Suggestion: for 25 cents more you can get the 16VAC 1000mA unit at mouser:
WAU16-1000 Triad Magnetics Plug-In AC Adapters
vs
WAU16-400 Triad Magnetics Plug-In AC Adapters
RocketScientist has them both on the BOM as alternatives. Take a look at the SPICE sims I posted in this section a couple of days ago for the O2 at max load. The (rms transformer secondary current rating) current de-rating factor for half wave rectifiers with capacitor input filters is 0.28 as per
http://www.hammondmfg.com/pdf/5c007.pdf
so that would give you about 112mA total, including amp quiescent, without saturating the trasformer core. That would be around 50mA per channel which is enough for a lot of - probably most - headphones. But with the 1000mA unit you can get the whole banana, 280mA max (130mA or so per channel plus quiescent), the amp is capable of.
I've seen some info about transformers generating noise bursts when the core comes out of saturation, but haven't been able to confirm it yet. May be total bunk - RocketScientist would have seen it in his noise measurements. But at the least core saturation on the current peaks would cause some temperature rise.
At any rate, if you are ordering an xformer from mouser anyway, may be worth paying the additional quarter.
The 1000mA unit is heavier any may add $1 to shipping too, don't know.
Last edited:
Thanks, agdr.
I might end up doing that just to have the versatility of an extra "wart" around.
This will be a portable amp for me, so it will be unplugged roughly 90% or more of the time. For my usage, the adapter is simply to recharge the batteries. However, I wasn't sure if that would happen if the output was less than 13.5V, which is why I questioned the use of the original 12-200.
As I've stated before though, my understanding of these things is very, very limited. It could be that the 12-200 is more than enough as a simple charger, even if it falls below the 13.5V minimum.
I *think* the diodes act like switches, and only allow current flow if the voltage is greater than 13.5. If that understanding is correct, then I'll probably get the 16-1000 or 16-400. If not, then the 12-200 is probably the right choice, as I won't need more for a simple charging circuit. Please feel free to correct me if I'm wrong.
Forgive my lack of knowledge about such things. If we were discussing something within my realm of study, I could actually ADD something to these boards. As it is, I fear I am purely a "taker" at this point.
Thanks again for everyone's help. You make DIY more fun.
I might end up doing that just to have the versatility of an extra "wart" around.
This will be a portable amp for me, so it will be unplugged roughly 90% or more of the time. For my usage, the adapter is simply to recharge the batteries. However, I wasn't sure if that would happen if the output was less than 13.5V, which is why I questioned the use of the original 12-200.
As I've stated before though, my understanding of these things is very, very limited. It could be that the 12-200 is more than enough as a simple charger, even if it falls below the 13.5V minimum.
I *think* the diodes act like switches, and only allow current flow if the voltage is greater than 13.5. If that understanding is correct, then I'll probably get the 16-1000 or 16-400. If not, then the 12-200 is probably the right choice, as I won't need more for a simple charging circuit. Please feel free to correct me if I'm wrong.
Forgive my lack of knowledge about such things. If we were discussing something within my realm of study, I could actually ADD something to these boards. As it is, I fear I am purely a "taker" at this point.
Thanks again for everyone's help. You make DIY more fun.
That is a good point about just being used for charging duty, tschuss. The maximum charging draw for just the batteries, without the amp powering anything, would be around [(11.8V - 8.0V) / 220R ] * 2 = 35mA, then add another 30mA or so for quiescent current from all the chips, for around 65mA or so. That current draw drops rapidly as the batteries charge, too.
The 200ma(rms) secondary would work out to be around (200mA)(0.28) = 56ma(dc) after the half wave rectifiers, so that 12v/200mA unit might be thin even for just battery charging. The 16VAC 400mA wall wart might do the best job in your case, for charging duty.
