Merry Christmas!
Not glad it happened at Christmas exactly, so I'll try to enjoy and forget about it. I am disappointed because there won't be high end audio for the hollyday. But I am curious to find out the reason why it happened.
My setup is: Te7022L cheap board from eBay with a 12MHz oscillator, to AKM4396 with I2S connection.
Few days ago I decided to remove the SLA battery and Salas shunt regulators and try the DAC straight on NiCd -> 5V for the analog and digital part of AKM4396 and 3,2V (made by 3 diodes in series for the voltage drop for the Te7022L). The 5V of the same pack for the oscillator.
In total 3 packs of Ni-Cd. One for the analog part, one for the digital part of AKM4396 and one for the Te7022L board.
I plugged it on and I was very happy with the sound! Been listening for hours and hours and hours. Better sonics than with the SLA and Salas regs.
Yesterday I've been listening to music the whole day. I decided to let the system play at night, while I was sleeping.
So I woke up at 08AM to hear oscillation from my tubes, like if there was no signal. I saw on the computer screen a message that my USB device is no longer recognized.
I thought that my batteries were discharged, though I wonder how it would be possible, because the DAC draws no more than 50mA and each pack is 1000mAh(2x1000mAh), 5000mAh for the analog part. Digital draws about 10mA and analog 40mA.
So I charged them a bit. Then I plugged the DAC again and the same message appeared - Problem recognizing USB. Then I touched the diodes going to the Te7022L chip, they were hot. WTF I thought, then I touched the Te7022L chip and it was hot like hell!
I really don't know what happened. The diode voltage drop is OK. The Te7022L chip is fed at its nominal voltage.
Not glad it happened at Christmas exactly, so I'll try to enjoy and forget about it. I am disappointed because there won't be high end audio for the hollyday. But I am curious to find out the reason why it happened.
My setup is: Te7022L cheap board from eBay with a 12MHz oscillator, to AKM4396 with I2S connection.
Few days ago I decided to remove the SLA battery and Salas shunt regulators and try the DAC straight on NiCd -> 5V for the analog and digital part of AKM4396 and 3,2V (made by 3 diodes in series for the voltage drop for the Te7022L). The 5V of the same pack for the oscillator.
In total 3 packs of Ni-Cd. One for the analog part, one for the digital part of AKM4396 and one for the Te7022L board.
I plugged it on and I was very happy with the sound! Been listening for hours and hours and hours. Better sonics than with the SLA and Salas regs.
Yesterday I've been listening to music the whole day. I decided to let the system play at night, while I was sleeping.
So I woke up at 08AM to hear oscillation from my tubes, like if there was no signal. I saw on the computer screen a message that my USB device is no longer recognized.
I thought that my batteries were discharged, though I wonder how it would be possible, because the DAC draws no more than 50mA and each pack is 1000mAh(2x1000mAh), 5000mAh for the analog part. Digital draws about 10mA and analog 40mA.
So I charged them a bit. Then I plugged the DAC again and the same message appeared - Problem recognizing USB. Then I touched the diodes going to the Te7022L chip, they were hot. WTF I thought, then I touched the Te7022L chip and it was hot like hell!
I really don't know what happened. The diode voltage drop is OK. The Te7022L chip is fed at its nominal voltage.
Smoked TE7022?
I wonder if running the digital on the AK4396 at 5 was a problem for the TE7022? And was the 5 really 5? Not much of a load...did you measure the NiCADs right off the charger? And were you applying the V you THOUGHT, to the TE7022? Did you have say 100ohms in series with each I2s leg to the DAC?. Were these separate modules? Were they fer sure kept at same ground-potential? Did you ALWAYS run the oscillator at 5V Before? And how is that signal coupled to the TE7022? a Cap, a R? DIRECT? [ I would use a 3.3V oscillator? Or maybe 2 R's...a voltage divider for the clock pulses?] And the TE7022L internal 1.8V pins, you didn't slip up and make a bridge from one of them to the 3.3 did you? pins 9, 19, 24, and 40 have TE7022 1.8V if you bridged any of those... or to any pins that don't want +3.3...
