RS! Actually I found these - 1.2M, 1.8M and 2.2M. Do you still prefer the 1.2M? I would prefer the 2.2M to have as long sessions as possible between reloads. Just let me know, soldering station is hot now...
All the thanks to RS fredoAlfredo - I just have a problem needing a fix. Amazing amp indeed.
Brgds
All the thanks to RS fredoAlfredo - I just have a problem needing a fix. Amazing amp indeed.
Brgds
Last edited:
OK, I tried with the 2.2M. There is now a quite loud clicking sound (about 1Hz) when the batteries run flat so I believe the duty cycle is down to 5/95 now. But it still wont shut off. I will now change to the 1.8M and then charge it for 5 mins and do another test.
1.8 gave silence at first but I waited for about 15s and it started again. It is now at 2Hz and the dutycycle has altered. The feeling I get is that the sound is smoother in the audible part of the cycle. I will as a last test test with what you proposed - 1.2M ;-)
Brgds
1.8 gave silence at first but I waited for about 15s and it started again. It is now at 2Hz and the dutycycle has altered. The feeling I get is that the sound is smoother in the audible part of the cycle. I will as a last test test with what you proposed - 1.2M ;-)
Brgds
Last edited:
OK, 1.2M is almost as 2.7M in my opinion with one exeption - it stayed quiet for 15s. I'll put everything back in the case now. A new volunteer is needed for further testing since I saw the original behaviour more as a signal to get som AC into it.
Now, you construct a nice top notch DAC RS ;-)
Brgds
Now, you construct a nice top notch DAC RS ;-)
Brgds
Thanks Turbon. That helps. 15 seconds of off time is a big improvement at least as it gives you plenty of time to shut off the amp before it goes into special effects mode.
I've done some experimentation with a bench supply here, and this should help even further:
R25 = 1.2M - 1.5M (or parallel a second 2.7M on top of the exiting 2.7M)
R9 = 33K - 36K (or parallel 270K on top of the existing 40.2K)
Lowering R9 to around 34K significantly raises the "turn back on voltage" to where it should stay off for a longer period of time if not until you power cycle the amp and/or charge the batteries.
I've done some experimentation with a bench supply here, and this should help even further:
R25 = 1.2M - 1.5M (or parallel a second 2.7M on top of the exiting 2.7M)
R9 = 33K - 36K (or parallel 270K on top of the existing 40.2K)
Lowering R9 to around 34K significantly raises the "turn back on voltage" to where it should stay off for a longer period of time if not until you power cycle the amp and/or charge the batteries.
You asked!
1) The DPDT push button switch should be labelled S1 and S2 - right now it just says S1 but the qty is correctly listed as 2. That probably happened when the gain switch was added later in the game.
2) Mouser seems to have dropped the cost of all the RN50C resistors to $0.31, like this one
RN50C1001FRE6 Vishay/Dale Metal Film Resistors - Through Hole
I think they were around $1 when the BOM was created. So maybe consider switching all the gain/feedback/input resistors back over. I think you had them as the RN50C originally. Just a few cents extra for the super low noise units.
3) One particular set of the BOM mosfets have 30V gate-source breakdown while all the others have 25V. Maybe switch over the "default" recommend to those 30V guys - they are what I'm using in my builds. Just a little extra voltage margin there for little-to-no additional $.
FQPF10N20C Fairchild Semiconductor MOSFET Power
FQU11P06TU Fairchild Semiconductor MOSFET Power
FWIW Avnet Express has them too and they also have the gold Tyco DIP8 sockets.
4) May be worth considering making the WAU16-400 the "default" transformer, with the 12V/200mA unit as one of the options. Just provides a little more current and voltage overhead, although no problem of any kind here of course. However - I just checked and both Mouser and Allied Electronics are out of the WAU16-400. Mouser has 71 WAU16-1000s left and Allied has 2 though. Mouser shows more -400s arriving 12/1 and only 2 weeks lead time on either, so no big deal there.
5) More dicey as to any actual benefit to do this, but consider switching C7 over to this guy
http://www.mouser.com/Search/Produc...tualkey64800000virtualkey81-RDEC71E225K1K103B
as per figure 17 here
http://www.fairchildsemi.com/ds/LM/LM7805.pdf
The O2 rocks!
Last edited:
One more came to mind and I'm past my 30 minute edit period.
