My Behringer DCX retrofit idea would need 3x ES9023 ...could just leave them on one card {or see if 3 of these boards can fit or stack} [I digress] I am also trying to force myself to learn Eagle; I get ideas...
I am VERY happy with the board, and WILL be able to do what I like, easily. Being able to take Micrel or LT is very nice! Once I hear it I'll not want to chop it up HAH. I'm sure my whole family will have them. I have them hooked on USB isolation now, I DO NOT SEE A USB isolator for max speed USB2 or USB3, that a working man can afford. Maybe a DIY project will surface, maybe I missed it.
I am VERY happy with the board, and WILL be able to do what I like, easily. Being able to take Micrel or LT is very nice! Once I hear it I'll not want to chop it up HAH. I'm sure my whole family will have them. I have them hooked on USB isolation now, I DO NOT SEE A USB isolator for max speed USB2 or USB3, that a working man can afford. Maybe a DIY project will surface, maybe I missed it.
There has been a project in elektor last year: USB-Isolator - ELEKTOR.de | Elektronik: Analog Digital Embedded Mikrocontroller Audio Messtechnik
you might need to google translate it though.
A readily build PCB is available from eurocircuits
best regards, Stefan
Last edited:
USB Isolation
note the "• Data transmission rate of 1.5 Mbit / s or 12 Mbit / s "
in the ad. USB2.0 can get up to 480Mbit/s. I have numerous Opticis fiber optic USB isolators and have given them to family, those? Same issue: great for say a TE7022 at 24/96 not something like a CM6631A at 24/192, however.
I think it is possible to construct a USB isolator for 480 and beyond, I have seen people mention it in forums, but have not seen a DIY.
note the "• Data transmission rate of 1.5 Mbit / s or 12 Mbit / s "
in the ad. USB2.0 can get up to 480Mbit/s. I have numerous Opticis fiber optic USB isolators and have given them to family, those? Same issue: great for say a TE7022 at 24/96 not something like a CM6631A at 24/192, however.
I think it is possible to construct a USB isolator for 480 and beyond, I have seen people mention it in forums, but have not seen a DIY.
You might have seen from my test setup picture that I was using acko's Amanero USB Isolator Re-Clocker between the Amanero USB and Daniel's DAC on the I2S signals. Domenico sort of recommends the use of a specific TI isolator whereas I'm using IL715E by NVE Corp with acko's little pcb.
I won't measure anything in this round, but I promise to build and measure at least two different setups in the next round. It won't be easy to make meaningful measurements, but I'll try my very best. Worst case I'll waste some time, with luck I might learn something.
I won't measure anything in this round, but I promise to build and measure at least two different setups in the next round. It won't be easy to make meaningful measurements, but I'll try my very best. Worst case I'll waste some time, with luck I might learn something.
Last edited:
I2S isolation then...
THAT sounds like I2S isolation. Or did you have access to something that somehow galvanically isolates 480Mbps USB2.0? BEFORE the Amanero?
And THEN my mind wishes we could use GB Ethernet-to-I2S "somehow".
I know there is USB tunneling going on via ethernet...not sure if/where any of THAt fits in terms of timing.
I have some of CHristian's Bragi boards and IL715 chips around but not much to measure with, well, 35Mhz dual-trace scope calibrated 25yr ago. I have TE8802 modules, CM6631... soon a CM6631A//CS4398 to hack. Now KlipschKid did get 24/96 mode working via 12Mbps isolation on a CM6631 I believe but that is not much help with today's faster and faster streams.
Have you seen anyone make a USB2.0 480 isolator for under $100 that just works? Devled into USB over ethernet?
THAT sounds like I2S isolation. Or did you have access to something that somehow galvanically isolates 480Mbps USB2.0? BEFORE the Amanero?
And THEN my mind wishes we could use GB Ethernet-to-I2S "somehow".
I know there is USB tunneling going on via ethernet...not sure if/where any of THAt fits in terms of timing.
I have some of CHristian's Bragi boards and IL715 chips around but not much to measure with, well, 35Mhz dual-trace scope calibrated 25yr ago. I have TE8802 modules, CM6631... soon a CM6631A//CS4398 to hack. Now KlipschKid did get 24/96 mode working via 12Mbps isolation on a CM6631 I believe but that is not much help with today's faster and faster streams.
Have you seen anyone make a USB2.0 480 isolator for under $100 that just works? Devled into USB over ethernet?
Leave it to HiFiDuino:
Which Digital Isolators for I2S? …or NOT | H i F i D U I N O
already dated but interesting
Which Digital Isolators for I2S? …or NOT | H i F i D U I N O
already dated but interesting
Awfully quiet around here 
Finished up my taxes mid last Sunday and spent a couple hours stuffing a prototype board. I should be able to put power to it this Sat. I've been listening to an Acko ES9023 I assembled that has the JG Filter installed. It really does elevate this DAC chip.
Daniel - keep pushing on V2, you really have something here. I hope to build a handful more for friends/family. [Very stupid question about casing options has been deleted. Thanks for the reality check Pat!]
Finished up my taxes mid last Sunday and spent a couple hours stuffing a prototype board. I should be able to put power to it this Sat. I've been listening to an Acko ES9023 I assembled that has the JG Filter installed. It really does elevate this DAC chip. Daniel - keep pushing on V2, you really have something here. I hope to build a handful more for friends/family. [Very stupid question about casing options has been deleted. Thanks for the reality check Pat!]
Last edited:
It's so quiet since we are all enjoying good music with our DACs, aren't we
To be honest I think that the Acko implementation will be at least on par with V1, probably be better. The upcoming Acko version doesn't have regulators and clock on board so it fits somehow different applications though...
