If the LED isn't lighting at all, that's rather odd. And your voltages are even more odd.
With the amp turned OFF, AC adapter connected, and no batteries installed, carefully measure the voltage at the two terminals of C6 and the two terminals of C7 (by the regulators). Both should be very close to 12 volts . If not, there's likely a problem with one of the regulators or there's a serious short circuit somewhere before the power switch (in which case one or both of the regulators should be getting quite hot).
If C6 and C7 are ok, measure across each of the battery terminals. They should both be above 11 volts each. If not, check D1, D5, R1 and R2.
If the battery terminals measure OK with the amp off, turn the amp on. Do they still each measure above 11 volts? If so, the LED should be on. If it's not on, I would look for a problem with the LED, R6 or the power switch.
If the battery terminals measure OK with the amp on, and the LED is on, measure the voltage across C8 and across C9. If one has a lot higher voltage than the other, you likely damaged one of the MOSFETs (Q1 or Q2), or less likely there's a problem with U2 or somewhere else in the power control circuit.
If you read a bit over 11 volts across C8 and C9 the power supply is OK.
A reamer and a tapered round file do wonders for this type of work. I've done all of my front panels by hand, no issues (except for the amount of time spent). It is ill-advised by the people who can't do it or don't appreciate the "art"
Good luck! I should have mine finished as soon as this weekend
Great! Thank you for the encouragement! I am going to give it a try. I will go slow and steady...
If you think the voltages are odd meaning they are high, you would be right, as I just found out that my Fluke DMM is busted. It is reading voltages way too high! I used another DMM I have and the readings are more in line now.
Both C6 & C7 are measuring 11.9 volts with the amp ON and OFF, but the LED is still OFF. I will replace it and R6 both, as the power switch seems to be OK.
I am not getting identical voltages across C8 & C9. C8 measures 11.7 volts while C9 is measuring at 2.5 volts, which leads me to think that I need to replace Q2. I'm not sure how to test U2 or what else to look for that is possibly damaged.
@samer0214
Fix the LED first before trying to replace Q2, etc. If you're going to order parts feel free to order a spare Q2 just to be safe. But it's probably fine.
The LED is the voltage reference for the comparator circuit. There should be around 1.8 volts with the negative DMM probe on pin 4 of U2 and the positive probe on pin 5. If not, the comparator circuit (U2) will not work properly and that's very likely your problem.
With the power switch on, carefully verify there is about 23 volts across the bottom pair of pins on the power switch (pins 3 & 6 on the schematic)--the pair farthest from the edge of the board. If so, that rules out the switch.
If the above measures OK, the LED should be on unless there's a solder bridge across it or it failed. Measure across the LED. If it's more than 2 volts the LED is damaged, the wrong part, or in backwards. If it's under 1.7 volts it may be shorted, R6 may be the wrong value or not properly soldered, or a trace on the PC board is damaged. The LED and R6 are literally connected directly to the power switch contacts measured above. So there's nothing else to go wrong. It has to be a PC board trace, solder connection on the LED or R6, or one of the two parts.
Figure out what's wrong with the LED and your amp will very likely work fine.
Fix the LED first before trying to replace Q2, etc. If you're going to order parts feel free to order a spare Q2 just to be safe. But it's probably fine.
The LED is the voltage reference for the comparator circuit. There should be around 1.8 volts with the negative DMM probe on pin 4 of U2 and the positive probe on pin 5. If not, the comparator circuit (U2) will not work properly and that's very likely your problem.
With the power switch on, carefully verify there is about 23 volts across the bottom pair of pins on the power switch (pins 3 & 6 on the schematic)--the pair farthest from the edge of the board. If so, that rules out the switch.
If the above measures OK, the LED should be on unless there's a solder bridge across it or it failed. Measure across the LED. If it's more than 2 volts the LED is damaged, the wrong part, or in backwards. If it's under 1.7 volts it may be shorted, R6 may be the wrong value or not properly soldered, or a trace on the PC board is damaged. The LED and R6 are literally connected directly to the power switch contacts measured above. So there's nothing else to go wrong. It has to be a PC board trace, solder connection on the LED or R6, or one of the two parts.