6) consider maybe making the Panasonic 470uF 35V caps that I used in my CRC mod the default cap, with the Nichicon as one of the optional caps
http://www.mouser.com/ProductDetail/Panasonic-Electronic-Components/EEU-FR1V471B/?qs=sGAEpiMZZMtZ1n0r9vR22eHFyjIavHtVAPa3WbxKD1I%3d
Even though there is no CRC filter here to cause more input ripple, these guys have twice the ripple current rating (even after de-rating for 60Hz) and 4 times the life rating - 8000 hours - than the Nichicons for just 23 cents more. Just a nice lifetime upgrade for not a lot of additional $$. Mouser has 2000 of them in stock. They are exactly the same lead spacing, diameter, and height as the Nichicons. The cover and lettering is even the same color! You can see them in one of the photos in my CRC mod post. The font of the lettering is just a bit different and in a different place on the label.
Mouser has these listed as a "new" part. I don't even think they were an option when you were making up the BOM.
6) consider maybe making the Panasonic 470uF 35V caps that I used in my CRC mod the default cap, with the Nichicon as one of the optional caps
http://www.mouser.com/ProductDetail/Panasonic-Electronic-Components/EEU-FR1V471B/?qs=sGAEpiMZZMtZ1n0r9vR22eHFyjIavHtVAPa3WbxKD1I%3d
Even though there is no CRC filter here to cause more input ripple, these guys have twice the ripple current rating (even after de-rating for 60Hz) and 4 times the life rating - 8000 hours - than the Nichicons for just 23 cents more. Just a nice lifetime upgrade for not a lot of additional $$. Mouser has 2000 of them in stock. They are exactly the same lead spacing, diameter, and height as the Nichicons. The cover and lettering is even the same color! You can see them in one of the photos in my CRC mod post. The font of the lettering is just a bit different and in a different place on the label.
Mouser has these listed as a "new" part. I don't even think they were an option when you were making up the BOM.
Last edited:
Turbon - good deal! I forgot to mention that the first mosfet listed, the FQPF10N20C, is the default right now on the BOM. So it would just be a matter of changing the other one on the BOM to the FQU11P06TU so that both have Vgs(max) = 30V. Note that one of these is an insulated TO-220 package and the other is an ipak, so they look a little odd right next to each other.
The 470uF caps are about 22 cents more x 4, so #6 would raise the total cost of the O2 by $1.
The resistors and 2.2uF cap all together are probably good for around another additional $1.50.
The WAU16-400 would add another $5 or so to that part of the BOM. Pondering it some more, a better suggestion for the default transformer may be the WAU12-500 at Mouser:
WAU12-500 Triad Magnetics Plug-In AC Adapters
That one one is $7.50 so would only be $2 more than the WAU12-200 but have 2.5 times the current available, even slightly better than that 400mA for the WAU16-400. The 12VAC rating would keep the efficiency high since the resulting DC is just enough overhead for the regulator chips.
From the data sheet the no-load voltage is given as 13.9VAC
http://system.netsuite.com/core/media/media.nl?id=5905&c=ACCT126831&h=d3e7ccb2166cb562df19&_xt=.pdf
which is great, similar to what RocketScientist designed for with the WAU12-200 (13.5VAC no load). That means we won't run into that issue of the transformer being designed for halogen lights and having loo low of a voltage.
Also a comment on those capacitor lifetime ratings. I know that RocketScientist knows this but I don't want to unnecessarily concern anyone. Those ratings are given at very high temperatures (cap max rated temp), 85C (185 F) for the Nichicon and 105C (221 F) for the Panasonic. The manfuacturers test the caps at those max temperatures for those given lifetimes (2000 hours and 8000 hours), then drop them back down to room temp (20C) and make sure the caps still meet specs. So if you just run them at room temperature or slightly higher the lifetime will be much greater than these numbers. That is why the TV set doesn't croak after only 1 year. I'm just a reliability freak and when I put on my "smart shopper" hat the 22 cents for 4x life rating seems like a good deal.
I forgot to mention that the first mosfet listed, the FQPF10N20C, is the default right now on the BOM. So it would just be a matter of changing the other one on the BOM to the FQU11P06TU so that both have Vgs(max) = 30V. Note that one of these is an insulated TO-220 package and the other is an ipak, so they look a little odd right next to each other. The 470uF caps are about 22 cents more x 4, so #6 would raise the total cost of the O2 by $1.
