Nice work pixel28.
So you can confirm that ADM7154 is excellent in that positions. What is now at AVCC position at your dac? You will use them also in AVCC lines?
I read that ADM is not so good for analog lines vs lt3045 because they are little bright and thin...
Mark you tested lt3045 in AVCC with or without boost Transistor? what is main difference That you found out in compare to opamp based AVCC regulator.
It is interesting how far was this project was gone... congratulation.
So you can confirm that ADM7154 is excellent in that positions. What is now at AVCC position at your dac? You will use them also in AVCC lines?
I read that ADM is not so good for analog lines vs lt3045 because they are little bright and thin...
Mark you tested lt3045 in AVCC with or without boost Transistor? what is main difference That you found out in compare to opamp based AVCC regulator.
It is interesting how far was this project was gone... congratulation.
Not me. I think Mikett and others have tried using the Chinese LT3042 boards that have a pass transistor. They always say opamp supply sounds way better. Iancanada and those folks also mostly have sworn off LDO regulators for dacs. Twisted pear uses ADM7150. Acko Dac uses ADM7154. Acko says they sound pretty close to ESS ES9311Q, not worth changing from what they use now.
Aslo, schematics of the new AK4499 dac chip evaluation board show they are using something that looks kind of similar to a Jung/Didden series regulator for their equivalent of what ESS calls AVCC. AKM is using one of the discrete voltage regulators for each of the left/right channels.
My interpretation of the foregoing is that one should probably proceed carefully if wanting to use LDOs for AVCC. For mobile ES9038Q2M chips it could be they don't draw enough AVCC current to work with LDOs like ADM7150 in the same way the PRO chips do. I am not aware of anyone having done a test to settle the matter one way or the other. That being the case, I can only recommend what I know works for sure in the case of this particular modding project we are doing here, and that means I recommend opamp regulators. Of course, this always has been a work in progress and if we find better ways to do things, my recommendations are subject to change.
Thank you 🙂 Still not finished, as one can see. It would help if we had other people who modded dacs up to the same point where I am at now and who could agree to divide up the work of investigating certain remaining issues (mostly on the digital end of things).
Then of course, there are alternate topologies that could be investigated. From what I have seen around, I suspect there could be some possible improvements to make with clocking of Sabre dacs, and of course with DSP.
Pretty sure there will be a detour soon though, given the impending release of AK4499. Soon most of the attention in diy circles will likely be shifted to it until we see what it can do. If its a dud, then back to ESS, and so on.
Last edited:
distortion, I don't trust but I would trust someone's word who prefers es sabre dacs
Distortion can be beautiful. I will never forget the meridian 206 cd player and its beautiful distortion, subtle. I had the bitstream conversion model. No distortion means paper specs which my ears do not prefer. Any company who releases dacs that sound like plastic has to do a serious makeover to regain my trust. Nothing compares to a garrard 301 in concrete with a rewired rega arm and an origin needle through a beard p-80 amp with 4 X 12ax7s and 4 gec kt88s into resurrounded, upgraded paper and oil cap crossover tannoy hpds 15" in birch plywood cabs. While one dac might sound nice for a while, headaches do set in for unknown reasons. The meridian 206 never gave me a headache except for when it malfunctioned and when playing cds like exile on main st that would just scratch to unusable condition from a slight drop. I bet it and that system had crappy specs on paper but man oh man, I want to go back! again and again. I'll never forget those bright valves (I shielded) as they sat there beaming radiation between charged plates.
I am using crossover free speakers these days which had whizzer cones I cut off with upgraded voice coils from the manufacturer here in Thailand. They say NASA on the back and when I bought them for 10 bucks a piece from a manufacturer who also sold very expensive speakers, immediately blew them, and bought more, until I had him put better voice coils on for an extra tenner a piece. he said "gaeng" which=smart in Thai when i preferred them over his expensive speakers. the whizzer cones had phase issues and are now about 1cm deep, after cutting them off with an exacto knife and they are bloody great. Finding decent plywood here was a challenge though. I almost got kicked out of an apartment for hand sawing the wood around midday but the landlord was thinking I was cutting up the bed or furniture I guessed. hahahahaha Anyway, reminiscing over.