Figure out what's wrong with the LED and your amp will very likely work fine.
Glad to report a succesful build, with everything checking out and working first time. Many thanks to RocketScientist for sharing his expertise and giving us all this excellent project; thanks also to all those involved in the group buys.
A couple of points - First, my O2 is considerably more powerful when running from the AC supply (15V 1A) than from batteries - Using a pair of HD25 Mk1, rated at 64 ohm, on AC it's possible to make them as loud as I would ever want with the gain switch set to the low position and the level about 2/3 max - however, on battery power, you need to set the gain to the high position and turn the level up almost to maximum to get the same output level. Still more than adequate for portable use, but I wonder if I might be better off investing in some different batteries, as it seems likely that they don't have enough current capacity - has anyone else noticed a similar phenomenon?
Secondly, I have been looking at finding a case - Farnell has already been mentioned as a suitable UK source, but their associated business, CPC, also stocks them - but they are priced differently - CPC charge £9.86 including VAT, whereas Farnell charge £11.21 for the same item. Of course, both businesses also have different minimum order and postage charges, but if you already use CPC, as I do, then they seem like the best deal, with free postage on orders over £45 excluding VAT.
A couple of points - First, my O2 is considerably more powerful when running from the AC supply (15V 1A) than from batteries - Using a pair of HD25 Mk1, rated at 64 ohm, on AC it's possible to make them as loud as I would ever want with the gain switch set to the low position and the level about 2/3 max - however, on battery power, you need to set the gain to the high position and turn the level up almost to maximum to get the same output level. Still more than adequate for portable use, but I wonder if I might be better off investing in some different batteries, as it seems likely that they don't have enough current capacity - has anyone else noticed a similar phenomenon?
Secondly, I have been looking at finding a case - Farnell has already been mentioned as a suitable UK source, but their associated business, CPC, also stocks them - but they are priced differently - CPC charge £9.86 including VAT, whereas Farnell charge £11.21 for the same item. Of course, both businesses also have different minimum order and postage charges, but if you already use CPC, as I do, then they seem like the best deal, with free postage on orders over £45 excluding VAT.
@greenalien, the gain of the O2 does not change from battery to AC. In fact, you can listen to the O2 while on battery power, plug in the AC adapter while the music is playing, and the volume will not change (unless perhaps it was badly clipping while on battery). If you're noticing dramatic differences, and have a DMM, check your battery voltages while the amp is playing fairly loudly and running on battery. Or perhaps you have D2 and/or D6 installed backwards, missing, or not soldered correctly?
The maximum output (which requires turning up the volume control while on AC to take advantage of) increases from about 5 Vrms on battery to about 7 Vrms on AC. That's only 3 dB difference.
The maximum output (which requires turning up the volume control while on AC to take advantage of) increases from about 5 Vrms on battery to about 7 Vrms on AC. That's only 3 dB difference.
As far as I know, the slew rate of an amplifier is a function of it's power output and required bandwidth. The higher the power output and the greater the bandwidth, the greater the slew rate required. Obviously, with a flea power headphone amp, not much of slew rate is required. Also consider the fact that headphones usually have much higher impedance than speakers, so I believe that further reduces the required slew rate.
I'm just starting out in High End Audio, and I'm very excited to pair up the O2 with my freshly ordered Sennheiser HD 650s. I mainly listen on my PC, so I'm still trying to pick out a nice DAC to complete the experience. I'm just wondering if I should wait for the Desktop version to come out. I'm really itching to get listening, but I may be able to wait if the Desktop version is coming soon. Otherwise, can I plan ahead and order certain components along with the portable version that will allow me to rebuild into the Desktop version once it's ready? Like which case is it going into? The Box Enclosures B3-080 was mentioned, just not sure if that's the one the Desktop version will be going into.
Thanks!
Thanks!
@dirkwright, you're half correct. It's the voltage output, not the power. And the impedance makes little difference (and in many cases has the opposite effect because many amps cannot swing as much voltage into a lower impedance load, so they need less slew rate).
