The Objective2 (O2) Headphone Amp DIY Project

The 1N400x diodes are rated at 30 A of peak current for a half wave cycle (8.3 mS) and don't even get lukewarm in this application under worst case conditions. I don't think it's a problem unless I'm missing something? The diodes are running at 1.9/30 or 6% of their rated current.

30A non-repetitive peak. I just found my mistake - forgot to add series resistance to the voltage source. That drops it down below 1A peak.
 
Last edited:
I have a boatload of windows open in safari which does tend to bog it down memory wise, but all the same there does seem to be something a miss.

just went to download it and couldnt (see below), looks like theyve got some honorable defending to do
 

Attachments

  • Picture 519.jpg
    Picture 519.jpg
    108.3 KB · Views: 4,355
30A non-repetitive peak. I just found my mistake - forgot to add series resistance to the voltage source. That drops it down below 1A continuous.

It's the duty cycle that ultimately matters. The "non-repetitive" issue, AFAIK, is mostly related to thermal issues. And if the diode is staying very close to ambient what matters are what sorts of other harm peak currents can cause. If the 1N400x can handle peaks of 30A without such harm, brief repetitive spikes of 1.9A should be a walk in the park.

But, it's good to know it's theoretically below 1A :). I never attempted a simulation on the power supply. The only simulation I ran for the entire O2 was was the power management circuit. I've mostly not had good luck with simulation representing the real world due to all the things it normally doesn't account for. And when I try to crank in all the parasitics etc., almost without fail, it fails to converge at all. But I'm sure others are better at it than I am. I'm a math and hands on guy mostly.
 
Diodes are so cheap I always spec them so that repetitive peaks stay under the rating. Probably over-enginering, but it has never done me wrong. :)

I've actually tried spice-ing the comparator circuit too, but no luck. I'm sure its on my end though. I've subbed in a LT comparator and I don't have models for the mosfets. Time permitting I may try adding the appropriate models in if I can find them.

A few months ago I was pondering a dual-battery protection circuits and was thinking about comparators on the rails. Never did design anything. But I more or less decided that any such circuit would be best as "supervisory", powered by a separate supply with a separate battery. Seemed that collapsing rails during a fault condition would produce random results with the protection circuit unless powered by something else.

But it looks like you've used comparators good down to 2V, and the LED reference is probably around 2V - haven't looked it up yet - so that likely long before the rail caps have discharged in a fault condition the protection circuit would have kicked in. I'm assuming you probably have designed in some hysteresis so the protection circuit doesn't bounce around the trip point. Haven't read through your ckt description yet.

Also during my past ponderings it seemed like a latching protection circuit would be a good idea, something that alerted the user a fault had occurred by a LED and required manual (power cycle) action to reset. Given that if the protection circuit ever operated then by definition it needs user attention..

bat A low
bat B low
bat A and B low
bat contact A disconnect
bat contact B disconnnect
both batt contacts disconnect
 
Last edited:
@agdr: I've tested the power management circuit every way I know how and, it took a few tweaks, but it works very well. It shuts down around 13- 14 volts (6.5-7 volts per battery) so there's no issue of the circuit itself not having enough power. I have a bipolar bench supply and have fed the O2 even conditions of extreme battery imbalance and it shuts down flawlessly every time. The trick is minimizing any transients at the headphone output due to asymmetrical rail conditions and I think I've done that. Yes it has hysteresis.

Please let me know if you have further comments after you read through the Circuit Description at the end of the article? It's always good to have another set of eyes or two checking for errors and if anything needs correcting I'll fix it ASAP.
 
Last edited:
I just made it back to the computer. :) Well congratulations on the protection circuit. Sounds like a very good design. A good protection circuit is really the special sauce to making a dual battery design safe and reliable.

I'm most useful with power supplies and other things close to DC. :) Hopefully someone here will eventually run a group buy of O2 boards.
 
I just made it back to the computer. :) Well congratulations on the protection circuit. Sounds like a very good design. A good protection circuit is really the special sauce to making a dual battery design safe and reliable.

I'm most useful with power supplies and other things close to DC. :) Hopefully someone here will eventually run a group buy of O2 boards.

Thanks. I agree about dual battery designs.

There are already some places interested in selling the bare PC board for a very reasonable price (hopefully < $10). So a group buy isn't required. But I'll be publishing the gerber/drill files so if a group buy works out cheaper, that's fine too.

Because the enclosure isn't available from Mouser, that would be a great item to include in any group buy to cut down on the total shipping costs. The same is true for the metal volume knob from DigiKey and perhaps the batteries purchased in a large lot from All-Battery (eBay) or elsewhere. If a group buy included the options of a PC board, enclosure, volume knob, batteries and perhaps even the FPE front panel, that might substantially broaden the appeal? Everything else is available from Mouser (as is a plastic volume knob).

The latest revision prototype boards are due to arrive next week. Once I build one up and torture it with the dScope, etc, I'll know if I can release the board design into the wild--including the vendors interested in offering them.

