StellarE,
Funny how I noticed the same jack on your amp. No, this is what Status Audio uses on OB-1 phones. There is a spring strain relief and of course a screw-on 3.5mm adapter (not currently used)
@Prasi - Thanks for the kind words. I may have a layout job for you coming up
Funny how I noticed the same jack on your amp. No, this is what Status Audio uses on OB-1 phones. There is a spring strain relief and of course a screw-on 3.5mm adapter (not currently used)
@Prasi - Thanks for the kind words. I may have a layout job for you coming up
Last edited:
StellarE - how big are your rail caps and are they FR's too?
Back in Black (and Brown)
Here is my current all electrolytically cap coupled daily listener with my new 4-driver IEM's. The combo is very intoxicating - deep deep bass and superb highs. My verdict on the FR's is that they are excellent and do not appear to need any help from a film cap bypass. I would say that the treble is very clear - almost too sharp for some. The 1000uF 16v OSCON plus 470uF Nichicon AK and 1uF Wima is still very good sounding to my ears - perhaps a little more laid back presentation but with more bass punch due to larger value of nearly 500uF more when using 16ohm phones. Albeit, these are 1o6dB sensitive so doesn't take much power.
I am using 2200uF 35v 105C rated (no-name Chongx) caps for rails, 1000uF 25v FR for output, 10uF 35v Silmic II for input (brown caps). I took it apart for the photo of the insides for y'all.
Here is my current all electrolytically cap coupled daily listener with my new 4-driver IEM's. The combo is very intoxicating - deep deep bass and superb highs. My verdict on the FR's is that they are excellent and do not appear to need any help from a film cap bypass. I would say that the treble is very clear - almost too sharp for some. The 1000uF 16v OSCON plus 470uF Nichicon AK and 1uF Wima is still very good sounding to my ears - perhaps a little more laid back presentation but with more bass punch due to larger value of nearly 500uF more when using 16ohm phones. Albeit, these are 1o6dB sensitive so doesn't take much power.
I am using 2200uF 35v 105C rated (no-name Chongx) caps for rails, 1000uF 25v FR for output, 10uF 35v Silmic II for input (brown caps). I took it apart for the photo of the insides for y'all.
Attachments
Glad that you hear it too, X! What value of R2 are you able to get away with when using those Silmic input caps?
Twocents, Post 889 lists all the parts and tweaks I used in my last build, except the output caps, which are the exact ones X is using. My rail caps are 12.5x21mm--absolute biggest that can fit in the tin unless you omit the battery wire harness between them.
Another heads up to folks wanting to use higher values at R4. My second build runs very hot! I think I have to heatsink it...
It helps immensely to add a layer of Kapton tape to the inside of the tin bottom and lid to enhance heat transfer. The bare metal has an emissivity of 20% whereas with the Kapton, it is about 90%. So you can get almost 5x better radiative transfer of the infrared heat coming off the PCB heatsink tabs to the case. I have run at 68mA bias and it is warm but not overly so.
I currently use 22k which should only require 4.7uF, but since 10uF is same size, why not?
Thanks X. I guess I needn't worry too much if you've been running with even higher bias than me. I put black electrical tape on the inside of my first build's tin, but I didn't realize its effect was so profound. I skipped the tape this time because I was eager to take the new amp on the road. 
In open air I had no problem but in a closed, un-taped tin, the MOSFETs were almost too hot to touch after 30 min...could only hold my finger on them for less than a second. Now where did I put that roll of electrical tape...?
I love the way Funch'a build looks with those tiny sinks on there. May try them too, but I suppose they're not much use without tape to help the heat escape the tin.
In open air I had no problem but in a closed, un-taped tin, the MOSFETs were almost too hot to touch after 30 min...could only hold my finger on them for less than a second. Now where did I put that roll of electrical tape...?I love the way Funch'a build looks with those tiny sinks on there. May try them too, but I suppose they're not much use without tape to help the heat escape the tin.
An externally hosted image should be here but it was not working when we last tested it.
Those should work. I wonder what Funch used to attach them? I have been using white silicone RTV and seems to work.
To insert in-line images, click on the yellow landscape (mountain skyline with sun) icon, paste your image url, and that's it.
Fascinating, X, do you think this is superior to the Silicon Harmony?
The specs are exceptionally low, of course, but there is something else here that gives it the sound, maybe the spatial characteristics, and the high levels of H2 even though it's a quasi, PP output stage.
Thank you for all the very careful, meticulous measuring..... I had been long convinced about your findings, but never actually measured it like you have here. This is proof that harmonic issues and reduced global fb are actually pivotal to the sound quality.
Hugh
The specs are exceptionally low, of course, but there is something else here that gives it the sound, maybe the spatial characteristics, and the high levels of H2 even though it's a quasi, PP output stage.
Thank you for all the very careful, meticulous measuring..... I had been long convinced about your findings, but never actually measured it like you have here. This is proof that harmonic issues and reduced global fb are actually pivotal to the sound quality.
Hugh
Hi Hugh,
Indeed I do think that the sound is perhaps among the best I have heard. It is very similar to Sikicon Harmony, but packs a greater punch. I do believe that zero or very low global feedback is key to great sound though. But a harmonic distortion profile like this playing at 14v peak to peak into 50ohm cans is an amazing thing to hear, if your ears scan take it. These are some ear-bleed SPL levels on 96dB sensitive cans.
Indeed I do think that the sound is perhaps among the best I have heard. It is very similar to Sikicon Harmony, but packs a greater punch. I do believe that zero or very low global feedback is key to great sound though. But a harmonic distortion profile like this playing at 14v peak to peak into 50ohm cans is an amazing thing to hear, if your ears scan take it. These are some ear-bleed SPL levels on 96dB sensitive cans.
