DIY talk

Discussion in 'DIY' started by Cspirou, Jul 25, 2021.

  1. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    Man, I'm really curious how this would fare with low impedance/low sensitivity planars, especially compared to the NuHybrid.

    I remember for that design pete said somewhere the NuHybrid was capable of delivering 1.5WPC into 32ohms (I think in the headfi thread, my numbers may be a bit off) while for the sangaku all I can find is a nebulous "1.4W".

    I DIY'd a garage1217 Project Sunrise III years ago that I still use with my HE-4, but recently I've become more curious about other hybrid designs that would play well with planars, especially since I'm planning on picking up some DCA Aeon RTs that really push the low-impedance envelope. The sunrise has a power output graph available on their site and from what I can tell it would only be capable of pushing 500mW into the impedance of the aeons which seems pretty light.
     
  2. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    Found this nice interview with Joseph Crowe. (Actual name is Troy. Joseph is his middle name). He goes into his design choices and projects he's done



    Especially interesting to me is when he mentions horns in pro audio vs home audio. In live sound reinforcement they want to maximize coverage and an even spread along the entire frequency range. The geometry needed for maximum coverage can lead to diffraction and reflections back into the horn, adding distortion. In the home environment low distortion is far more important than even dispersion so a smooth exponential horn is favored in that case. You wouldn't want to use pro horn at homes and

    Primary reason we don't see horns for home audio is because a good horn is big, hard to design and expensive. You can get extremely good results now with a simple box and inefficient drivers with lots of watts. Back when horns were a bit more in fashion, amplifiers were very low wattage so everything was done to boost sensitivity. This meant big drivers, big cabinets and horn loading whenever possible.
     
  3. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    the output uses OPA551PA opamps and from the datasheet it looks like you would get 2W into 50Ω. Current is limited to 200mA whether you are balanced or not.

    If you have something like the LCD-3 at 110Ω then you have 4W available.
     
  4. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    TIL about pro vs home audio horns!

    Hmm, so it looks like in terms of power there wouldn't be much difference between the nuhybrid and the sangaku, since the nuhybrid uses the same opamps in its output. I wonder if there's much reason for me to bother with the sangaku then, esp considering I'm not planning on using the balanced in/out anyway.

    200mA looks like it should be enough for my applications, from this chart it looks like I'd only need more current if I plan on causing myself hearing loss.

    Are there any other DIY high current hybrid tube designs recommended for low impedance/sensitivity planars? I wonder if I should maybe be looking into millet's older designs or derivatives of them, or if there's another DIYer I should look into I haven't even considered yet. I'm still in the very very opening stages of this build so I'm not looking to make any firm decisions just yet, mainly just wanting to see what's out there. It's been so long since I've DIY'd anything that I'm still just trying to get my bearings again, lol.
     
  5. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    Pretty sure you could get 100 mA of genuine class A goodness out of the older Millett hybrid MOSFET versions. But I'd be surprised if you could find boards, and have no idea what sort of tube stocks are still available. If anyone can help you out, it would be @TomB.

    Here is the starting page, and the setup and biasing suggests output stage class A bias currents up to 125 mA.
     
    Last edited: Sep 10, 2021
  6. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,469
    Trophy Points:
    113
    Location:
    Winnipeg
    There's also the proximity effect. Once you get out to a reasonable distance, a lot of those distortions kinda smooth out. It's a bit like thinking you're doing a "controlled" turbulence effect in the beginning in order to get smooth dispersion later on because everything is evenly mixed, vs "laminar" flow early on which will not give you even mixture down the line (obviously this is a gross misuse of the terms turbulence and laminar).

    Some pro horns can be used well in a home theatre setting, but you still want to be a couple body lengths distance away from the speakers. A friend of mine built a floating sound proof theatre in his basement to watch movies with some horns in there, though in honest truth it's probably just where he and his wife go to get away from the kids... huh, guess horns are involved either way
     
  7. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    Thanks! I'd stumbled on those pages diving deeper into pete's site and was about to ask about whether those designs are still good. Looks like beezar still has the PCB available, I'll have to try to pick one up if their stock starts to get low. As far as tubes I'm sure I should be able to find tubes somewhere, looks like beezar does still have the 12FK6 in stock which seems to be the most relevant for me anyway considering the lower gain tubes are encouraged for lower impedance headphones on the tube page of the milletmax site. I'll have to keep monitoring the stock of both and pounce when they're close to out of stock I guess. If nothing else, I may end up having to buy a few single tubes and figure out how to try to match them myself (or ship them to someone else that can)

    EDIT: lol, just bothered to check tomb's profile and realized they're beezar :D
     
    Last edited: Sep 10, 2021
  8. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    Yep, @TomB is Beezar! :D

    I reckon the Millett Max would hold up pretty well for a nice DIY project. If nothing else, they are a fun build, with plenty of cool customisation options. I built several over my early years, all BJT versions. I did end up selling them all, because my M3 was always a slightly better and more versatile amp.
     