Even the crappy es 9028 dac with new pins and a better regulator sounded close to my old system and no headache after 2hrs listening, but the crappy JRC chip (opamp?) has to go!. akm was way off the mark. I do not believe paper. Most of our hearing sub 1000hz comes partially or some say mostly from bone conduction as well, it takes 1/4th of a sound wave to fit in the ear canal I have read and ((300m/s)/1000)/4=7.5cm, bigger than an ear canal can fit, or just at the threshold. As I think about that theory however, each frequency should have a distinct gradient that the ear can hear sub 1000hz. One thing is for sure, I trust people with similar tastes, and my feelings/hearing over paper specs.
Have people here compared the ess9028 to the 9038 soundwise?
I can't judge akm forever from two dacs I have heard.
I do like es sabre and cirrus logic.
Distortion can be beautiful. I will never forget the meridian 206 cd player and its beautiful distortion, subtle. I had the bitstream conversion model. No distortion means paper specs which my ears do not prefer. Any company who releases dacs that sound like plastic has to do a serious makeover to regain my trust. Nothing compares to a garrard 301 in concrete with a rewired rega arm and an origin needle through a beard p-80 amp with 4 X 12ax7s and 4 gec kt88s into resurrounded, upgraded paper and oil cap crossover tannoy hpds 15" in birch plywood cabs. While one dac might sound nice for a while, headaches do set in for unknown reasons. The meridian 206 never gave me a headache except for when it malfunctioned and when playing cds like exile on main st that would just scratch to unusable condition from a slight drop. I bet it and that system had crappy specs on paper but man oh man, I want to go back! again and again. I'll never forget those bright valves (I shielded) as they sat there beaming radiation between charged plates.
I am using crossover free speakers these days which had whizzer cones I cut off with upgraded voice coils from the manufacturer here in Thailand. They say NASA on the back and when I bought them for 10 bucks a piece from a manufacturer who also sold very expensive speakers, immediately blew them, and bought more, until I had him put better voice coils on for an extra tenner a piece. he said "gaeng" which=smart in Thai when i preferred them over his expensive speakers. the whizzer cones had phase issues and are now about 1cm deep, after cutting them off with an exacto knife and they are bloody great. Finding decent plywood here was a challenge though. I almost got kicked out of an apartment for hand sawing the wood around midday but the landlord was thinking I was cutting up the bed or furniture I guessed. hahahahaha Anyway, reminiscing over.
Even the crappy es 9028 dac with new pins and a better regulator sounded close to my old system and no headache after 2hrs listening, but the crappy JRC chip (opamp?) has to go!. akm was way off the mark. I do not believe paper. Most of our hearing sub 1000hz comes partially or some say mostly from bone conduction as well, it takes 1/4th of a sound wave to fit in the ear canal I have read and ((300m/s)/1000)/4=7.5cm, bigger than an ear canal can fit, or just at the threshold. As I think about that theory however, each frequency should have a distinct gradient that the ear can hear sub 1000hz. One thing is for sure, I trust people with similar tastes, and my feelings/hearing over paper specs.
Have people here compared the ess9028 to the 9038 soundwise?
I can't judge akm forever from two dacs I have heard.
I do like es sabre and cirrus logic.
amanero outs
Did you have to drill holes for the extra i2s pins? are those necessary if I set dsd and sample rates manually? Would I do that through the four buttons and the LCD screen, setting them manually? On yout pic it looks like you are not using the mclk from the amanero, just fsclk bclk, and data, with three grounds. My amanero is a leaf audio module and it only has 2 ground pins at the i2s out place, although there appears to be another one near the usb input. It has the same chips, I just liked the look of the caps on the unit so I got it. It has 3 clocks. I wolud rather set samppling rates manually and dsd too. DO I still need to drill holes and add pins if I do that? can I do that via the 4 buttons and avoid some drilling?
Should I use the Mclk from the amanero or the DAC?
my amanero has sctf 12.000, 24.576. and 22.5792mhz clocks
I would rather not replace clocks on the DAC or the ak4137 unit. mine is pcb1.3. i got 22.5792 and 49.172 clocks on it
Did you have to drill holes for the extra i2s pins? are those necessary if I set dsd and sample rates manually? Would I do that through the four buttons and the LCD screen, setting them manually? On yout pic it looks like you are not using the mclk from the amanero, just fsclk bclk, and data, with three grounds. My amanero is a leaf audio module and it only has 2 ground pins at the i2s out place, although there appears to be another one near the usb input. It has the same chips, I just liked the look of the caps on the unit so I got it. It has 3 clocks. I wolud rather set samppling rates manually and dsd too. DO I still need to drill holes and add pins if I do that? can I do that via the 4 buttons and avoid some drilling?