Also, as a general FYI to everyone, I've substantially revised the first two articles. If you haven't looked at them in a while there's quite a bit of new info. I've also tried to trim some of the "fat" and redundant content. For example I'd left the output stage details out of the Design Process article and have added quite a bit there. And I've also added new measurements to the first O2 Headphone Amp article.
 
yeah i ended up just using ltc2935 for battery supervisor hanging off one of the 4 x 3v3 cells (nominal 3v3, 3v6->3v range) the one with the largest current draw and sensing its own supply; at 3.15v, as the PFO pin is pulled low, an led on the front panel lights for low battery via an xor gate that sniffs the pin and opens the gate, which flows current via a transistor with beta 10; given just enough ic to produce 10ma with RLED chosen to make sure 10ma flows over the led, at 3v the RST pin is pulled, which shuts down all the regulators into a low current SHDN state and an LED (via an identical mechanism but with different RLED and threshold to cope with a dropping VCC) and lights to say 'charge me now'

the led remains lit until I switch the relays, disconnecting the battery so i can connect to the external high current battery balancer/rapid charger (its a 2300mah 14v LiFePO4 battery). the LEDs are indicators only, not directly powered by the monitor and the regulators are effectively off in shutdown. the relays are normally closed types, so even if i miss all the signals the relay coil will eventually drain the battery to the point where it can nolonger stay open, so i'm never in a position where there is no charge to ready it for charging.

the battery is still a pretty safe voltage when the RST pin is pulled, but after 3v it starts to fall more steeply. this is a balanced, DC coupled DAC/Amp unit, with each channel of the amp section powered by a bipolar LDO regulator from LT and a differential opamp, so i have little to worry about wrt offset, the load will just center itself around the middle of the rails and will never swing close enough to the rails for me to have anything to worry about, plus even if it did it would be a common mode error and thus deleted at the driver; plus if there is some sort of disconnect on one phase, the signal wouldnt be complete, because the headphones dont know what ground is

all these regs are insanely efficient, never needing more than the dropout to survive, but not what you would call cheap.

yeah Safari sux, used to be much better, i run both chrome and firefox too, but i had safari open today as it has a plugin i needed and didnt get around to doing anything else. that being said, that big honking error message has nothing to do with the browser.
 
@qusp, there certainly are ways to do Li-xxx batteries relatively safely. But even LiFePO4 has issues. I had a defective LiFe prism cell puff up and explode before it was ever used, charged, or connected to anything. It didn't catch on fire, it just built up pressure internally from a manufacturing defect and made a mess. I did some searching online and it turns out that's not uncommon. Many have blamed it on abusing the battery but it happens even with brand new unused batteries. Apparently if the wrong impurities get into the the wrong places when they're made they turn into ticking time bombs.

And I'm not aware of any HPA suitable through hole charger controller/power management IC for a bipolar Li-xxx supply. They're all surface mount.

What's really weird to me is FiiO essentially cloned the AMB Mini3 with their new E11 amp. Except instead of a 9 volt battery they used LiPoly and a DC-DC converter. The could have very easily made it a bipolar DC-DC converter as in the E9 and the amp would have 0 volt referenced inputs and outputs--i.e. a real ground. Instead, however, they used a single DC-DC converter and the same inefficient expensive OPA690 virtual ground op amp that seriously degrades the Mini3's performance. Why?
 
@khaos974, You mean Google Docs is censored across the board in China? I don't know where they ship from, but Mouser has a Chinese branch: Mouser China

Google Blogger has restrictions on maximum image size, but I've made some changes and if you click on the schematic, board drawing, and/or BOM images in the article you'll get much larger versions. They're fuzzy but legible. The PDF from Google Docs is much better but if that's not working for some of you, I've tried to provide another option.
 
Last edited:
I just got the power management circuit to simulate! The LED was drawn backwards on the schematic PDF I downloaded yesterday. Just took a look at the Google site and see you found/fixed that in the latest drawing.

Here is a failure voltage drop spice sim on the negative rail with 8.4V constant on the positive rail. Negative rail goes from -8.4V to -3.2V at 3 seconds in. The converse parameters with the drop moved to the positive rail with the negative rail held constant produce similar results.

I've subbed in an LT 1716 for the comparators from the LT spice model base. It is rail to rail, over the top, and good from 2.7 to 44v. Yours is better suited with the 2V min supply but this does the job for sim. IRFP240/IRFP9240 subbed in for the mosfets - so in real life they would gate-source overvoltage short when operating. :D

The legend is on the top. Blue -vbatt is the negative rail drop at 3 seconds in. Orange +vout is the positive rail shutting down, while light magenta -vout is the remainder of the negative rail shutting down.The green trace for the output of comparator 1 is kind of hard to see, but the other blue trace is the output of comparator 2 toggling from high to low. You can see the bright magenta voltage divider sample voltage crossing over the grey LED reference voltage at about 3.05 seconds, with the comparator(s) toggle happening just a hair after it.
 

Attachments

  • Comparator Neg Pulse Circuit.png
    Comparator Neg Pulse Circuit.png
    46.9 KB · Views: 2,559
  • Comparator Neg Pulse Sim Plot.png
    Comparator Neg Pulse Sim Plot.png
    19.4 KB · Views: 2,499
Last edited:
@agdr Thanks for the sim! If I'm reading your graph correctly, you show about a 60 mS lag between when each crosses 3 volts? I don't think that's true in real life but I'd have to double check. It turns out that's the critical parameter. Below that is where the audio op amps can become unstable. As long as each rail is above |3| volts the entire audio circuit doesn't care how unbalanced the rails are.
 
You are right about the lag, but I've subbed so many parts in here just to fit the LT Spice model base that the real thing will likely be different.

Pretty impressive design, though! It does work for a fault in either rail. I haven't tried a combined rail fault, but just logically since its all between the rails that is just a change in point of reference. Should also work just fine.
 
Last edited: