Yes, diff. outputs are the drawback(but I don't like output trafo idea, just to respect a max purity). That fact dramatically reduce chances for commercialization without big brand support, so I not asking any money for that - just hate when I do something neer to perfect and put that to the shelf. The only brand I recollect when asking myself who could be interested in that amp, it is HiFiman(planar's and headphones amps production). And I asked them a couple days ago about that concept - no reply.. Actually, I made it 100% for myself. I probably too looong time sit in front of my PC or laptop, and I've no space for any fancy phones amps at all. So I like to use the USB power and 24/192 stream for small enough amp, which could hang as a bit fatty cable. Please note, the USB part of this cable could be a long to keep the analog cable(amp's output to headphones) as short as possible! Ideally would be to put the amp inside of headphones cap, to get kind of high-res USB-headphones 
And finally, try to google for open-loop(with no feedback) amp with THD+N = 0.03% at -0.8db before clipping, do you believe you'll found something else? And how it is achieved, maybe with cutting-edge GaAs output stage? Nop, 64pcs LVC gates, that's it! I apologize if I may look like a braggart but using standard logic gates as a power stage is the spectacular idea
And finally, try to google for open-loop(with no feedback) amp with THD+N = 0.03% at -0.8db before clipping, do you believe you'll found something else? And how it is achieved, maybe with cutting-edge GaAs output stage? Nop, 64pcs LVC gates, that's it! I apologize if I may look like a braggart but using standard logic gates as a power stage is the spectacular idea
Last edited:
BTW, does exist one more drawback.. USB cable has to be from good one to gold plated one, probably. It is doesn't affect the sound quality but strongly affect the connect robustness. I have 3pcs USB cables and two of them losing connection if I shake the cable. Next, need to take off and plug again USB to get amp connected. I see no any difference between these cables at all but only one doing its job well. I've no idea what I can recommend potential users regarding the USB cable, probably gold plated one would be fine. I couldn't say if it is the serious drawback, rather no, however, it's quite annoying when you lay down on the sofa enjoying the feeling and suddenly gets silence. Probably I may consider another USB types of connector instead of B-type, which actually looks big and good for max current ability.
About schematics, I already shared USB high-res PCB design here CM6631 usb audio interface .... any good ? Power DAC schematic see attached:
Attachments
Deal! And I have a question about the output arrangement, is it good/bad idea to add there some 3.5 connectors yet? I found some jack but it seemed too thick, see attached. I can try to put jacks to the left and right walls as well if it's not too ugly. In another hand, fewer connectors = more sound purity, and we may simply solder outputs wires into the PCB without any jacks?
Attachments
badboygolf16v, do you have a 3D printer? The problem is I have no that instrument so far So I do not promise to send the printed case with PCBA(PCBs assembly), only the 3D model. As I said, I didn't found 3.5 jack suitable for 20mm toy's case hight(actually, even for 23mm 8mm jack's diameter is too big). In the same time, I 100% disagree increase the toy's size because it has to be extremely compact, ideally like a continuation of cable. I can replace a big film caps(yellow stacked type bricks, you can see my latest pic, they overlaying with jacks) with small SMD types, however, It will affect the sound, no doubt. Hence, again no. I've tried to put jacks to the left and right walls - no way, jacks are too large. If you can offer me some bright idea about jacks placement, you are welcome. Otherwise, let's forget about jacks. God bless a purity! I'll leave two 4mm holes for two cables(Left/Right), which should be soldered to PowerDAC PCB
PS: 2.5mm jack(the female one) is a too rare beast, I tell you.. I trying to find some stacked type film caps with less hight to fit 20mm case. Probably Nissei has something..
So I do not promise to send the printed case with PCBA(PCBs assembly), only the 3D model. As I said, I didn't found 3.5 jack suitable for 20mm toy's case hight(actually, even for 23mm 8mm jack's diameter is too big). In the same time, I 100% disagree increase the toy's size because it has to be extremely compact, ideally like a continuation of cable. I can replace a big film caps(yellow stacked type bricks, you can see my latest pic, they overlaying with jacks) with small SMD types, however, It will affect the sound, no doubt. Hence, again no. I've tried to put jacks to the left and right walls - no way, jacks are too large. If you can offer me some bright idea about jacks placement, you are welcome. Otherwise, let's forget about jacks. God bless a purity! I'll leave two 4mm holes for two cables(Left/Right), which should be soldered to PowerDAC PCB PS: 2.5mm jack(the female one) is a too rare beast, I tell you.. I trying to find some stacked type film caps with less hight to fit 20mm case. Probably Nissei has something..