Should I use the Mclk from the amanero or the DAC?
my amanero has sctf 12.000, 24.576. and 22.5792mhz clocks
I would rather not replace clocks on the DAC or the ak4137 unit. mine is pcb1.3. i got 22.5792 and 49.172 clocks on it
Last edited:
They all tend to sound the same and very good if the implementation is good enough. A lot of the sound of them is often the sound of the circuit they are in, as much as anything else. Hopefully, when you get done modding you will have a very good implementation.
Yes. I used a dremel tool.
It is not related to manual settings. The added ground pins and ground wires are to minimize jitter and preserve the quality of the I2S signals. Ribbon cable with every other wire connected as a ground wire acts a lot like a set of parallel coax cables. It is a very good way to connect RF between boards while maintaining very good signal integrity.
The buttons are used for a manual settings. Again, it is unrelated to the number of ground wires.
Correct. MCLK would typically be for use with AKM dacs.
Your choice if your connect all the signal wires or not. You might have to check if you have a real Amanero or a Chinese clone and let us know, as that is a big difference. Anyway, if only the three I2S signals are used, then one would only need four ground wires, one ground wire between each of the signal wires, and one ground wire at each outer edge of the ribbon cable. That is all shown in the pics if you look carefully.
You don't have to. The dac will sound significantly better if you change the clock on it, however. We have a number of posts I can link to that explain exactly how to do it. If cost of a clock is the issue, you can get a very good NDK 80MHz clock for about $8 plus postage:NDK NZ2520SD 3.3V 80Mhz 49.152Mhz 45.1584Mhz 24.576Mhz 22.5792Mhz 12Mhz Ultra low phase noise oscillator - DIYINHK
So far, nobody, not even beginners, have had a problem with clock replacement if they read up on all the info we have on how to do it. (I use and recommend a Crystek 525 clock, but other clocks work too. And any recommend clock will be an improvement.) The main trick for removing clocks almost effortlessly is to use Chip Quik de-soldering alloy. That stuff is amazing. Just don't scratch up the solder mask and soldering in a new clock is very easy too.
The AK4137 does not need clock replacement. However, I removed one of the clocks on mine and replaced it with a u.fl connector. It was easy soldering work. That was so I could run the AK4237 DSD mode synchronously with the dac clock, for better sound quality. However, everything is optional with this project. Nobody has done everything yet except me. Most people seem to have had enough of dac modding long before that 🙂
Last edited:
Gosh I have a dremel in USA. I hate to buy a new one. About the AKM DACs I have heard them in crap circuits as well as sabres and sabre did better in a more crappy circuit. There are elements of sound I can hear through and other fundamentals I need. I can hear through most circuit problems but if the fundamental noise generator has a problem, I cannot hear through that, The way I am able to hear through crappy circuits is adding memory to what I hear. If the fundamental source is not to my liking, No Way Jose! I have spent enough time with top studio equipment to have a good memory to spice up what I hear with. Also I spend half of my days, most of my days, ~6-8hrs playing unamplified (priceless) acoustic instruments I worked my *** off for, That may affect my perception as well.
also grew up on a piano my grandma did not pay much for that recently sold for ~15 grand, no auction, a dealer paid that, she took the first offer so i assume it was priceless as well. She had good ears too I guess.
Honestly sometimes I prefer the sound of a cheap 60s suitcase record player or a lenco goldring over most digital stuff.
But one thing is for sure, power sources make all the difference in the world. I use >10kg lab grade ones I pick up from scrap markets in thailand that are replaced with switching modules. And flourescent lights, acs, everything on my personal supply need to be OFF for a listening experience. Not sure if the light drowns out the experience as much as that tiny buzz some appliances add to the mains. Shielding is also important.
also grew up on a piano my grandma did not pay much for that recently sold for ~15 grand, no auction, a dealer paid that, she took the first offer so i assume it was priceless as well. She had good ears too I guess.
Honestly sometimes I prefer the sound of a cheap 60s suitcase record player or a lenco goldring over most digital stuff.
