O2 headamp output booster PCB - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Headphone Systems

Headphone Systems Everything to do with Headphones

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 19th October 2013, 06:21 AM   #11
diyAudio Member
 
Join Date: Feb 2009
There is a pretty good layer of plastic in these packages. You can easily file off 0.5mm or even more.
I still would fit some resistance between OPAs and LMEs supply rails

In this little, single battery powered project, second and forth harmonics dropped around 8-10dB when I increased timeconstant of that RC filter around front end opamp. And it was driving just around 1Vrms at 33Ohm.
Attached Images
File Type: jpg port_1.jpg (161.9 KB, 304 views)
File Type: png port_2.png (39.5 KB, 290 views)
File Type: png port_3.png (39.9 KB, 256 views)

Last edited by Sergey888; 19th October 2013 at 06:34 AM.
  Reply With Quote
Old 19th October 2013, 08:47 AM   #12
diyAudio Member
 
Join Date: Feb 2009
Addition
OPA1642 or OPA2140 can give you comparable (may be a tiny bit worse) performance for much lower price, and there is a dual part available.

Last edited by Sergey888; 19th October 2013 at 08:50 AM.
  Reply With Quote
Old 19th October 2013, 11:34 PM   #13
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Sergey888 - I agree about the package sanding. Taking 0.5mm or so off the top really shouldn't have any effect. Most of that plastic is for mechanical strength when the leads are bent. Having the chip body butt up tight against the top would seem to be good for heat sinking, but I know that very little heat actually transfers through the plastic body. Most of the heat transfer is through that metal tab. Having the entire PCB bottom as heat sink trace via those thermal vias should help quite a bit.

Intersting about those series resistors on the power supply leads! Makes sense. I don't think I've read a thread about that anywhere but it sounds like a good idea. Device PSRR is usually frequency dependant and drops off at higher frequencies. Your power lead RC combination would help reject higher frequency stuff on the rails.

I've read read several posts about those OPA164x chips. The specifications look great but most of the posters seem to think they didn't like the sound of the result. So far most of the posts about the end result of the OPA827s I've read have been positive. The OPA2140 I haven't heard of before. I'll take a look at that one.

I see that both the OPA827 and LME49600 are $2 less per chip at Arrow, the big chip house, than Mouser and they are willing to sell them in single quantities:

http://parts.arrow.com/item/detail/t...ents/opa827aid

That would save $8, although the other parts probably are not available from Arrow so a Mouser order would be needed anyway and both would have a shipping charge.

I forgot to post a picture showing the clearance underneath a PCB in that top slot vs. the top of the O2 board. Fits great. I've added pads for a resistor to V- on the LME49600 bandwidth pins in case someone is running it on AC and doesn't care about current draw. I made that one a small 0603 sized SMD though to save space, so if anyone wants those two resistors they will have to work for them a bit when solding.

All the wires go to the two DIP headers replacing the NJM4556A chips except for ground. I have it set up to get ground in any of 3 ways. Solder to the middle battery terminal that the PCB butt up against, solder through a hole I've put in front to a wire that goes around a case screw ( same way RS used to ground the case ) or solder a wire to another hole I added above the output jack to the center ground on that O2 hole he has next to the jack.

If the board works when I get it back, stuffed, and tested I'll post the Gerbers again.
Attached Images
File Type: jpg IMG_2075.JPG (135.6 KB, 290 views)

Last edited by agdr; 19th October 2013 at 11:53 PM.
  Reply With Quote
Old 20th October 2013, 05:37 AM   #14
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Sergey888 - it looks like you are right about the usefulness of those series resistors on the IC power rails. I was just looking at the OPA827 datasheet and was kind of surprised by the amount the PSRR drops off after 10Hz for the negative rail, figure 20 on page 8:

http://www.ti.com/lit/ds/sbos376h/sbos376h.pdf

By 20kHz the neg rail PSRR is down to 60dB.

Unfortunately I don't see a good way to fit the resistors in from a space standpoint, and I'm eager to get V1.0 out to fab, so I'm going to skip them on this round and add those resistors to my wish list for future revisions. I did manage to fit in the feedback caps on the 827 though.

I'm going to see if the mods would be willing to split this O2 output board stuff out to a separate thread since it really doesn't have anything to do with the ODA board.
Attached Images
File Type: png O2 output PCB OPA827 LME49600 layout lyr 1_4.png (129.2 KB, 274 views)
File Type: png O2 output PCB OPA827 LME49600 circuit.png (38.7 KB, 82 views)
  Reply With Quote
Old 20th October 2013, 07:09 AM   #15
diyAudio Member
 
Join Date: Feb 2009
You can try to attack only the negative one, which is worse. Also 0603 resistors solve a lot of problems with space.
  Reply With Quote
Old 20th October 2013, 11:26 PM   #16
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Thank you for the thread split, Mooly!

Sergey888 - I was thinking about a negative-rail-only RC! That much might fit, especially if I did go to 0805 or 0603 sized SMD for a few things. The net result should just be slighly reduced maximum output swing, hitting the negative rail first. I'm still thinking about whether that might alter the offset voltage. If nothing else I could just add 2 pads and try it in V2.0.