  9. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    Heh, well fun is about half the reason for DIYing at all! I haven't tried a ton of SS DIY designs (I think the pimetav2 was the most complex one I ever bothered with) but the M3 has always caught my eye as one to maybe try at some point. I've always tended to stay away since the HE-4 really needs the extra warmth of a hybrid, but I have thought that I should probably have at least one good higher-powered SS amp around for when I start exploring other planar options that may benefit more from that kind of transparency.

    The RJM Sapphire also caught my eye as a decent SS design, although it looks like even the high-current version wouldn't have very much headroom. To be honest though, from what I've read about the DCA Aeon RT (my current planned next planar, conditional on audition) I doubt I'd find a hybrid design too warm with those cans. Maybe I should build a good SS amp to have around in case I find another set of cans that would benefit from it though.

    I am curious what you liked better about the M3, and which opamps you decided to go with there. I'd always read the OPAs were supposed to sound warm (and the pimeta I built used the AD opamps so I definitely saw the more "harsh" side there) so I'm curious what your thoughts were on that amp. If it is as warm as some of the writeups act like I may need to add an m3 to my shortlist of DIY projects too.
     
  10. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    Pretty standard build with AD8610 at the time. It's been 10+ years, but I remember that the M3 was just more precise and detailed than the Millett hybrids. Some people interpret that as harsh, but the M3 also has a reputation as being on the warmer and more lush side of the SS spectrum. But given the nature of DIY and OPAMP rolling, it is pretty hard to make sweeping conclusions.

    I've got ADA4627B in there now, and I didn't do particularly convincing comparisons, but they seem somehow more musical and more detailed than the 8610.

    Oh, and the M3 has a much lower noise floor. Like, vanishingly low. I attribute that to the S11 power supply and custom magnetically-shielded transformer I use (and obviously, not tubes!).
     
    Last edited: Sep 10, 2021
  11. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    Not exactly. opa551 on the nuhybrid has a lower voltage and there's clipping at much lower volume than the Sangaku.

    Also depends if you had enough voltage to reach that current to begin with. In order to reach 2W at 50Ω the opa551 needs +/-14V. on the nuhybrid the voltage is +/-12V.
     
  12. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    Too bad Glass Jar Audio decided to close. They had some Cavalli hybrid PCBs. Can probably still make your own if you find the gerber files somewhere
     
  13. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    Hmm, interesting. Given the AD opamps seem to have a reputation as the more "analytical" sounding anyway, that makes me think I'd be pretty into the amp with the burr-browns instead.

    Makes sense, although I've gotta be honest, the fact that I wouldn't have to bother building a $100+ PSU is kinda an attractive feature of the mosfet max, hehe.

    It's funny, I remember when I first found out about the M3 years ago I'd wanted to build one, but got somewhat scared away by the anti-virtual ground craze at the time. Then after I built the pimeta I kinda figured SS amps would always be a bit too harsh for my taste. But now talking with you it sounds like it'd be exactly the type of SS amp I should give a try. It's gonna be quite awhile before I start buying parts and such again, but this is definitely another design I'll need to keep on the shortlist.

    I do see that the M3's "bass boost" feature can be repurposed as an adjustible gain. I wonder how versatile that kind of setup would be with both the low impedance/sensitivity planars I mentioned as well as higher impedance cans like HD650s, or if I'd just be better off trying to optimize the design for one or the other. That's a much more minor concern though, given when I do finally get a pair of 6xxs I'll inevitably end up building a Crack to use them with anyway.

    Ahh gotcha. My inexperience strikes again (I didn't even consider the difference in implementation). It's sounding more and more like the nuhybrid might be interesting to mess around with later but the mosfet max would probably be a better overall fit for its price/performance ratio. My curiosity has definitely been piqued about the mosfet diamond buffer of the max, although I'll need to do more research to see what differences that topology would make compared to a circuit with an opamp output.

    Thanks for everyone answering all my questions so far, I know my inexperience with circuit design and such is definitely showing through these discussions, but I really appreciate everyone's patience. Not really sure which resources I should try to look into to better familiarize myself with some of this stuff. Maybe I should look back over the tangentsoft site again, I seem to remember that site explaining a fair bit about amplifier circuits.
     
  14. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    I have a Crack as well. The Crack is the better amp with the HD650. The M3 is a better amp with everything else.

    The virtual ground is a whole heap of plusses and minuses, that are mostly academic. I think there is great value in preventing ground pollution, and not needing the ground to source and sink all current through a headphone. The virtual ground solves this, but so does going balanced. The virtual ground adds additional distortion and increases the amp's output impedance, but so does going balanced. Balanced obviously has infinitely better stereo crosstalk than a virtual ground, but in a world where people use crossfeed, I'm not too worried.

    I don't think you'd design an amp today with a virtual ground, you'd just go balanced. At the time, going balanced was really expensive, because not many sources or headphones were set up for it, so it made sense historically.

    Yep, I originally had mine with variable gain from 2-12 (voltage gain, that is). It was never above the minimum setting. It is set to a fixed gain of 3x voltage now, and I find I get very little movement on the volume pot even with a cool 1.4VRMS source.

    So it really depend what volume you listen at. My highest comfortable listening level with the HD650 corresponds to about 0.25VRMS. An absolute piece of cake even at unity gain. Even if you think/calculate you might need 10 times that, we're still at 2.5VRMS, which is achievable at a gain of 2 and maximum on the volume pot. A gain of 3, and you are approaching clipping from a 2VRMS source. TLDR, most headphone amps have WAY too much gain, even for high impedance headphones.

    Oh, and also consider the entire amp's global negative feedback loop goes through the bass boost (or gain) pot. If you are going to do variable gain, a switch and fixed resistors is a better strategy.
     
    Last edited: Sep 10, 2021
  15. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    That was kinda my thinking too. I'm thinking whenever I do get around to getting the 6XXs it may be a bit before I pull the trigger on the crack+speedball purchase, so it'd be nice to have another cheaper amp to use in the meantime that could then be used exclusively with my planars once I do build the crack.


    Wow, that's a really succinct explanation. While I always knew nwavguy was kinda full of shit, I never fully understood why. It helps too that I've been reading a lot on balanced connections since I was considering building a Sangaku, so that historic context really helps a ton in understanding. Thanks for that brief writeup.

    Thanks for that info, it sounds like a gain switch would probably be ideal in my case too. I'm not sure exactly what gain I'd need for the planars I'd intend on using with it. It looks like according to this chart the HE-4 only needs a bit higher voltage for 90dB but far more current, which if I'm understanding right the gain doesn't really affect anyway (I could absolutely be wrong on that point though).

    Damn, if I'm understanding all this right, the M3 might just be the SS all-rounder I've been looking for to complement my sunrise III for now. Seems like it'd be more than capable of handling my planars and could get the job done with my HD6xx until I splurge on the crack+speedball.
     
  16. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    A word about power availability.

    Even if you won't go near the rated power it is often better to have a higher power amp than a lower powered one. Why? Because it will be more linear in the usable region.

    A major reason why sensitive speakers sound better is because you aren't stressing the amp to the point of distortion. Often remaining in the Class A region of a Class A/AB amp

    So yes, using 200mA on HE-4 would make you go deaf but that doesn't mean it's unwise to have that much power on tap
     
  17. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    If you haven't seen the DIY thread I started here, you should take a look. Basically, it's my angst at trying to rationalise my DIY gear versus the reality that the DIY community is dying, and commercial gear is amazing.

    I haven't posted about the M3 yet, I will in the next couple of months. But as a preview, the M3 is a great DIY amp, and I still reckon there are few better. But when you can buy a Schiit Jotunheim for not much more than the cost of some nice casework and panels alone? It's hard to argue that the M3 makes sense. Not that I am trying to discourage you, but DIY just isn't the no brainer it used to be.
     
  18. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    Absolutely, that's what I was trying to say in that sentence (as in, this has way more power than I'd ever need anyway), apologies for not being clearer there.

    I'd glanced at that thread previously but I'll take a deeper dive into it. I feel like the hackability and customization of DIY will never get old for me, since I'll be able to tweak amp designs to really optimize them for the cans I intend to use them with, but it is nice at least that the market has been shifting to allow great performance at lower costs. A big reason I got into DIY at all was that the equipment I could build for $200-300 would rival stuff I could only buy at the $700+ price point back in 2013-2014 or so, so back then it was a no-brainer to build stuff yourself.

    I really need to try out a jotunheim. Hell, the thing has some damn impressive power output specs, and at $400 you're right that I'll be hard pressed to finish a mosfetmax or m3 build for a significant savings over that cost. I even see that one of the vanity photos has a DCA Aeon in it. I guess my brain is just still having trouble getting used to the thought of buying any retail gear at all :D
     
  19. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    Yeah, that's exactly the problem I was having. Looking at new products and upgrades after a long break, and realising it would be way cheaper to buy something, but not wanting to.

    I think with the M3, the recasting and upgrades I implemented still make it a compelling 'product'. And the DIY DAC I am building with a Soekris dam1021 module makes great sense versus other options. But this might be my last hurrah before submitting. I just don't have the talent to design anything myself, nor the time to do deep dives into the sort of projects that people are producing.
     
  20. 3l3tric

    3l3tric New

    Joined:
    Sep 8, 2021
    Likes Received:
    3
    Trophy Points:
    3
    Location:
    Texas
    It's funny you mention the soekris, I've had my eye on doing a build with one of the DAM boards as well (since my current "best" dac is still a gamma2). I'm tempted to build one of the phidacs as well since it seems like a more vintage-style R2R design to complement the soekris, but beyond those two I don't think I've really seen many other DIY DAC projects worth messing with.

    Beyond those dacs and the small handful of amps I've talked about, I'm probably pretty much towards the end of my DIYing too, although my last hurrah will most likely end up being a full speaker build with an a10 preamp and NuClassD poweramp or something similar. Seems like speaker equipment hasn't really had the budget retail renaissance that headphone equipment has.

    Then again, speaking of headphone equipment, are there any decent retail alternatives to DIY Hybrid designs? Seems like both full tube amps and SS designs have some really compelling retail options, and maybe I just haven't looked anywhere near hard enough, but it doesn't really seem like there are any retail alternatives to the garage1217 amps or millet hybrid designs at the price point you can still build those.
     

Share This Page