Jelle, Nice idea, but it's a bit late in the game to make changes like that. Rocket Scientist would have to redo the board to account for the different connector(s)(and number of connections), as well as the extra board track to connect them together..
Beside the fact that doing a search for trrs on Mouser, Digikey and Newark brings up nada in the way of connectors.. It might be too vendor specific
Part of the design criteria for the O2 was "easy to build with readily available parts"...
Beside the fact that doing a search for trrs on Mouser, Digikey and Newark brings up nada in the way of connectors.. It might be too vendor specific
Part of the design criteria for the O2 was "easy to build with readily available parts"...
Jelle Schrijver that's an interesting idea but, for now, you would have to find a panel mounted 4 conductor 3.5mm jack that could be added in addition to the on-board jacks. If you find a part number for such a jack, please share it here?
Mouser only had I think two styles of stereo through-hole 3.5 mm jacks that don't have metal bushings (which would have created an ugly input-ground to output-ground loop through the front panel). They don't have any TRRS 3.5mm through hole jacks. Even if you can find an acceptable 4 conductor jack somewhere it would require a new PCB layout.
But I can see it as a desirable feature in a portable amp. Perhaps if there's ever an SMT version of the O2 that could be incorporated.
Mouser only had I think two styles of stereo through-hole 3.5 mm jacks that don't have metal bushings (which would have created an ugly input-ground to output-ground loop through the front panel). They don't have any TRRS 3.5mm through hole jacks. Even if you can find an acceptable 4 conductor jack somewhere it would require a new PCB layout.
But I can see it as a desirable feature in a portable amp. Perhaps if there's ever an SMT version of the O2 that could be incorporated.
speaking of using mobiles with the O2, have you done any testing wrt RFI noise immunity? in particular response to the location signal chirp put out just before a call or message comes in, or when changing towers (or switching from 3g service to some other service in black spots) this is a problem with many portable amps and i believe has more to do with the input stage filter and grounding than the enclosure, which i believe unless made from mumetal provides very little by way of attenuation.
you arent using any really wideband devices so induced oscillations shouldnt be a problem? but the sidebands down in the audio range can produce painfully loud RFI with lowZ cans
you arent using any really wideband devices so induced oscillations shouldnt be a problem? but the sidebands down in the audio range can produce painfully loud RFI with lowZ cans
Last edited:
qusp, The O2 has input filtering and I've noticed zero problems with my (cdma) phone in my pocket all the time. It's also been used close to a GSM phone quite a bit without problems. Even out of the enclosure working with bare board and long cables strung all over I've not heard, or seen on the dScope or my oscilloscopes, any hints of a problem.
One of the best things you can do for RFI immunity is to keep the important loop areas small and I've taken pains to do just that with the O2 PCB layout.
I'm not a mumetal expert, but I believe it's only required for low frequency magnetic interference (like 60hz hum from a power transformer). Even fine wire mesh made out of just about any metal works great at cell phone frequencies (just look at the door to any microwave oven--that mesh is blocking 1000+ watts of cell phone-ish 2.4 Ghz RF!).
So the O2's input grounded all aluminum enclosure should provide excellent shielding. The issue is more what the cables pick up. I suspect if you wind the headphone, input, and AC power cable around your phone you will probably hear some of the usual noises.
As an aside, the O2 has DC blocking between the stages, but fully direct coupled amps without input filtering (like the Mini3) can easily rectify the RF at the op amp's input stage and it turns into substantial DC. You just hear "clicks" in the headphones but if you look at it on a scope it can be pretty nasty stuff depending on the nature of the RF. A sustained RF carrier signal creates sustained DC at the amp's output.
One of the best things you can do for RFI immunity is to keep the important loop areas small and I've taken pains to do just that with the O2 PCB layout.
I'm not a mumetal expert, but I believe it's only required for low frequency magnetic interference (like 60hz hum from a power transformer). Even fine wire mesh made out of just about any metal works great at cell phone frequencies (just look at the door to any microwave oven--that mesh is blocking 1000+ watts of cell phone-ish 2.4 Ghz RF!).