I wonder if running the digital on the AK4396 at 5 was a problem for the TE7022? And was the 5 really 5? Not much of a load...did you measure the NiCADs right off the charger? And were you applying the V you THOUGHT, to the TE7022? Did you have say 100ohms in series with each I2s leg to the DAC?. Were these separate modules? Were they fer sure kept at same ground-potential? Did you ALWAYS run the oscillator at 5V Before? And how is that signal coupled to the TE7022? a Cap, a R? DIRECT? [ I would use a 3.3V oscillator? Or maybe 2 R's...a voltage divider for the clock pulses?] And the TE7022L internal 1.8V pins, you didn't slip up and make a bridge from one of them to the 3.3 did you? pins 9, 19, 24, and 40 have TE7022 1.8V if you bridged any of those... or to any pins that don't want +3.3...
1. Diodes are HER602
2. The DAC supply was unknown, but sure lower V before blowing up, because it ran for almost a day. 4,95V for the analog, 5,05V for the digital AKM4396 stage, and 4,98 and 3,4V for the digital part of TE7022L
3. I have no idea, I don't have a scope to verify.
4. I don't have a schematic, but I used this: TE7022L USB Module Card Assembled Board for DAC5 by Weiliang | eBay
The boards are connected via star ground
1. I have no idea.
2. It was really 5 in the first run, when I got the NiCds off the shop
3. After charging the NiCds were 5,4V.
4. Yes, I have 100 ohms with each I2S leg to the DAC
5. Yes, separate modules
6. Yes, same ground potential.
7. Yes, the oscillator had a 5V series regulator which I removed.
8. OSC signal is coupled via a 100R to the TE7022L.
9. Haven't bridged or touched anything to the 1,8V pins
In the worst case, 0,3 volts over the Te7022L and 5,4V on the OSC, not sure if these could have been immediately fatal. But I imagine the DAC has stopped working in the previous morning.
What I don't like at all is the lack a DETAILED DATASHEET about TE%@!@#$!@L !
How can I get an idea at least about the MAX SPECS?
Last edited:
http://www.yoyodyneconsulting.ca/downloads/General/ODAC/ODAC-release.pdf
http://www.gfec.com.tw/documents/10157/11043/TE7022SL_PB-v21.pdf
do YOU have a schematic of the TE7022 board from Min? And did Oscillator have a step-down to 3.3? Or it takes +5 from one of the 10 gold pins at card's edge? I KNOW you need 4 pins for I2s, 3-4 for the RAW USB and +/- PWr [on USB cable], ground...
DO you have separate AGND and DGND?
I am trying to find a TE7022 schematic SHOWING a TTL oscillator not a crystal, no luck YET, maybe I'll find it yet.
http://www.gfec.com.tw/documents/10157/11043/TE7022SL_PB-v21.pdf
do YOU have a schematic of the TE7022 board from Min? And did Oscillator have a step-down to 3.3? Or it takes +5 from one of the 10 gold pins at card's edge? I KNOW you need 4 pins for I2s, 3-4 for the RAW USB and +/- PWr [on USB cable], ground...
DO you have separate AGND and DGND?
I am trying to find a TE7022 schematic SHOWING a TTL oscillator not a crystal, no luck YET, maybe I'll find it yet.
"3. After charging the NiCds were 5,4V."
http://www.asahi-kasei.co.jp/akm/en/product/ak4396/ak4396_f00e.pdf
Says max 5.25V...some digital goes to 5.5V...not this one
http://www.asahi-kasei.co.jp/akm/en/product/ak4396/ak4396_f00e.pdf
Says max 5.25V...some digital goes to 5.5V...not this one
yep, overvoltage seems the problem, NiCd is without a doubt considerably noisier and higher impedance than your Salas reg and SLA. NiCd has a noise floor of around -80dB -> -90dB, a good 50dB (100000x) higher (at least) than the salas reg. using batteries directly is only going to equal or better shunt regs if you use quieter chemistries like LiFePO4 and NiMH isnt bad either, but their main advantage is isolation.
without any type of protection or current limiting, it will have just fried itself at the first sign of trouble. i'm surprised you didnt have DC on the output terminals to the amp->speakers.
so my advise is to just try and get it working how it was before.
without any type of protection or current limiting, it will have just fried itself at the first sign of trouble. i'm surprised you didnt have DC on the output terminals to the amp->speakers.
so my advise is to just try and get it working how it was before.