But one thing is for sure, power sources make all the difference in the world. I use >10kg lab grade ones I pick up from scrap markets in thailand that are replaced with switching modules. And flourescent lights, acs, everything on my personal supply need to be OFF for a listening experience. Not sure if the light drowns out the experience as much as that tiny buzz some appliances add to the mains. Shielding is also important.
Last edited:
I have many clocks from studio equipment I scrapped. I will dig them all up and ask you about them, Mark. I am not sure which ones are good but they look nice. You will probably know what I should hang on to and what is trash. I hung onto them because I know there are little crystals inside of them.Some are crystek at the same frequency as the ones in the DAC. Big ones.... Old clocks, I have seen in pics, are fascinating instruments to me.
Last edited:
Some news for those considering ADM7150 for ES9038Q2M AVCC use. Just tried it myself with one of Iancanada's FIFO_Pi dacs, and something sounded pretty distorted with the sound quality. I also had a 'New Class D' 3.3v, 3-terminal regulator to try, but it needed more than 5v input since it is not an LDO. I found it would run okay with 5.7v input power. Using the New Class D voltage regulator for AVCC improved the sound quality a lot, so I definitely would recommend it over ADM7150 for use with Q2M dacs. It could be with higher current levels of a PRO Sabre dac the ADM regulator would sound fine, but it didn't sound good for the case in which I tried it with a Q2M dac.
Last edited:
You can disconnect the ground from the I2S between the two boards, digital only needs a rough ground. Both boards already share a grounding point through the supply.
More in general, faulty grounding can wreck heavoc, orders of magnitude larger than any clock improvement can deliver. I find grounding, let's say lay out including managing return currents etc., much more complicated than the electronic design itself. The latter can in large part be found in text books, manufacturers info (NS blue book) and sometimes even on this site. Plus that math works. When it comes to the former, it is each to their own. It requires much more head simming capacity, at least for me. As always Bruno Putzeys has written some insightful stuff on this in his LA article on a flawless opamp based pre-amp: follow the signal.
By having two grounding points as per your example, you are asking for trouble. Lifting one is a good start, but still further study of board lay out etc. is required to determine the best landing spot. On most DAC boards I have seen, it is best to land your ground as close as possible to the first opamp from the DAC-chip. Imo.
No, I don't think so. It may be recalled that clocks and RF signal propagation are analog, not digital. RF needs a solid ground to avoid reflections and standing waves due to abrupt impedance mismatches. Try removing the coax ground from your satellite dish to its receiver and ground them though the power line ground instead. What a mess.
I already determined that particular I2S connection required gold pins and alternating ribbon cable ground wires for lowest jitter. Seems silly to mess it up by making the RF ground return go through the power supply ground returns. For one thing that path is pretty inductive at RF frequencies. The better course of action under the circumstances was to effectively deter RF from using the power supply ground at all.
EDIT: However, I would agree that grounding problems are very common in most diy audio systems, and it causes audible degradation of the sound. Maybe that explains part of why some folks are attracted to battery power supplies and audio transformers. People don't seem to understand the underlying sources of SQ problems and try to apply bandaid fixes, IMHO. Of course, there is also some mystique around tubes and transformers as though they make sound quality better, which they don't. They generally add more distortion, and even adding nice sounding distortion to a reproduction system is too much like adding the same spice to every meal. It's not better in reality. All the foregoing is IMHO only, of course 🙂
Last edited:
2 of the clocks only have 2 connections, so I am guessing the crystal is vibrating at a lower order and they are not suitable for the 4 point solder connection clock replacements, even though they still sell for 20 bucks a pop on ebay. The other, I will need to write black lion mod to ID unless anyone here knows what kind of 4 point connection clock the digi 002 uses, the print has seemingly rubbed off and digidesign schematics are proprietary so probably only BLM knows🙁
Last edited:
phi112358,
For ES9038Q2M we would normally be looking for an ultra-low jitter 100MHz clock. The NDK 80MHz is also very low jitter. These are different than most audio clocks that are below 50MHz frequency. Also, if two pins then that would be crystal, and if four pins then it would be a clock module. It is by far the best for ES9038Q2M to use a module. If it would be too much trouble to get the kind we normally use (and that are unlikely to be in most audio gear) then it should be okay to keep using the original clock for a long as necessary. If at some point buying a new clock becomes of interest, we could recommend one then. The thing is, I don't want to get hung up on just one part. It would be great if you could find parts to make an output stage, etc. Also, I don't remember if we talked about an AVCC supply. Normally we use a dual audio opamp for that. For the output stage and AVCC we prefer to use OPA1612 (or OPA1611) but maybe hard to find. Next best is probably LME49720, LME49860, or LM4562. A pair of AD797 can also be used for AVCC, but they may be hard to find. Less good, but still usable are common 5532 audio opamps. Any of them can work, although the first ones in the list will probably sound better.