The resistors and 2.2uF cap all together are probably good for around another additional $1.50.
The WAU16-400 would add another $5 or so to that part of the BOM. Pondering it some more, a better suggestion for the default transformer may be the WAU12-500 at Mouser:
WAU12-500 Triad Magnetics Plug-In AC Adapters
That one one is $7.50 so would only be $2 more than the WAU12-200 but have 2.5 times the current available, even slightly better than that 400mA for the WAU16-400. The 12VAC rating would keep the efficiency high since the resulting DC is just enough overhead for the regulator chips.
From the data sheet the no-load voltage is given as 13.9VAC
http://system.netsuite.com/core/media/media.nl?id=5905&c=ACCT126831&h=d3e7ccb2166cb562df19&_xt=.pdf
which is great, similar to what RocketScientist designed for with the WAU12-200 (13.5VAC no load). That means we won't run into that issue of the transformer being designed for halogen lights and having loo low of a voltage.
Also a comment on those capacitor lifetime ratings. I know that RocketScientist knows this but I don't want to unnecessarily concern anyone. Those ratings are given at very high temperatures (cap max rated temp), 85C (185 F) for the Nichicon and 105C (221 F) for the Panasonic. The manfuacturers test the caps at those max temperatures for those given lifetimes (2000 hours and 8000 hours), then drop them back down to room temp (20C) and make sure the caps still meet specs. So if you just run them at room temperature or slightly higher the lifetime will be much greater than these numbers. That is why the TV set doesn't croak after only 1 year.
I'm just a reliability freak and when I put on my "smart shopper" hat the 22 cents for 4x life rating seems like a good deal.
Last edited:
Thanks for all the suggestions agdr. I'll put some of them in place.
The low noise resistors are genuinely useless for gains less than 7X. I've run tests of three different resistors: The Xicon generic ones, the Vishay SFR16S series, and the military spec RN50 series. Likely because the critical resistors in the O2 are already low values, any differences are only a fraction of a dB worth of noise (less than a 1% change) except at really high gains (> 7X) where it might be a single dB difference.
The RN50 resistors are also the only thing on the BOM that are not RoHS compliant (they have lead in them) which could be an issue for some--especially those wanting to sell assembled O2 boards or amps.
The best "upgrade" Mouser AC adapter is, no contest, the WAU16-400. The WAU12-500 isn't worth the extra money over the WAU12-200 unless the cheaper one is out of stock. The comments in the Obtaining The Components section go into the isolated cases where the WAU12-200 might not be the best choice.
The Panasonic caps are perhaps worthwhile for those who might run their O2 7x24 for a decade on AC power. But they're over four times more expensive. Yeah it's only a buck total, but a buck here and a buck there adds up for those on a tight budget. I'll add them to the options list in place of the Panasonics that are currently listed.
Right now Blogger is having behind the scenes problems so it's not easy for me to update the O2 Details article. They changed their security policies on their servers and offline editors no longer have access to blogs--including mine. So I can't retrieve and edit the current O2 Details article (I have only an older backup).
Google is aware of the problem but has been typically silent about communicating what they're planning to do about it. From their perspective you're just supposed to use the really annoying WYSIWYG web-based editing tools, but they're largely incompatible with normal HTML--such as my articles.
This is yet more proof the "Google Cloud" still has some ugly storms lurking within. From now on I'll be keeping offline current copies of everything rather than periodic backups. Google seems really good at coming up with cool stuff but not always so good about keeping it running smoothly once its established (Gmail and web search, of course, excepted). As you can read at the start of this thread Google Docs also has had problems.
The low noise resistors are genuinely useless for gains less than 7X. I've run tests of three different resistors: The Xicon generic ones, the Vishay SFR16S series, and the military spec RN50 series. Likely because the critical resistors in the O2 are already low values, any differences are only a fraction of a dB worth of noise (less than a 1% change) except at really high gains (> 7X) where it might be a single dB difference.
The RN50 resistors are also the only thing on the BOM that are not RoHS compliant (they have lead in them) which could be an issue for some--especially those wanting to sell assembled O2 boards or amps.