So the O2's input grounded all aluminum enclosure should provide excellent shielding. The issue is more what the cables pick up. I suspect if you wind the headphone, input, and AC power cable around your phone you will probably hear some of the usual noises.
As an aside, the O2 has DC blocking between the stages, but fully direct coupled amps without input filtering (like the Mini3) can easily rectify the RF at the op amp's input stage and it turns into substantial DC. You just hear "clicks" in the headphones but if you look at it on a scope it can be pretty nasty stuff depending on the nature of the RF. A sustained RF carrier signal creates sustained DC at the amp's output.
i've found that cable shielding (no matter how much, or what type) will do bugger all if the input filter is crap, or if overly wideband devices are used without bandwidth limiting. therefor i dont use any at all on any of my portable cables, in a portable situation imo its more trouble than its worth wrt microphonics and general physical transparency.
i figured you had it covered, but thought i would mention it, because i actually think its a big deal, lots of people use their phones, or other wifi devices either as source, or very close while they are using portable amps; some manufacturers have a handle on it, some dont seem to even factor it in.
i was only joking about the mumetal, it would cost as much as the entire amp and its pretty heavy. ive used it before in my dac (for AC as you say) its more about the layout imo, if apple and others can make phones that themselves sound pretty good without a hint of interference, then layout and design technique can stop it in the amp too.
it comes up every now and then over at HF, at the moment there is a certain tube hybrid portable that has a really bad case of it, my guess is the tube itself, being they are pretty wideband and its the VAS.
myself, i dont have any problem with any of my amps, especially now i'm running balanced, but even the SE units i have are pretty good in that regard, with only the odd tiny background chirp if i'm going through a black spot, non-3g seems to be much worse for some reason. with customs its really important, as not only are they lowZ and sensitive, but you cant just rip them out of your ears.
i figured you had it covered, but thought i would mention it, because i actually think its a big deal, lots of people use their phones, or other wifi devices either as source, or very close while they are using portable amps; some manufacturers have a handle on it, some dont seem to even factor it in.
i was only joking about the mumetal, it would cost as much as the entire amp and its pretty heavy. ive used it before in my dac (for AC as you say) its more about the layout imo, if apple and others can make phones that themselves sound pretty good without a hint of interference, then layout and design technique can stop it in the amp too.
it comes up every now and then over at HF, at the moment there is a certain tube hybrid portable that has a really bad case of it, my guess is the tube itself, being they are pretty wideband and its the VAS.
myself, i dont have any problem with any of my amps, especially now i'm running balanced, but even the SE units i have are pretty good in that regard, with only the odd tiny background chirp if i'm going through a black spot, non-3g seems to be much worse for some reason. with customs its really important, as not only are they lowZ and sensitive, but you cant just rip them out of your ears.
One past week, I have worked on my own O2 case and finally finished.
Just share my work process
Thank you for watching~
Just share my work process
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Thank you for watching~
I caught the wink, but just so everyone knows, as long as the source impedance is low (< 150 ohms or so) I don't think EMI is an issue with the O2 completely unshielded in non-hostile environments.
MoonL you should remove C1 by the AC power jack on your board (and your friend's board). It's been eliminated from the current design. That will improve the noise performance of your O2, especially without a ground plane underneath the PCB (i.e. using a wood enclosure). And C16 and C21 should be changed to 0.022 uF and C22 increased to 1.0 uF if you want to minimize the turn on, turn off, and low voltage shutdown transients. (C22 became C1 on the latest revision--just follow the current design documents if there's any confusion).
Have you seen this opamp?
I don't know if you have already seen this opamp, but I haven't been able to find much information about it.
It is $4.20 in 100 pieces from digikey. They also make a ridiculously expensive $75/ piece opamp, the muses 1 or muses 2.
Expensive doesn't always mean better, but I thought that the 8820 might be worth checking out. They also make an interesting looking integrated volume control chip, which might work in a higher end desktop version of the O2.
Since you already have this company's opamps in the O2, you probably already saw this opamp, but I just thought I might point it out in case it looked interesting.
Is it anything special, or just a marketing gimmick?