Honestly XRK971audio goes too fast for me if there is going to be new version every 2 or 3 month where conclusion is the newest always seems greatest and as normal in diy sector schematics are not released. To keep diy customer as me interested here in Silicon Harmony plus HyQu and upcoming releases then at least please document versions with pro performance measurements for for a varity of load impedances so we can see some of what we buy into and what to expect. For example Audio Precision APx 525 measurement service is available here https://www.neurochrome.com/performance-measurements/.
PCA is not included into above speculation because X was encouraged to release kit from users and schematics is openly shared, hope there is some understanding for point of view in a thread where every one seems to agree that even changing one capacitor or change layout a bit can or will change subjective sound experience (real world improvement or just bit different).
XRK971audio goes too fast for me if there is going to be new version every 2 or 3 month where conclusion is the newest always seems greatest and as normal in diy sector schematics are not released. To keep diy customer as me interested here in Silicon Harmony plus HyQu and upcoming releases then at least please document versions with pro performance measurements for for a varity of load impedances so we can see some of what we buy into and what to expect. For example Audio Precision APx 525 measurement service is available here https://www.neurochrome.com/performance-measurements/.PCA is not included into above speculation because X was encouraged to release kit from users and schematics is openly shared, hope there is some understanding for point of view in a thread where every one seems to agree that even changing one capacitor or change layout a bit can or will change subjective sound experience (real world improvement or just bit different).
Hi X,
i am a little confused running the amp at such high dissipation. For example the Beyer Dynamic with 250 Ohm impedance claim maximum input of less than 100 mW, that translates to 20 mA and 5V. The Sennheizer HD598 indicates 112dB/1V rms and impedance is 50 ohm, translating in 1V at 20 mA.
Typical SE average dissipation is about 4 x required power output, in this case <100 mA The voltage swing of 5V rms is enough to drive 250 ohm headphones.
Total dissipation is only 0.5watt. I cannot see the need for running the amp so hot, all you are doing is wasting energy promoting 9V battery sales.
i am a little confused running the amp at such high dissipation. For example the Beyer Dynamic with 250 Ohm impedance claim maximum input of less than 100 mW, that translates to 20 mA and 5V. The Sennheizer HD598 indicates 112dB/1V rms and impedance is 50 ohm, translating in 1V at 20 mA.
Typical SE average dissipation is about 4 x required power output, in this case <100 mA The voltage swing of 5V rms is enough to drive 250 ohm headphones.
Total dissipation is only 0.5watt. I cannot see the need for running the amp so hot, all you are doing is wasting energy promoting 9V battery sales.
Honestly
PCA is not included into above speculation because X was encouraged to release kit from users and schematics is openly shared, hope there is some understanding for point of view in a thread where every one seems to agree that even changing one capacitor or change layout a bit can or will change subjective sound experience (real world improvement or just bit different).
Hi Byrtt,
I took data for 51ohms and 270ohms at low, moderate, and high drive. Were you looking for a different impedance like 16ohm? I can tell you that if it plays this clean with 51ohms and 270ohms, anything in between is good. Why would I want to have tomchr to run an AP measurement on this amp for me? My amps are only -66dB to -85dB THD with a limit of -126dB noise floor - within capability of consumer $100 audio interfaces and RMAA or REW. I don't have a need to measure down to a -160dB noise floor or -145dB THD.
New amps come out and tend to sound or perform better because we get smarter as designers. Although the difference in sound quality between HyQu and SH is very slight - HyQu can just get louder.
I agree if you have 250ohm headphones that are 96dB to 110dB sensitive, all you need is 60mA bias. I could back HyQu down to 60mA bias and 18v supply rail and it would still be a great amp since it can drive a higher swing than a SE Class A amp.
Here is the problem child headphone that I have been designing my amp for:
Considered by some to be the best sounding headphone in the world. It's only 83.5dB sensitive and requires a 2w amp.
Headphones & portable audio - HIFIMAN.com
Don't know if its because of businees but miss objective point of view to develop and improve tech, also amp electrics live their electric bandwidth way out of 100$ audio interface and can't believe as designers not interested to see how and learn from those pro data.
Maybe its just me but if i want amps that subjective is claimed sound more or less better or is it just different there is plenty of builds to choose from in other sector of diyA, and as costumer to a commercial product one of the worst thing to happen is few months after spending ones money there is released new version claimed to be better.
When you say it tend to sound or perform better because we get smarter as designers, well smarter can mean many things when we into audio stuff, remember CHN70 was subjective smarter but objective it was worse distorting unit that cheated the human ear, backing up products with pro objective data will help show honestly designer share products real performance even also when data is worse than other products but maybe then sound better for some other reason.
I don't think you need a $40k AudioPrecision to get 98% of the info that matters. As the great work that you, Wesayso, and others have shown in other threads. A world class (best in my opinion) speaker system was built and measured by Wesayso with a $100 mic and REW.
Of course measurements are important but the fact that an amp is zero global feedback and running pure SE Class A enforces a certain sound harmonic profile and directness that subjectively will beat a Class AB amp with feedback in a blind test every time. I will set up an AB switcher again to compare. But I never said the HQ was *better* than SH. They are close but serve different goals. If one needs multiple watts at 50ohms - SE Class A is very hard to achieve.
Of course measurements are important but the fact that an amp is zero global feedback and running pure SE Class A enforces a certain sound harmonic profile and directness that subjectively will beat a Class AB amp with feedback in a blind test every time. I will set up an AB switcher again to compare. But I never said the HQ was *better* than SH. They are close but serve different goals. If one needs multiple watts at 50ohms - SE Class A is very hard to achieve.