The required bandwidth, in pretty much any rational audio playback application, is limited by the digital audio source or vinyl turntable. In all cases it's unusual to find anywhere close to a 0 dBFS (full output) signal at even 20 Khz let alone higher frequencies. But the O2 will play back 0 dBFS signal at full output into any load at 40 Khz without slew limiting even though such a signal doesn't exist anywhere beyond a test bench. For more see:
NwAvGuy Heaphone Amp Measurement Recommendations (InnerFidelity)
The required bandwidth, in pretty much any rational audio playback application, is limited by the digital audio source or vinyl turntable. In all cases it's unusual to find anywhere close to a 0 dBFS (full output) signal at even 20 Khz let alone higher frequencies. But the O2 will play back 0 dBFS signal at full output into any load at 40 Khz without slew limiting even though such a signal doesn't exist anywhere beyond a test bench. For more see:
NwAvGuy Heaphone Amp Measurement Recommendations (InnerFidelity)
Still, the voltage output of a headphone amp is pretty small, so obviously the required slew rate is small.
Here's the formula for calculating slew rate:
"Let's say, 500 watts out, 4 ohms, 50 kHz bandwidth at least. ( I know how to calculate slew rate)
SR= 2*pi*f*SQR(2*P*R) = 2*3.14*50000*SQR(2*4*500) = 19 V/us = absolute minimum. Some overhead is wanted, let's say 25-50 V/us is what you need."
Here's a thread on it:
http://www.diyaudio.com/forums/solid-state/47277-slew-rate-calculations-how-much-do-i-need.html
So, obviously the above formula is converting power @ a particular impedance into a voltage. For the Sennheiser HD650 headphones, with a sensitivity of 115dB/V, one volt should be plenty loud enough, so for a bandwidth of 40kHz, the minimum slew rate is 0.25V/us. Some feel that a good margin is required, like 10 times, so even then, your 5V/us should be plenty good enough.
@kiwimonk don't by anything for the ODA (Objective Desktop Amp) yet. Because it's likely going to have an internal hi-resolution USB DAC option the case may well be different. I hope to get the first ODA article published in December as well as start a new thread here on it.
@dirkwright the voltage swing on the O2 is about 20 volts peak-to-peak so that, along with whatever worst case signal you want to assume, determines the slew rate. The common rule of thumb from Doug Self, Bob Cordell and other amplifier experts is 0.2 V/uS per volt RMS of output. The O2 has a max output of 7 Vrms so that yields 1.4 V/uS. For a healthy margin, it's more than twice that.
The HD650 needs 2 - 3 Vrms so that's a slew rate of 0.4 - 0.6 V/us per the guideline.
I don't agree with 10 times the requirement. Once you get past around 2X or 3X anything further speed increase is often more of a liability with zero benefit. Faster than needed isn't better. It's like putting a Porsche 911 engine in your lawnmower. Excess speed often creates stability and other problems and requires other trade offs that may have a real-world negative impact on the amplifier. The science is pretty clear and black and white on slew rate (unlike say jitter).
@dirkwright the voltage swing on the O2 is about 20 volts peak-to-peak so that, along with whatever worst case signal you want to assume, determines the slew rate. The common rule of thumb from Doug Self, Bob Cordell and other amplifier experts is 0.2 V/uS per volt RMS of output. The O2 has a max output of 7 Vrms so that yields 1.4 V/uS. For a healthy margin, it's more than twice that.
The HD650 needs 2 - 3 Vrms so that's a slew rate of 0.4 - 0.6 V/us per the guideline.
I don't agree with 10 times the requirement. Once you get past around 2X or 3X anything further speed increase is often more of a liability with zero benefit. Faster than needed isn't better. It's like putting a Porsche 911 engine in your lawnmower. Excess speed often creates stability and other problems and requires other trade offs that may have a real-world negative impact on the amplifier. The science is pretty clear and black and white on slew rate (unlike say jitter).
You should go through the steps detailed on NwAvGuy's blog: NwAvGuy: O2 Details
Scroll down to the section titled "Initial DIY Testing". All the info is there.