One thing I should add to the above post, that issue of "music power". These are all DC current calculations, which would directly apply to battery charging. But as for the amp playing music the occasional music peak can come out of the filter caps, giving you some additional current capability there. But I always pick the conservative side (how many peaks will there be in any given music, and how long will they last?) and go with the DC values, especially since the transformer price is nearly the same for 400mA as 1000mA.
The 200ma(rms) secondary would work out to be around (200mA)(0.28) = 56ma(dc) after the half wave rectifiers, so that 12v/200mA unit might be thin even for just battery charging. The 16VAC 400mA wall wart might do the best job in your case, for charging duty.
One thing I should add to the above post, that issue of "music power".
These are all DC current calculations, which would directly apply to battery charging. But as for the amp playing music the occasional music peak can come out of the filter caps, giving you some additional current capability there. But I always pick the conservative side (how many peaks will there be in any given music, and how long will they last?) and go with the DC values, especially since the transformer price is nearly the same for 400mA as 1000mA.
Last edited:
This is assuming it can output 13.5V while loaded, correct?
My concern is not about the *current* output capability of the 12-200, it's about the *voltage* output. If it can't supply a minimum of 13.5VAC under load, then the current is a moot point, right? The diodes simply won't "open" until they see that voltage.
Again, please forgive me if this understanding is not correct. I'm just trying to gain some knowledge.
Thanks!
Yes, I'm making that assumption.
You're correct about the diodes as switches, but the transformer voltage has to do with the ripple and peak charging voltage seen by the reservoir capacitors, combined with the action of the regulators. The regs need a bit of "overhead" (the dropout voltage) to maintain 12V regulation.
You're correct about the diodes as switches, but the transformer voltage has to do with the ripple and peak charging voltage seen by the reservoir capacitors, combined with the action of the regulators. The regs need a bit of "overhead" (the dropout voltage) to maintain 12V regulation.
I'm kicking around the idea of having an external DC option. It would deliver +/-15V to the regulator inputs using a DC-DC converter, and could be powered via battery pack, USB, lighter plug, etc. RocketScientist has some DC-DC coverage in his blog, and I'm thinking the regulators will provide maybe -30dB rejection, making it a little quieter than being directly connected. Charging convenience is obvious. Any advice/opinion/scolding that may be helpful to me?
Advice: Haha. Yeah, right. I receive the advice around here. I'm in no position to give it.
Opinion: That sounds great! I would really be interested in such an option, especially the ability to use a lighter plug.
Scolding: See "Advice", above.
I have to admit to a brain hiccup on my charger current calculations, above. Thinking about it some more this afternoon I realized that the typical half-wave rectifier math, as in that Hammond app note, is for the typical configuration of just one diode. With one diode the other half of the AC waveform gets thrown away, which is why the secondary current de-rating factor is less than 50%. Not very efficient.
But RocketScientist's circuit is essentially a double half-wave rectifier, or more accurately a voltage doubler circuit, which IS making use of the other half of the incoming AC waveform. Which means the load on the two channels has to be treated separately and matched up with the capability of each rail's half wave rectifier. In the case of the 12VAC 200mA(rms) transformer, each rail gets 200mA * 0.28 = 56mA max to avoid core saturation, which is better. However that gets mitigated a bit by my underestimating the total chip quiescent current draw. I whipped out the data sheets and the total for all the chips is around 42mA, with some of it rail-specific (the LM regulators) and some of it rail to rail (the op amps).
The net result is that to the 17mA or so of max charging current per battery, add around 30mA of quiescent current and you are at around 47mA per rail. So that 56mA available per rail from the 12vac 200mA(rms) charger works just fine, plus leaves some for the amp output itself (and even more as the batteries charge and the charging current goes down).
That result makes more sense as a double-check, given that RocketScientist did the original measurements with that particular adaptor and it worked just fine for the headphone loads he tried. For using the amp with music though, along with charging, that 16vac 400mA unit would be a good idea. Figuring in the two half wave rectifiers now, and adding in battery charging current which I forgot initially, that one would leave 70mA - 90mA free per channel (rail) for the amps themselves, much more than that 50mA I came up with above. That much current will probably cover most headphones just fine. And for the full 130mA or so per channel the amp is capable of, get that 16VAC 1000mA transformer.