DUG, check the datasheet carefully. On MAX operating conditions, it is indicated 6.0 volts. 5,25 volts is the max recommended value.
Check this out (for battery noise) http://tf.nist.gov/timefreq/general/pdf/1133.pdf
Tomorrow I'm going to replace the I2S board with PCM2706 and see if the AKM4396 is intact.
Check this out (for battery noise) http://tf.nist.gov/timefreq/general/pdf/1133.pdf
Tomorrow I'm going to replace the I2S board with PCM2706 and see if the AKM4396 is intact.
Last edited:
Yes, it is good to 6V.
But I noticed another line where the inputs cannot be greater than VDD+.3V.
Maybe what Mooly and others suggested in that the supplies coming on in different orders or different discharge potentials may have caused damage.
I would NEVER run battery supply of any sort on any audio gear without very large electrolytic caps in parallel(after current limiting resistors or fuses) to lower impedance AND voltage regulation or max voltage limiting(i.e., zeners). I have personally found, although not the Tenor chips specifically, that the newer the digital audio chip(receiver, dac, etc.), the less tolerant it is of even momentary voltage peaks over it's ratings. For example, I've seen a number of Burr-Brown DIT4096 spdif transmitter chips completely fail after very brief exposure to 5.5V.
The Tenor part is here TE7022L - GFEC
See that this has 3V3 and 1V8 supplies, so the oscillator probably should be 3V3 as well.
No absolute max is given, but I would expect 3V6 for the main supply.
Power supply sequence is very important for multi-supply parts and take care that there are no inputs before the supply is up.
See that this has 3V3 and 1V8 supplies, so the oscillator probably should be 3V3 as well.
No absolute max is given, but I would expect 3V6 for the main supply.
Power supply sequence is very important for multi-supply parts and take care that there are no inputs before the supply is up.
i've seen quite a bit on battery noise and NiCD is one of the worst ive seen, it is certainly much worse than Salas regs. the paper you link mentions a single AA cell in measurements 20 years ago, no mention of a load higher than 0-1ma and current noise is shown separately. battery noise increases substantially with load because its the chemical reaction + output resistance that causes the noise.
the paper you link is using them as a bias element in measurement systems, not actually powering a load. as they mention, used like this the current is so low a single AA cell lasts for many years.... pretty different to running a much larger capacity (noisier already), much larger voltage (and again), much larger load (the kicker) or should I say, it actually has a load.
without a load there is little noise, so they are great for that purpose, while regulators create noise regardless because they need operating power. the situation changes drastically with load. I like batteries I use them a lot, but never NiCd.
the paper you link is using them as a bias element in measurement systems, not actually powering a load. as they mention, used like this the current is so low a single AA cell lasts for many years.... pretty different to running a much larger capacity (noisier already), much larger voltage (and again), much larger load (the kicker) or should I say, it actually has a load.
without a load there is little noise, so they are great for that purpose, while regulators create noise regardless because they need operating power. the situation changes drastically with load. I like batteries I use them a lot, but never NiCd.