Then other parts for output and AVCC would be good to look around for. There is a schematic and BOM attached to post #3003 in this thread: https://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-301.html#post5577605
For +-15v power supplies for the opamps, it might be good to consider eziitis' Nazar regulator design. It should be used without any more electrolytic caps than shown on the schematic or it might oscillate, but its a good sounding audio regulator design: https://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-386.html#post5698160
For ES9038Q2M we would normally be looking for an ultra-low jitter 100MHz clock. The NDK 80MHz is also very low jitter. These are different than most audio clocks that are below 50MHz frequency. Also, if two pins then that would be crystal, and if four pins then it would be a clock module. It is by far the best for ES9038Q2M to use a module. If it would be too much trouble to get the kind we normally use (and that are unlikely to be in most audio gear) then it should be okay to keep using the original clock for a long as necessary. If at some point buying a new clock becomes of interest, we could recommend one then. The thing is, I don't want to get hung up on just one part. It would be great if you could find parts to make an output stage, etc. Also, I don't remember if we talked about an AVCC supply. Normally we use a dual audio opamp for that. For the output stage and AVCC we prefer to use OPA1612 (or OPA1611) but maybe hard to find. Next best is probably LME49720, LME49860, or LM4562. A pair of AD797 can also be used for AVCC, but they may be hard to find. Less good, but still usable are common 5532 audio opamps. Any of them can work, although the first ones in the list will probably sound better.
Then other parts for output and AVCC would be good to look around for. There is a schematic and BOM attached to post #3003 in this thread: https://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-301.html#post5577605
For +-15v power supplies for the opamps, it might be good to consider eziitis' Nazar regulator design. It should be used without any more electrolytic caps than shown on the schematic or it might oscillate, but its a good sounding audio regulator design: https://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-386.html#post5698160
That is true. Especially for those 1960's suitcase record players but Bob Dylan 45 "like a rolling stone" would sure sound great on those. Unfortunately for vinyl, it is heavy, constantly degrading as it is plastic, particularly not suited for hot climates, and the little diamond needle shaves a layer off every time and persistently, periodically, breaks.... I also went digital so I could travel without a cargo container just for the music collection, which would have been the case if I had carried on buying vinyl. I gave all my vinyl away to a drummer who plays with the Buzzcocks member Steve. He is the best drummer I ever played with and the most hospitable friend, or one of the most hospitaable friends I have ever had. One of those guys who could match the notes coming off the drum at microsecond precision perfectly to the music. He does alot of studio work off of Denmark st. these days.
And I have gotten into and many more music types than I was into in the past. Valves can have low distortion but the amp needs to weigh about 80kg to achieve transparency. 4 kt88s stay within class A at very high volume. No valve sound at all, just 100% analog amplification through those charged plates. But they have to be replaced over and over. That is a pain like record needles and vinyl. So glad to be free of the hassle. My father warned me but I thought I was smart and at least knew I could make a profit on everything but the records I bought and did. I sold my record player, the tannoys, and the KT88 valve amp and made a small profit in the end. Nice to have a smaller valve amp for certain music types and some occasions though. Gramophones do a great sounding piano reproduction, interesting things to have heard, and beautiful antiques for a collector.
MOSFET amps with single high quality transistor pairs can sound perfect to me, they can be as "warm" as valves and they are more reliable, portable, probably could run for a century straight in a dry environment without breaking down.
.
And I have gotten into and many more music types than I was into in the past. Valves can have low distortion but the amp needs to weigh about 80kg to achieve transparency. 4 kt88s stay within class A at very high volume. No valve sound at all, just 100% analog amplification through those charged plates. But they have to be replaced over and over. That is a pain like record needles and vinyl. So glad to be free of the hassle. My father warned me but I thought I was smart and at least knew I could make a profit on everything but the records I bought and did. I sold my record player, the tannoys, and the KT88 valve amp and made a small profit in the end. Nice to have a smaller valve amp for certain music types and some occasions though. Gramophones do a great sounding piano reproduction, interesting things to have heard, and beautiful antiques for a collector.