The best "upgrade" Mouser AC adapter is, no contest, the WAU16-400. The WAU12-500 isn't worth the extra money over the WAU12-200 unless the cheaper one is out of stock. The comments in the Obtaining The Components section go into the isolated cases where the WAU12-200 might not be the best choice.
The Panasonic caps are perhaps worthwhile for those who might run their O2 7x24 for a decade on AC power. But they're over four times more expensive. Yeah it's only a buck total, but a buck here and a buck there adds up for those on a tight budget. I'll add them to the options list in place of the Panasonics that are currently listed.
Right now Blogger is having behind the scenes problems so it's not easy for me to update the O2 Details article. They changed their security policies on their servers and offline editors no longer have access to blogs--including mine. So I can't retrieve and edit the current O2 Details article (I have only an older backup).
Google is aware of the problem but has been typically silent about communicating what they're planning to do about it. From their perspective you're just supposed to use the really annoying WYSIWYG web-based editing tools, but they're largely incompatible with normal HTML--such as my articles.
This is yet more proof the "Google Cloud" still has some ugly storms lurking within. From now on I'll be keeping offline current copies of everything rather than periodic backups. Google seems really good at coming up with cool stuff but not always so good about keeping it running smoothly once its established (Gmail and web search, of course, excepted). As you can read at the start of this thread Google Docs also has had problems.
RocketScientist - all sounds good.
I was wondering if those RN50C resistors would even register a difference on your equipment given how low the noise is already. I thought you dropped them due to dollar cost but apparently was lack of benefit, too.
Agreed about Google! Very frustrating to deal with them.
Mouser may have bumped the price way up on those Nichicon 470uFs on you. They show them at $0.48 right now but I just noticed that the BOM has them at 8 cents.
UVR1V471MPD1TD Nichicon Aluminum Electrolytic Capacitors - Leaded
So the total case cost of the O2 on the BOM should be $1.60 higher at $30.33. Those Panasonics are "only" 50% more then. That is a good idea you have - listing the Panasonics as one of the options rather than the default. Saves $1 more on the base cost of the O2 and folks can make the change if they wish. The BOM must be where I got 8 cents from in my CRC mod writeup. I had used that number in the dollar cost math, then later couldn't find out where I got it from!
I just did a search on Octopart. Both Future Electronics and Allied have the Nichicons listed for around 8 cents, but that is minimum of 500 pieces in both cases, following the "buy now" links out to the sites.
Octopart Part Search
I was wondering if those RN50C resistors would even register a difference on your equipment given how low the noise is already. I thought you dropped them due to dollar cost but apparently was lack of benefit, too.
Agreed about Google! Very frustrating to deal with them.
Mouser may have bumped the price way up on those Nichicon 470uFs on you. They show them at $0.48 right now but I just noticed that the BOM has them at 8 cents.
UVR1V471MPD1TD Nichicon Aluminum Electrolytic Capacitors - Leaded
So the total case cost of the O2 on the BOM should be $1.60 higher at $30.33. Those Panasonics are "only" 50% more then. That is a good idea you have - listing the Panasonics as one of the options rather than the default. Saves $1 more on the base cost of the O2 and folks can make the change if they wish. The BOM must be where I got 8 cents from in my CRC mod writeup. I had used that number in the dollar cost math, then later couldn't find out where I got it from!
I just did a search on Octopart. Both Future Electronics and Allied have the Nichicons listed for around 8 cents, but that is minimum of 500 pieces in both cases, following the "buy now" links out to the sites.
Octopart Part Search
Last edited:
Those are places to connect an external input jack (P1) and an external output jack (P2). For example if you were putting the O2 PCB in a larger B3-080 chassis and using the additional panel space for larger 1/4" input and output jacks. In that case you would just leave the PCB jacks off the board entirely and run wires from those P1 and P2 holes to the 1/4" jacks mounted in holes right on the front panel.
I need to order the B2 enclosure very soon and I think I'll just go with newark to canada since I'm not sure if one of the forum members has too much on his hands right now to sell me one. So when I'm ordering on that site, it's asking me a few questions I'm unsure of:
-"Is this shipping to a U.S. freight forwarder for export?: Yes or No?"
-"Backorder preference: Backorder allowed or Ship order complete?"
And they also have a *required company name field which I'm not sure what to do since it's just a personal order
-"Is this shipping to a U.S. freight forwarder for export?: Yes or No?"