Actually I got feedback from some of the testers that they will need some time since they have other thinks to do... I myself have three boards up and running and I am about to put them into my miniSHARC based DSP system.
V2 is almost finished and a first batch will probably be ordered this weekend.
Stay tuned...
To be honest I think that the Acko implementation will be at least on par with V1, probably be better. The upcoming Acko version doesn't have regulators and clock on board so it fits somehow different applications though...
Actually I got feedback from some of the testers that they will need some time since they have other thinks to do... I myself have three boards up and running and I am about to put them into my miniSHARC based DSP system.
V2 is almost finished and a first batch will probably be ordered this weekend.
Stay tuned...
Been following the thread for a while and I've got to say, I'll be interested to see the minisharc setup curryman. How're you managing the mclk in that setup? I've toyed with the idea of a ES9016 for multichannel output from minisharc, I know there are others thinking about that too. I'm waiting for ADC options before I start on the minisharc myself.
I've got one of Acko's 9023 boards in my headphone amp setup and need to find some time to rewire the DAC to put the JG filter in it, looking forward to hearing it soon though! [I put that off because I wanted to not make too many changes all at once and only just put Ian's FIFO + Si570 + isolator into the system]. Acko's PSU section on the current 9023 board is pretty damn nice I have to say. My JG filter is just going between DAC and stepped attenuator. Stepped attenuator goes straight to a 'The Wire' SE-SE headphone amp. Will be interesting to see how much difference that filter makes in this configuration!
I've got one of Acko's 9023 boards in my headphone amp setup and need to find some time to rewire the DAC to put the JG filter in it, looking forward to hearing it soon though! [I put that off because I wanted to not make too many changes all at once and only just put Ian's FIFO + Si570 + isolator into the system]. Acko's PSU section on the current 9023 board is pretty damn nice I have to say. My JG filter is just going between DAC and stepped attenuator. Stepped attenuator goes straight to a 'The Wire' SE-SE headphone amp. Will be interesting to see how much difference that filter makes in this configuration!
Last edited:
I was running a very similar chain Chris. For testing, I was using the Amanero, Acko reclocker, Ian's si570 for RCLK (PSYNC), Ian's Battery Management board + LifePO4, Acko 9023, JG Buffer, to Wire SE.
My main system is a pretty maxed BIIISE board (NTD1, Auricaps, TX's, etc) and I was astonished at how well this mini-chain performed against it.
When Acko's clock boards are available, I'll use that in the mini chain and case it up. One thing though; while I like what the JG buffer brings, the direct DAC output is pretty damn good and has more slam. The Acko board allows you to tap output before and after the buffer easily, so I think I'll expose both through the case
My main system is a pretty maxed BIIISE board (NTD1, Auricaps, TX's, etc) and I was astonished at how well this mini-chain performed against it.
When Acko's clock boards are available, I'll use that in the mini chain and case it up. One thing though; while I like what the JG buffer brings, the direct DAC output is pretty damn good and has more slam. The Acko board allows you to tap output before and after the buffer easily, so I think I'll expose both through the case
I have ~110 ES9023 here in the USA. I prefer not to have to fill out customs forms and have to drive to post office to ship internationally. $3Ea. I have quite a few BETA boards left too $3ea any qty. I do not have full kits made up, but 95% of the parts to do so, are here - that could be a big job {cutting strips, marking and taping to a BOM}. FET-grading 200 FETs was a chore -- the readings vary widely, assembling this buffer without grading the FETs would probably not work well. I acquired another 200 FETs but no time or sanity to sacrifice yet. If we were to do a group buy we'd need to grade more like 2000 FETs. The tester we've been using to grade the FETs seems to work fine, they run under $20 USD from China, shipped.
Hey Pat. My tester landed, so I'd be willing to share the work of grading fets. Daniel, I think your board has the potential to satisfy the pent up interest in a GB for a quality 9023 board for the masses. A quick look at the Subbu DAC waiting list bears that out I think. I am in NO WAY diminishing that DAC or making comparisons, other than to say V2 as it is designed - with the possible addition of a balanced option may fit the bill very nicely.
Just saying
Just saying
I found that I am often just slower than the cycle time of the TESTER going to sleep, as I fumble with the the tweezers and next FET. As you saw {in your mailing}, I stop and write down the readings for each on the grid I draw. SO grading FETs can be taxing. After grading comes matching: finding close quads of each type of FET used. The more devices graded the easier matching would be. Easier, but it all takes time.
how are you matching them? single point IDSS, curve trace, transconductance at operating point, multipoint IDSS, VGS, VDS etc ? single point IDSS is not terribly useful really. not to diminish your efforts, but selling properly matched fets is not a value proposition for the seller and the still rather significant effort required to test in a simple way is not really that much better than picking quads from next to each other on a strip when it comes to dynamic performance, especially if it is not even a single point that is the same as the operating point in the design they will be used in.
you are using plastic tweezers? metal tweezers will act as a heatsink and effect the reading depending on the time of day/temperature, which will also be effected by how long each one is tested for and if each is not biased up for the same amount of time, the reading will vary based on self heating. the only way to get even somewhat close results for even a basic IDSS measurement is to be very quick so that self heating is not as much a factor
you are using plastic tweezers? metal tweezers will act as a heatsink and effect the reading depending on the time of day/temperature, which will also be effected by how long each one is tested for and if each is not biased up for the same amount of time, the reading will vary based on self heating. the only way to get even somewhat close results for even a basic IDSS measurement is to be very quick so that self heating is not as much a factor
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.