend dac setup later.

You have made great efforts already! The system is capable of higher performance even with the DAC module that you have chosen.

I have mentioned this before: once the signals leave the re-clocker there is nothing left to prop them up and therefore quality of interconnects come into play. So connections should be short and direct to the DAC inputs and the re-clocker, particularly the MCLK and RCK signals as these are ~100MHz for the full sync setup of yours.

Also, I am a little curious as to what would happen if the div/4 link from Clock Board is disconnected. If the Amanero is truly using external clock then the sound will cut off.
 
Thanks! I will double check later but when I initially made connection to the dac, if the div/4 link is disconnected no sound was heard. think i will use short ufl cables with the ends removed soldered to the header location directly as was suggested.

Also I have a ESS9018 dac chip, anyone know of any pcb with provisions for ufl conectors and excellent layout etc. that I could purchase?
 
Last edited:
iancortez, saw you put your name in for the u.fl adapter stuff from Ian, honestly its a waste of money for your purposes, it isnt made to match other dacs, so it will not improve anything, it will probably be worse than using a short ribbon cable.

dont get me wrong, I love u.fl connectors, in fact I pretty much championed them into popularity here, but they are all about having well terminated, controlled impedance connections and a u.fl adapter hacked into another PCB with bent wires or pins to make it fit, without proper termination is a cosmetic mod more than anything; it will not bring any real benefits IMO
 
I'm not 100% sure on this, but looking at the E/D pin function on the CCHD957 datasheet, it looks like it is enabled when there is an open circuit on the E/D pin. So I think you can just leave that stuff all alone, leave it open circuit and the clock will just power up and run.

You can of course, build it without those parts, then see if it works or not, certainly won't damage anything. Which is only a useful suggestion if you have all the parts already. If you're working out what to order then maybe simplest to just order those parts, they aren't expensive ;).
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.