-"Backorder preference: Backorder allowed or Ship order complete?"
And they also have a *required company name field which I'm not sure what to do since it's just a personal order
Regarding the ODA
If you plan to use a microcontroller for the audio decoding i would like to see a FTDIchip FT2232 device on it. Those might be a bit pricey but those provide one jtag interface for programming and then anyone can reprogram it (downloading a pre-compiled binary is simple given a detailed tutorial). Then anyone can update the software at any time. It also gives the feature of being able to control the entire device from the computer using an additional usb cable (and more software which some skilled guy can develop)...
I have tested other USB to serial devices (not with jtag) and their driver are terrible (especially prolific). I have never had any trouble with the FTDI drivers on either XP, Win7 or linux (various versions).
The AckoDAC uses the hiface usb board, but that one isn't very fun at the moment because the driver development for linux is dead in the water unless i have missed something (please enlighten me).
Was looking at the USB DAC - AKD-26, but i really want to see your design as well because that sounds promising and i like the measurements you do. The lack of measurements for the ackodac is a bit of a letdown because it looks cool...
https://sites.google.com/site/ackodac/home
I would really like to see you do some measurements of this and some buffalo DAC not to mention the ODA when it's up and running...
Regarding SMDs, im in for those. It takes a bit of time to get into it but in the end its not so difficult if you at least have a good tip and a decent station. I think most people should mange 0805 but otherwise you have the 1206 and those are huge. A bit of patience and some extra flux should get you far...
If you plan to use a microcontroller for the audio decoding i would like to see a FTDIchip FT2232 device on it. Those might be a bit pricey but those provide one jtag interface for programming and then anyone can reprogram it (downloading a pre-compiled binary is simple given a detailed tutorial). Then anyone can update the software at any time. It also gives the feature of being able to control the entire device from the computer using an additional usb cable (and more software which some skilled guy can develop)...
I have tested other USB to serial devices (not with jtag) and their driver are terrible (especially prolific). I have never had any trouble with the FTDI drivers on either XP, Win7 or linux (various versions).
The AckoDAC uses the hiface usb board, but that one isn't very fun at the moment because the driver development for linux is dead in the water unless i have missed something (please enlighten me).
Was looking at the USB DAC - AKD-26, but i really want to see your design as well because that sounds promising and i like the measurements you do. The lack of measurements for the ackodac is a bit of a letdown because it looks cool...
https://sites.google.com/site/ackodac/home
I would really like to see you do some measurements of this and some buffalo DAC not to mention the ODA when it's up and running...
Regarding SMDs, im in for those. It takes a bit of time to get into it but in the end its not so difficult if you at least have a good tip and a decent station. I think most people should mange 0805 but otherwise you have the 1206 and those are huge. A bit of patience and some extra flux should get you far...
@QRikard, thanks for the suggestion. I'm not currently planning an MCU-based interface but I have worked with the FT2232 on other projects and I agree FTDI does a better job with drivers. But a USB DAC cannot use an FTDI interface for high resolution native-driver audio. It cannot enumerate as a USB audio device and does not have the bandwidth needed. Such a device could only be used for programming and/or remote control of the DAC.
The ODA DAC will not require any drivers as the plan is it will conform to the standard USB audio specification which has native support in XP and later, OS X, and even Linux (although I haven't yet tested Linux).
The ODA DAC will not require any drivers as the plan is it will conform to the standard USB audio specification which has native support in XP and later, OS X, and even Linux (although I haven't yet tested Linux).
I'll vote for SMD's as well. They are doable even for a DYI'st. Buy some cheap testboard, a band of whatever resistors (0805), thin solderlead, thin tip and positive googles from your favourite gas station and you are all set for an educational trip.
Alcohol or other drugs are not recommended even if a large gin&tonic or two have been proposed in some posts.
Now, it isn't that hard and never mind the looks... My 9023 dac board looks like Pearl Habour -41, but those parts work as long as both ends are connected to the board.
RS! In Linux we have native support for UAC2, in windows only UAC1.
Alcohol or other drugs are not recommended even if a large gin&tonic or two have been proposed in some posts.
Now, it isn't that hard and never mind the looks... My 9023 dac board looks like Pearl Habour -41, but those parts work as long as both ends are connected to the board.
RS! In Linux we have native support for UAC2, in windows only UAC1.