Last edited:
http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=pcm2706&fileType=pdf
looks like the PCM2706 is a square flat IC, man if you don't have a breakout PCB for that you're pretty SOL. You place the mode pins in the right states and then the thing will spew out the 4 I2S signals and ground. You were on the right track before, find something that sends out I2s easily, and use that. JUST having the IC is no fun. I buy breakout stuff like that made for Arduino prototyping [which I could care less about]. I would place female pins on edge of a proto that holds the AK4396, or did you buy a Ak4396 DAC board and not put on all the parts? In that scenario there are usually 4x 100ohm resistors feeding I2s to AK4396 and ground, as you must know. I have never tried it but I guess you might be able to wire PCM2706 "dead bug style" [upside down glued to something upside down], wire-wrap wire...hmm, touchy. Breakout-board much easier!! If you bought a PCM2706 "USB sound card" it will not have easy access to I2s and would be in the wrong mode most likely. Various sellers have adopted RJ45 jack as a means to transmit I2s short distances [CAT5] although the pin-outs for that link may differ. Audio GD sold me TE8802 boards and RJ45 boards for those...some have the TE7022 out to RJ45 or regular onboard header holes for say machined gold female pins? I have a CM6631 assy with 5-pin white male plug [5P IDC ribbon]. Anyway, so you had battery power but still no isolation from common-mode noise on the USB DATA lines? I use a 12Mbps fiber optic isolator with remote power injection, I can hear the difference. MY hope is to try things like you are, but with some sort of standard card-edge or breakout system. Hard to justify in a way, since premade DAC boards are out there with AK4396 for not a lot, and I do not use op amps so don't have costly PSU expenses there. On the premade boards there are 4 discrete resistors for I2s [huge = easy]. Failing that I may make a DAC lab...I have purchased the ultra low noise regulators at 3.3, 3.6, and I think +5 [someone mentioned 1.8V that is something the Tenor TE7022L does on-chip]. I would probably use a 12V source, and then a 7808? then whatever low-noise 5V and 3.6 and 3.3 [3.6 for ESS ES9023 like ODAC does] or multiple 3.3V regulators is also very safe. And Foxconn? makes a tiny 3.3V 12Mhz TCXO [via Digi-key in the USA]. I tend to go caps-out but that ignores the differential... so my next phase is a Broskie Unbalancer made with 6N16B-V tubes and a $15 HV PSU, though I wonder about SSM2141 ICs on a board...ebay... I wish you luck with PCM270# but doubt it will be simple. You HAD the right idea from the start; TE7022 is a pretty nice place to be. My first experiences with TE7022 were modding the TE7022//WM8761 boards [warm and nice for what they cost, PSU mods and direct-out mod] , next the TE7022 CS4398 [same thing], then USB isolation. [I am using nothing but FLAC]. My Behinger rack-mount gear all uses AK4393 so I share your interest in the AK4396 [but not wrecking the gear with my learning SMD soldering].
looks like the PCM2706 is a square flat IC, man if you don't have a breakout PCB for that you're pretty SOL. You place the mode pins in the right states and then the thing will spew out the 4 I2S signals and ground. You were on the right track before, find something that sends out I2s easily, and use that. JUST having the IC is no fun. I buy breakout stuff like that made for Arduino prototyping [which I could care less about]. I would place female pins on edge of a proto that holds the AK4396, or did you buy a Ak4396 DAC board and not put on all the parts? In that scenario there are usually 4x 100ohm resistors feeding I2s to AK4396 and ground, as you must know. I have never tried it but I guess you might be able to wire PCM2706 "dead bug style" [upside down glued to something upside down], wire-wrap wire...hmm, touchy. Breakout-board much easier!! If you bought a PCM2706 "USB sound card" it will not have easy access to I2s and would be in the wrong mode most likely. Various sellers have adopted RJ45 jack as a means to transmit I2s short distances [CAT5] although the pin-outs for that link may differ. Audio GD sold me TE8802 boards and RJ45 boards for those...some have the TE7022 out to RJ45 or regular onboard header holes for say machined gold female pins? I have a CM6631 assy with 5-pin white male plug [5P IDC