I don't know if you have already seen this opamp, but I haven't been able to find much information about it.
It is $4.20 in 100 pieces from digikey. They also make a ridiculously expensive $75/ piece opamp, the muses 1 or muses 2.
Expensive doesn't always mean better, but I thought that the 8820 might be worth checking out. They also make an interesting looking integrated volume control chip, which might work in a higher end desktop version of the O2.
Since you already have this company's opamps in the O2, you probably already saw this opamp, but I just thought I might point it out in case it looked interesting.
Is it anything special, or just a marketing gimmick?
I had not seen that, thanks! NJR is to be commended for keeping the slew rate to a sane value of 5 V/uS when most of the high-end op amps are needlessly fast for audio use which just creates other significant problems.
It might have some advantages in highly demanding applications like a phono preamp, mic-preamp, pro gear with a dozen stages (where little tiny things ultimately add up), etc. But I really doubt it will offer any audible, or perhaps even measurable, benefits in typical applications like the O2's gain stage, a DAC I-V converter, input/output buffer, etc.
But, that said, it would still be fun to test one.
An electronic volume control would be a cool thing to have to improve low end tracking and add a mute function, especially if it used an encoder and not those slow/annoying up/down buttons. Since the pot is in the middle on the O2 the 1st stage output voltage swing could go up to 2-3V from the rails. The volume chip would probably need to handle fairly high signal voltages. Unfortunately those njr volume chips don't appear to, but this one does:
ATtiny.com | MAS9116 / WM8816 Digital Stereo Volume Control
That fellow shows a rotary encoder feeding a small and cheap Atmel 8 bit microcontroller, the AT90S2313, off a separate 5V (digital) bus. Then he uses that to control the MAS9116 volume chip which will accept +/-17V rails and signal voltages. Looks like that chip might now be the MAS6116 and might not be widely available, but there are probably similar parts out there. Also looks like surface mount.
Last edited:
Digital controlled analog volume chips do have some advantages. And that might be a fun upgrade for the more experienced DIY person set up to program microcontrollers, etc. Or they could buy a ready made board with an already programmed MCU. It can also allow remote control.
The Cirrus/Crystal volume chips are the best ones I know of but they're expensive, only SMT, require an MCU with custom firmware, and need 3 additional power supply voltages. So they're out of the question for the desktop O2.
Even at 2.5X gain I have no audible channel balance problems with any of my headphones except at really low levels with the SuperFi 5's. But they should be run at 1X gain and, at that setting, they're also fine. Personally, I think the best solution is optimizing the gain so you stay out of the first few percent of volume control's range.
Desktop amp builders are also free to use DACT or other stepped attenuators, higher end bigger pots, etc. if they want. Short of the weird noises, Doug Self and some other high-end designers like Krell, have decided relay switched ladder attenuators are the best possible solution. They have less distortion than volume chips, they last longer than stepped attenuators, and they allow some interesting options when under software control including variable step sizes, etc.
The Cirrus/Crystal volume chips are the best ones I know of but they're expensive, only SMT, require an MCU with custom firmware, and need 3 additional power supply voltages. So they're out of the question for the desktop O2.
Even at 2.5X gain I have no audible channel balance problems with any of my headphones except at really low levels with the SuperFi 5's. But they should be run at 1X gain and, at that setting, they're also fine. Personally, I think the best solution is optimizing the gain so you stay out of the first few percent of volume control's range.
Desktop amp builders are also free to use DACT or other stepped attenuators, higher end bigger pots, etc. if they want. Short of the weird noises, Doug Self and some other high-end designers like Krell, have decided relay switched ladder attenuators are the best possible solution. They have less distortion than volume chips, they last longer than stepped attenuators, and they allow some interesting options when under software control including variable step sizes, etc.
That's a good point. Rather than include something on-board it may be best for folks wanting to mess with it to cable off-board (1st stage output and 2nd stage input) to whatever they want. The stepped attenuator, relay attenuator, or electronic volume control. Makes a good individual builder value-add.
Last edited:
I had not seen that, thanks! NJR is to be commended for keeping the slew rate to a sane value of 5 V/uS when most of the high-end op amps are needlessly fast for audio use which just creates other significant problems.