@RocketScientist Awesome! With an internal DAC it's everything I need to get started. I think i'll build a portable anyway just to have that as well.
Now I'd really like to say Thank You very much for all the hard work you have put into the O2, and answering everyone's questions. Also, after reading a few of the reviews you've written, the way you honestly and scientifically lay things out has really helped me sort through the hype and marketing that rule the Hi-Fi audio world.
Can't wait to start listening!
Now I'd really like to say Thank You very much for all the hard work you have put into the O2, and answering everyone's questions. Also, after reading a few of the reviews you've written, the way you honestly and scientifically lay things out has really helped me sort through the hype and marketing that rule the Hi-Fi audio world.
Can't wait to start listening!
You really seem to like to argue with people, even when they are agreeing with you to some extent. You don't need to be so defensive.
You can argue with John Curl, Nelson Pass and others about the 10 times slew rate requirement. I was just repeating what they said.
You didn't state a bandwidth for your slew rate calculation.
slewrate=2xPixfxV
Sorry, I'm just on a mission to try and correct at least some of the more obnoxious audiophile myths--especially ones that often cause more harm than good like excessive slew rate.
I left bandwidth out of the equation because, as I explained in an earlier post, the bandwidth requirement is determined by the music you're playing. And nobody has ever demonstrated any music source that requires more than the 0.2 V/uS per volt RMS guideline. The guideline also allows full bandwidth sinewave bench testing to 20 Khz. Digital audio, and vinyl, are both inherently bandwidth and slew constrained. So, for that matter, are magnetic tape and even microphones. The only way to stress an amplifier beyond that guideline is to use an artificial signal.
So it doesn't matter if the amp has bandwidth to 200 Khz as sources of music don't. Not even SACD/DSD. If you want to side with Pass and Curl, that's fine. But you might want to read (if you haven't) what Self and Cordell have to say in their latest books about slew rate. They back up their claims with hard science, references, measurements, math, etc. Pass and Curl tend to be far more subjective and artful in how they dance around the science when it doesn't suit their claims and/or designs. I'm not going to start any fights with legends in audio, but if one of them wants to start a thread here on slew rate I'll gladly participate.
@Kiwimonk (and others), You're welcome. Thanks for the encouragement!
Last edited:
@RocketScientist -
- perhaps I was being too dramatic describing the difference as 'considerably more powerful' on AC - you are quite correct, there is no difference in gain between AC and battery power except at maximum volume, even after running on batteries alone for a couple of hours - the only noticable difference is that the LED glows fractionally brighter when the AC lead is connected. In any case, the O2 is a great device and more than powerful enough for me, and I think mine's working properly - thanks again for such an interesting project.
Hey RocketScientist, I dont know if this has been answered, but if youre building a desktop amp with optional DAC, does that mean there will be a standalone DAC option as well for when we are at home with our o2 portables? That would be ultimate for me, a DAC designed with the same goals as the o2, I just hope you dont choose TI dac's because thats all I've built so far and would really like to build an AK based dac, or something else....
just some splew.
just some splew.
@forsakenrider, the current DAC daughter board for the ODA can be used stand-alone from USB power with a line level (not headphone) output. It doesn't use the ancient TI PCM2xxx series USB chips as they don't support 24 bit operation over USB and have mediocre performance in some areas.
I believe the ODA DAC will match or outperform its two closest competitors--AMB and Twisted Pear--in most ways for substantially less money. Like the O2 it will offer maximum performance at the lowest possible price instead of focusing on the the latest FOTM or esoteric components. And unlike AMB, Twisted Pear, Schiit Audio, Matrix, Audinst, AudioGD, numerous eBay products, etc. the performance will be fully documented on professional equipment.
I believe the ODA DAC will match or outperform its two closest competitors--AMB and Twisted Pear--in most ways for substantially less money. Like the O2 it will offer maximum performance at the lowest possible price instead of focusing on the the latest FOTM or esoteric components. And unlike AMB, Twisted Pear, Schiit Audio, Matrix, Audinst, AudioGD, numerous eBay products, etc. the performance will be fully documented on professional equipment.