ribbon]. Anyway, so you had battery power but still no isolation from common-mode noise on the USB DATA lines? I use a 12Mbps fiber optic isolator with remote power injection, I can hear the difference. MY hope is to try things like you are, but with some sort of standard card-edge or breakout system. Hard to justify in a way, since premade DAC boards are out there with AK4396 for not a lot, and I do not use op amps so don't have costly PSU expenses there. On the premade boards there are 4 discrete resistors for I2s [huge = easy]. Failing that I may make a DAC lab...I have purchased the ultra low noise regulators at 3.3, 3.6, and I think +5 [someone mentioned 1.8V that is something the Tenor TE7022L does on-chip]. I would probably use a 12V source, and then a 7808? then whatever low-noise 5V and 3.6 and 3.3 [3.6 for ESS ES9023 like ODAC does] or multiple 3.3V regulators is also very safe. And Foxconn? makes a tiny 3.3V 12Mhz TCXO [via Digi-key in the USA]. I tend to go caps-out but that ignores the differential... so my next phase is a Broskie Unbalancer made with 6N16B-V tubes and a $15 HV PSU, though I wonder about SSM2141 ICs on a board...ebay... I wish you luck with PCM270# but doubt it will be simple. You HAD the right idea from the start; TE7022 is a pretty nice place to be. My first experiences with TE7022 were modding the TE7022//WM8761 boards [warm and nice for what they cost, PSU mods and direct-out mod] , next the TE7022 CS4398 [same thing], then USB isolation. [I am using nothing but FLAC]. My Behinger rack-mount gear all uses AK4393 so I share your interest in the AK4396 [but not wrecking the gear with my learning SMD soldering].
I have to laugh every time I see some expensive usb spdif interface or usb dac that does separate supply or battery & alleges to be high end, when NO isolation has been done on the data side of the usb, so still exposed to the noise pollution of the data ground from the pc. Was looking at a $1000 usb/spdif unit a few days ago that a friend was considering, and it was exactly like that- boasting about it's exotic supply, but no isolation on the data side. Sad. Too bad ethernet hasn't gotten popular for audio interfacing. It's full transformer isolated by format spec, AFAIK. USB is just a terrible choice for audio, IMO.
USB over ethernet
there IS a open source project for USB over Ethernet [however in what state] .
I use a Opticis M2-100-03 ? USB 12Mbps fiber optic isolator. I bought 36x and have 30 left to sell, no good for the 480Mbps world but awesome for the TE7022 12Mbps USB, honestly you can hear the lack of garbage. I will be making a eBay auction for the 30x as 1ea items soon they still sell at $189, I'll try my luck at about 1/2 that, keeping a few in reserve. I speak with forked ongue since I have various higher speed USB audio receiver assys around that would not work via traditional USB isolators.. Soon we will have theHifimediy uae23+
http://www.diyaudio.com/forums/atta...mediy-es9023-te7022-24-96-usb-dac-uae23-2.jpg
that unit will have 12Mbps USB isolation and external PWR jack with FW bridge after that.
PM me if you want at 12Mbps OPticis m2-100-03 3M fiber optic USB isolator at under 1/2 retail, and I cannot get any more at that price ever. [Includes switching 5V PS with ferrites at both ends. Or you supply 5V some other UUBER way.
there IS a open source project for USB over Ethernet [however in what state] .
I use a Opticis M2-100-03 ? USB 12Mbps fiber optic isolator. I bought 36x and have 30 left to sell, no good for the 480Mbps world but awesome for the TE7022 12Mbps USB, honestly you can hear the lack of garbage. I will be making a eBay auction for the 30x as 1ea items soon they still sell at $189, I'll try my luck at about 1/2 that, keeping a few in reserve. I speak with forked ongue since I have various higher speed USB audio receiver assys around that would not work via traditional USB isolators.. Soon we will have theHifimediy uae23+
http://www.diyaudio.com/forums/atta...mediy-es9023-te7022-24-96-usb-dac-uae23-2.jpg
that unit will have 12Mbps USB isolation and external PWR jack with FW bridge after that.
PM me if you want at 12Mbps OPticis m2-100-03 3M fiber optic USB isolator at under 1/2 retail, and I cannot get any more at that price ever. [Includes switching 5V PS with ferrites at both ends. Or you supply 5V some other UUBER way.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.