Somebody asking me in Private_Messages about the BOM cost(parts+PCB1+PCB2, assembling is free), I didn't calculate it exactly but roughly it is $25-30 + delivering I guess $2-20. So I think the total should be within $50. And guys, I really not sure if my PowerDAC would be good for your DYNAMIC type headphones(*), hence, I could not recommend you spend your money and my time for that
* The PowerDAC is open-loop system(means no feedback at all), hence an output impedance not as low as typical headphones class A/AB amps has with tons of feedback. An input impedance of dynamic headphones isn't flat vs frequency, unlike planar one, so an open-loop output may sound tonally unneutral, colored(I can expect some bumpy, not well-controlled bass etc). Maybe not, but an ideal load is a planar headphone with differential(balanced, 4 wires) inputs, that's it.
* The PowerDAC is open-loop system(means no feedback at all), hence an output impedance not as low as typical headphones class A/AB amps has with tons of feedback. An input impedance of dynamic headphones isn't flat vs frequency, unlike planar one, so an open-loop output may sound tonally unneutral, colored(I can expect some bumpy, not well-controlled bass etc). Maybe not, but an ideal load is a planar headphone with differential(balanced, 4 wires) inputs, that's it.
I did a little math on what the output impedance might be, based on a guess at what LVC output buffer you're using. I went for Nexperia 74LVC2G34 dual buffer. Its worst case Zout at 4.5V is 22ohm in high state, 17ohm in low state. I'd expect typically it would be under half the worst case but let's go worst-case.
With 16 buffers in parallel that's about 1.4ohm per side of the H, so totally 2.8ohm. I have not included the impedance of the FBs nor that of the 10uH output inductors.
My conclusion is - I suspect in regards to output impedance there will be no problem driving dynamic over-the-ears headphones. IEMs with multiple armatures would be problematic but they're likely unsuitable for other reasons like noise.
<afterthought> Worst case Zout is (Zout(H)+Zout(L))/16 which is 2.4ohm, not 2.8ohm.
With 16 buffers in parallel that's about 1.4ohm per side of the H, so totally 2.8ohm. I have not included the impedance of the FBs nor that of the 10uH output inductors.
My conclusion is - I suspect in regards to output impedance there will be no problem driving dynamic over-the-ears headphones. IEMs with multiple armatures would be problematic but they're likely unsuitable for other reasons like noise.
<afterthought> Worst case Zout is (Zout(H)+Zout(L))/16 which is 2.4ohm, not 2.8ohm.
Last edited:
abraxalito, maybe you are right and dynamic type is ok too, but I saw HFiMan HE400i at taobao for 900RMB/$143, I guess it is cheapest planar on the market to try my PowerDAC BTW, I just ordered 20 PCBs(10 sets) which will be delivered next 4 days. News #2, 3.5mm jacks are no problem with 74x44x20mm case(from my latest pic). So far we are lucky!
BTW, I just ordered 20 PCBs(10 sets) which will be delivered next 4 days. News #2, 3.5mm jacks are no problem with 74x44x20mm case(from my latest pic). So far we are lucky!
Actually, my big project Tiny TAS5558 +TAS5624 "Power_DAC"+SMPS is quite similar to headphones PowerDAC, and I can implement all fancy things regarding IOS/Android control of internal DSP as well. It will not affect the size, only you'll get the ability to control 7pcs parametric EQs, loudness, treble/bass, and compressor which is a pretty interesting feature for open-back headphones placed into a noisy environment(or for vinyl-like compression simulation).
badboygolf16v, you mean you did measure SPL vs frequency, or you have a feeling of bump on 8kHz? To be honest, a planar headphone technically so simple thing(I believe HFman will be fine even if will offer that for $25), that I can't imagine if someone really tries to design it with some shaped frequency response because the membrane is so thin film that virtually has no resonants within audio range. Maybe a front chamber with auricle may form some nonlinear shape of the frequency response but I see nothing inside of HE400i caps to suspect that. What I personally feel when listening different planars vs dynamics, planars are sounds more neutral and lighter-effortless vs dynamic, and IMHO opened planars are very similar to each other's sounding headphones. Different sensitivity, different materials, and tactile feeling but the same "signature" to me.
PS: I just tested PowerDAC with dynamic Kicker headphones(crap one) and didn't notice any freq.resp. changes, THD@1kHz is also similar but higher at 100Hz..
PS: I just tested PowerDAC with dynamic Kicker headphones(crap one) and didn't notice any freq.resp. changes, THD@1kHz is also similar but higher at 100Hz..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.