Also FWIW I finally found a good writeup on why you want to avoid transformer core saturation in the first place
Core Saturation - Transformers
That "noise" I had read references to in saturation is harmonics of the line fundamental frequency (60Hz or 50Hz) that get generated by the core saturation, plus waveshape distortion occurs. The harmonics heat up the transformer and can conceivably find their way onto the power supply rails depending upon regulator and filter rejection.
But RocketScientist's circuit is essentially a double half-wave rectifier, or more accurately a voltage doubler circuit, which IS making use of the other half of the incoming AC waveform. Which means the load on the two channels has to be treated separately and matched up with the capability of each rail's half wave rectifier. In the case of the 12VAC 200mA(rms) transformer, each rail gets 200mA * 0.28 = 56mA max to avoid core saturation, which is better. However that gets mitigated a bit by my underestimating the total chip quiescent current draw. I whipped out the data sheets and the total for all the chips is around 42mA, with some of it rail-specific (the LM regulators) and some of it rail to rail (the op amps).
The net result is that to the 17mA or so of max charging current per battery, add around 30mA of quiescent current and you are at around 47mA per rail. So that 56mA available per rail from the 12vac 200mA(rms) charger works just fine, plus leaves some for the amp output itself (and even more as the batteries charge and the charging current goes down).
That result makes more sense as a double-check, given that RocketScientist did the original measurements with that particular adaptor and it worked just fine for the headphone loads he tried. For using the amp with music though, along with charging, that 16vac 400mA unit would be a good idea. Figuring in the two half wave rectifiers now, and adding in battery charging current which I forgot initially, that one would leave 70mA - 90mA free per channel (rail) for the amps themselves, much more than that 50mA I came up with above. That much current will probably cover most headphones just fine. And for the full 130mA or so per channel the amp is capable of, get that 16VAC 1000mA transformer.
Also FWIW I finally found a good writeup on why you want to avoid transformer core saturation in the first place
Core Saturation - Transformers
That "noise" I had read references to in saturation is harmonics of the line fundamental frequency (60Hz or 50Hz) that get generated by the core saturation, plus waveshape distortion occurs. The harmonics heat up the transformer and can conceivably find their way onto the power supply rails depending upon regulator and filter rejection.
Last edited:
re: underestimating the total chip quiescent current draw
I used a battery cutoff of 6.7V, which is 7V at the battery, due to D2; (~1V/cell was the design target). But when charging only, there is no current past the power switch. FWIW, I think RS spec'd the charge current at 2 times 24mAmax with 220 ohms and the Tenergy NiMH.
I used a battery cutoff of 6.7V, which is 7V at the battery, due to D2; (~1V/cell was the design target). But when charging only, there is no current past the power switch. FWIW, I think RS spec'd the charge current at 2 times 24mAmax with 220 ohms and the Tenergy NiMH.
Okay, so this was my first DIY project and I got down assembling the board. I'm afraid I may have screwed up the MOSFET though. When I was soldering it on, I accidentally bridged the solder between two of the pins.
Now I'm in the testing phase and everything checks out -- except when I check the DC voltage on the output (P2). If I plug into the wall, or if I have both batteries in, the voltage is very low (~3mV). However, when I pull one battery out and check, it is ~200mV). Did I mess up? How can I fix it?
Also, I just tried testing with some sacrificial headphones. There's no loud thump upon turn on/off, just a slight barely-noticeable thud. Also, the amp works when I play through AC-power, or when two batteries are in, but does not play music with only one battery in.
Thanks in advance for any help/advice I can get.
Now I'm in the testing phase and everything checks out -- except when I check the DC voltage on the output (P2). If I plug into the wall, or if I have both batteries in, the voltage is very low (~3mV). However, when I pull one battery out and check, it is ~200mV). Did I mess up? How can I fix it?
Also, I just tried testing with some sacrificial headphones. There's no loud thump upon turn on/off, just a slight barely-noticeable thud. Also, the amp works when I play through AC-power, or when two batteries are in, but does not play music with only one battery in.
Thanks in advance for any help/advice I can get.
Last edited:
Hi, all, First-time DIYer here. I have two questions I'm hoping you'll entertain. First, regarding the transformer: given the two transformers below and the default option on the BOM, which would be the appropriate choice when I'll be driving orthodynamic headphones (50 ohm) and low-impedance dynamic headphones? (Is there any reason for me to upgrade?)
Second, I'd like to be able to adjust the gain, if need be, so I'd like to install the collet sockets mentioned in the BOM, Mouser part 535-10-0518-10. However, this component is on backorder. Is there a Mouser alternative to order?
Thanking you in advance!
Second, I'd like to be able to adjust the gain, if need be, so I'd like to install the collet sockets mentioned in the BOM, Mouser part 535-10-0518-10. However, this component is on backorder. Is there a Mouser alternative to order?
Thanking you in advance!
@Questhate
You didn't harm the FET. Solder bridges are unavoidable, and common with SMT. You did right by checking your work. Don't, though, turn on the O2 with just one battery. It's not good for any of the equipment. There's even a protection circuit in the O2 to prevent damage.
@Harold
For transformers, I think it comes down to the 12-200 as minimum, the 16-400 as suitable for most phones at most listening levels, and 16-1000 as the "red line it" option.
I don't know much about the gain option sockets. It may have been removed to make room for the gain switch. A look at Mouser only turned up a 575-801010, but at twice the price.
You didn't harm the FET. Solder bridges are unavoidable, and common with SMT. You did right by checking your work. Don't, though, turn on the O2 with just one battery. It's not good for any of the equipment. There's even a protection circuit in the O2 to prevent damage.
@Harold
For transformers, I think it comes down to the 12-200 as minimum, the 16-400 as suitable for most phones at most listening levels, and 16-1000 as the "red line it" option.
I don't know much about the gain option sockets. It may have been removed to make room for the gain switch. A look at Mouser only turned up a 575-801010, but at twice the price.
Did you read Initial DIY Testing (bottom of the page).
A low DC voltage at P2 is good, high voltage is bad. Read the Low Voltage Shutdown section again, you may have a problem with the power management circuit. If it is working properly, pulling either battery should disable the power supply, so that no voltage is on P2.
Thanks guys for the responses so far.
I was running the tests from the Initial DIY Testing section on the blog. What a letdown that the very last test failed.
I tested both the P2 pads, as well as the J3 pins (center pin to the outer pins), and they give the same measurement. Very low DC (~3mV) from the AC or two-battery supply, and a much higher DC (~200mV) from only one battery. So it appears there is something wrong with the power management system.
I'm wondering if anyone knows if there is any way to localize the problem? Did one (or both) of the MOSFETs go bad (and how to measure)? Maybe one of the diodes got ESD damage when I dropped it on the carpet?
And what are the limitations of an amp with a faulty power management system? It appears to work fine with headphones, so I'm assuming that the power management circuit is only there to turn off the board when the battery voltage drops below a certain point? So, assuming I can't fix this problem, would this amp be suitable for desktop-only (no batteries) use without any adverse effects? I'd hate to waste this entire board.
Thanks for all the help guys.
I put one together last night- here are some shots of it if anyone is interested:
O2 Headphone Amplifier - Photography by Travis Dodd
Note: my version does not have the battery mounts or charging circuit. I wanted an A/C only amp.
Super easy to put together and started right up with 3.3mV/2.8mV DC offsets.
Still not sure how I'm going to enclose it...
O2 Headphone Amplifier - Photography by Travis Dodd
Note: my version does not have the battery mounts or charging circuit. I wanted an A/C only amp.
Super easy to put together and started right up with 3.3mV/2.8mV DC offsets.
Still not sure how I'm going to enclose it...