It might have some advantages in highly demanding applications like a phono preamp, mic-preamp, pro gear with a dozen stages (where little tiny things ultimately add up), etc. But I really doubt it will offer any audible, or perhaps even measurable, benefits in typical applications like the O2's gain stage, a DAC I-V converter, input/output buffer, etc.
But, that said, it would still be fun to test one.
I wish they were available in smaller quantities. I bet that we could do a group buy, I would like to try it in a hagerman bugle phono preamp, and maybe the o2 when I build one.
I found a commercial amp that uses one. For sale here, and reviewed Korean Pocket Rocket: The Audinst Amp-Hp | Headfonia.
Not a very good review, and not much info is given on the design. With a 3.6 volt battery it has to be handicapped in some way.
Sorry for the derail. Hope that you are able to test one somehow. Maybe someone here has a spare. I certainly would be curious to find out what you think of it.
Thank you. But thinking of it. It will create a ground loop with normal TRS jacks. So I guess it will ruin the measurements. I did not do my homework if these parts where available. I will have a look around.
As long as where dreaming:
1) Tilt EQ
2) Crossfeed
3) Big beefy batteries (I will use it mainly portable)
4) A option to mix in the microphone signal, so you can keep hearing the enviroment.
5) Build in a HQ SDcard musicplayer with simply display and controls (like the QLS one)
6) And how about a bluetooth module, so you can still make phonecalls with your setup without haven to switch or take out headphones.
Or am I now discribing the clip+ rockbox'ed? ;-)
@Jelle, I don't think using the ground system of the O2 would hurt the phone's performance with the TRRS jacks. But I've never tried. You wouldn't want to run an extra ground wire between the jacks. Just let the grounds be connected through the O2 PCB as they are now.
It would also be possible to get some little tiny PCBs made and fit them with SMT TRRS jacks and small "L" mounting brackets or otherwise get creative if you can't find panel mounted TRRS jacks.
But it sounds like you would be happier with a more feature laden commercial solution. As Shike said in his review, the O2 isn't really "wearable" anyway. He prefers his FiiO E5 for that and I have to agree with him. The E5 makes the Clip+ look bulky.
It would also be possible to get some little tiny PCBs made and fit them with SMT TRRS jacks and small "L" mounting brackets or otherwise get creative if you can't find panel mounted TRRS jacks.
But it sounds like you would be happier with a more feature laden commercial solution. As Shike said in his review, the O2 isn't really "wearable" anyway. He prefers his FiiO E5 for that and I have to agree with him. The E5 makes the Clip+ look bulky.
With a TRRS jack the 2e r carries the mic signal. so 2e R should be passed straight through from output to input. But when you use a normal TRS. The 2e R from the terminal connects to the S from the jack. In that case it will have a ground loop. So it's not a ground wire but the mic/remote wire that is gone be connected to ground when a normal TRS jack is used instead of a TRRS.
But is there such a feature laden device that can meassure up to O2?
It seems to me a rockboxed clip+ goes a great way, but only is limited on battery capacity.
RocketScientist, how about board mount Mini XLR's?
Since we are all DIYers who build our own equipment, why do we use inferior connectors on our builds? If you can build an O2, you can recable a headphone, or build a cable.
I would like to see a board where something like these board mount Mini XLR's were used, or at least a board where both holes are put on the board so they are at least an option.
TRS connectors and RCA's are terrible designs that are bettered now. It would be nice to easily get away from them when we are building something. It could even make your measurements a little (I mean very little) bit better. Perhaps you could consider this for the desktop version, or a revision down the road?
Since we are all DIYers who build our own equipment, why do we use inferior connectors on our builds? If you can build an O2, you can recable a headphone, or build a cable.
I would like to see a board where something like these board mount Mini XLR's were used, or at least a board where both holes are put on the board so they are at least an option.
TRS connectors and RCA's are terrible designs that are bettered now. It would be nice to easily get away from them when we are building something. It could even make your measurements a little (I mean very little) bit better
. Perhaps you could consider this for the desktop version, or a revision down the road?