MOSFET amps with single high quality transistor pairs can sound perfect to me, they can be as "warm" as valves and they are more reliable, portable, probably could run for a century straight in a dry environment without breaking down.
.
Last edited:
Mark, I trust the 100 MHz clock to solve any jitter related issues. In that situation, I don't see how a dirty digital ground can do any harm.
for my friends
Here is something I am sharing with my friends I just tried, learned and believe works. I read that scientists in orbit for longer than two weeks and originally those in underground faraday cages removed from earths natural frequencies for longer than two weeks often experienced insomnia and sickness. And a scientist Schumann measured earth's natural frequencies to be 7.83hz. Allegedly astronauts in outer space recover or do not experience sickness when the frequency is played. Also in cities, there are many foreign RF signals from AM/FM Mobile phones and even the grid. I tried playing a 7.83hz wave through my speakers and it cured insomnia that nothing could cure for me, instantly. I live in a dense city but immediately felt like I was at the ocean and got to sleep in 30 min. I have tried seemingly every sleeping aid and all their cousins and nothing has worked better than a 7.83hz wave played through decent woofers. I immediately felt like I was at the beach or in a forest and was asleep within 30 min. People also sell solfeggio generators. They apparently play a 7.83hz wave and work but I have not tried those yet.
Here is something I am sharing with my friends I just tried, learned and believe works. I read that scientists in orbit for longer than two weeks and originally those in underground faraday cages removed from earths natural frequencies for longer than two weeks often experienced insomnia and sickness. And a scientist Schumann measured earth's natural frequencies to be 7.83hz. Allegedly astronauts in outer space recover or do not experience sickness when the frequency is played. Also in cities, there are many foreign RF signals from AM/FM Mobile phones and even the grid. I tried playing a 7.83hz wave through my speakers and it cured insomnia that nothing could cure for me, instantly. I live in a dense city but immediately felt like I was at the ocean and got to sleep in 30 min. I have tried seemingly every sleeping aid and all their cousins and nothing has worked better than a 7.83hz wave played through decent woofers. I immediately felt like I was at the beach or in a forest and was asleep within 30 min. People also sell solfeggio generators. They apparently play a 7.83hz wave and work but I have not tried those yet.
Ah, I see what you are saying now. We all bought into the Sabre story of flawless ASRC, etc. What else could we do, the story was a good one, and nothing in the data sheets or other information gave reason to question it. Only over time and with quite of few people working on it some people have come to understand something closer to the reality. And ESS has most of them tied up with NDAs, that or potential copyright violation claims.
It is now pretty much public knowledge that reducing DPLL bandwidth by as much as possible improves SQ. Also public, synchronous mode sounds better than ASYNC. Still kind of emerging, DSD sounds better than PCM. (all in reference to particular devices, not speaking generally).
Then there is the issue that many sub-types of S-D dacs are *very* sensitive to slight amounts of jitter, far more so than old R-2R dacs.
Thus, my DPLL=1, I am feeding it synchronously clocked DSD256, and it sounds better to me. Still needs some more work too.
Anyway, to respond to your question in this revised context, I think grounds, I2S signal quality, etc., are all big issues for solving the last few little SQ imperfections. That being the case, I have to stick with the grounding scheme I have arrived at.
For other dac applications, some people may not notice any effects from how the grounds are connected, or even if there is an unresolved ground loop. But, those things absolutely do make a difference when one is trying to get the thing to sound its very best.
Last edited:
Sorry for not responding earlier. I have a lot of work now.
I have not yet finalized the AVCC line, but I plan to do it soon. I do not want to use ADM in this line, like LT3045, TPS, and so on. I plan to make the circuit on the recommendation: Accurate and low-noise reference voltage source and an operational amplifier at the output.
This is a very good experience, Mark. But so sorry that he is not so successful. I still can not understand why this is happening, because it is a good, low noise, low PSRR voltage converter. Why exactly does an op amp work well there? And did it mean that in the VCCA, DVCC, Clock ADM lines would work too badly?
Today I received a Crystek oscillator. I already dropped the old, it was very easy to do. I just warmed up the metal case with a soldering iron at 350C and he fell off himself 🙂
There is one more spot for soldering a smaller oscillator. What do you think, is it worth removing it, leaving only the necessary?