Well V1.0 is out to fab. With any luck I'll have it back in a week. I had to use Seeed Studio's 5cm x 10cm 4 layer service for the 28mm x 80mm board. I realized today that I could panelize 3 of these side by side (nice wide 8mm cut channels inbetween) and then use Seeed's 10cm x 10cm 4 layer service, which is only $20 more for 10 boards. The per-board cost, including DHL, would drop from $10.50 USD to $4.50 or so.

I have the files below also at a Google Drive link, but I have no idea how long the link will continue to work. My Goole-fu on permalinks isn't doing well.

https://drive.google.com/folderview?...Gc&usp=sharing

The V+ rail is now a polygon on layer 3 rather than traces, to match a polygon ground on layer 2 and the full V- plane on the bottom and top heat sink foil areas.

I've added optional 4.99k resistors across the O2's power rails, R1 and R2, to implement a no-thump mod I posted a while back. Puts a 10K resistive load across the load side of the O2's mosfets to allow the post-mosfet power rails to discharge linearly all the way to zero when the mosfets shut off. RocketScientist / NwAvGuy left it with just the chips as load on the mosfets. Their discharge rate becomes rather undefined when the rails drop below the chip's minimum operating voltage, which can result in one rail dropping more than the other, which can create a power-off thump.
Attached Images
File Type: png O2 output PCB OPA827 LME49600 layout lyr 1_4.png (129.2 KB, 64 views)
File Type: png O2 output PCB OPA827 LME49600 circuit.png (38.7 KB, 56 views)
Attached Files
File Type: pdf O2 output board.pdf (63.8 KB, 25 views)

Last edited by agdr; 20th October 2013 at 11:40 PM.
  Reply With Quote
Old 21st October 2013, 06:15 AM   #17
diyAudio Member
 
Join Date: Feb 2009
LME49880 may be a good alternative, and it is dual. It is a bit power hungry though.
  Reply With Quote
Old 21st October 2013, 04:07 PM   #18
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Quote:
Originally Posted by Sergey888 View Post
LME49880 may be a good alternative, and it is dual. It is a bit power hungry though.
Hey good find! I didn't know that chip existed. It looks like a FET input version of the LME49990. They say "total" on those quiescent current numbers, so I'm assuming they mean for both sections together, which would make the draw about the same as the 2 OPA827's. They pricing is fantastic, about $2.50 at Mouser for the dual chip. The one big hiccup I see so far is the offset voltage, about 5mV vs. the 125uV on the OPA827. I guess that is where the extra $$ goes, for more DC precision. This one would probably need a DC servo if used in an output stage.

Well this is a great chip find, Sergey888. I may take a look at this for the gain stage on V2.0 of the ODA. You mentioned once about having a FET input there. This might be the one! As a dual it would condense 2 chips into one and save some board space. The offset voltage wouldn't matter at all for the first stage since there is a blocking capacitor in the middle of the amp. The amp's input ground-return resistor could be raised to 50k from 10K to give a high impedance input to devices that need it. The johnson noise would be higher, but not for devices wehre the source impedance was less (parallel combo of the two). Interesting!

Edit: I just bought one soldered onto a DIP8 adaptor to mess around with. The adaptor even has the power pad soldered down. Might be fun to try it in for the gain stage in the O2, if the adaptor will fit.

Last edited by agdr; 21st October 2013 at 04:18 PM.
  Reply With Quote
Old 22nd October 2013, 06:24 AM   #19
diyAudio Member
 
Join Date: Feb 2009
You can not connect C5 C11 just like that. Resistors must be added between LME's output and opamp input. Sorry didn't noticed it earlier.
I tried it in simulator. With opa827 and without fitting anything at all the phase margin looks a bit to small. Fitting 10p + 1k resistor brings it to more reasonable level.
The alternative way would be using opamp with smaller GBW, like OPA1642(1) I mentioned before.

Last edited by Sergey888; 22nd October 2013 at 06:34 AM.
  Reply With Quote
Old 22nd October 2013, 07:34 AM   #20
diyAudio Member
 
Join Date: Feb 2009
As far as I remember the main driver for using JFET input opamp is a low input bias current.
Some BJT input devices have quite low input bias currents. Have a look at OPA209 or its dual version OPA2209.

Last edited by Sergey888; 22nd October 2013 at 07:44 AM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Zen headamp. PCB, my ver. Stream Pass Labs 43 21st May 2014 09:28 AM
FS - Fully populated & working O2 headamp Lil Knight Swap Meet 0 16th January 2013 03:30 AM
Grommes S35 booster amp output? Dill Pickle Tubes / Valves 1 9th October 2012 08:23 PM
WTT O2 headamp PCB for CCS PCBs regal Swap Meet 0 13th October 2011 04:42 AM
FS: PCB's for the Morgan Jones tube headamp Nisbeth Swap Meet 0 19th October 2004 08:43 PM


New To Site? Need Help?

All times are GMT. The time now is 05:01 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2