Gilmore Lite mk2 Measurements

Discussion in 'Headphone Amplifier Measurements' started by purr1n, Nov 29, 2019.

  1. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    Throwing these out here. A few observations:
    • At these levels, distortion is lower into 32-ohm load than 300-ohm load. -4dBu would be near the top of the range for RAD-0 or HD650 for typical listening levels.
    • There is of AC mains noise, which maybe audible with sensitive headphones.
    • Only one gain setting: 16db or x6.3. it's a bit high for most headphones.
    • AC mains harmonics keeps going. We can see the grass, the multiples of 60. Don't know this is an issue that will correlate to anything we hear.
    0dbu into 32-ohms
    0dbU 32.png

    -4dbu into 32ohms
    -4dbU 32.png

    0dbu into 300-ohms
    -0dbU 300.png

    -4dbu into 300-ohms
    -4dbU 300.png
     
    Last edited: Nov 29, 2019
    Joshvar, ultrabike, trl and 4 others like this.
  2. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    Here are THD+N vs dBu plots.
    • Keep in mind that dBu is input from the analyzer signal generator. For output dBu, don't forget to add 16db (the gain of then amp).
    • I'll let you guys do the dBu to Vrms conversions.
    • Bandwidth of the analyzer was set to 22kHz
    • GREEN = no high pass; RED = with 400Hz high pass (to filter out AC noise)
    Gl 32.png
     
    Last edited: Nov 29, 2019
  3. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    These were done with Modi 3 as the signal generator. Level was set to 1dbU in the THD+N "sweet spot" of the amp.
    GL MThigh.png
    GL MTlow.png
     
    Joshvar, Sqveak and Lyander like this.
  4. Vtory

    Vtory Illogical Spock

    Friend MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    3,983
    Dislikes Received:
    37
    Trophy Points:
    113
    Location:
    TX
    D2 = D3 for 32 ohm (for both output levels)
    D2 (increased) > D3 (reduced) for 300 ohm (roughly by 20+ db) -- seems an interesting behavior.

    Of course harmonics may vary depending on output levels, load Z, gain setting, etc etc. But when it comes to "relative ratio bw D2 and D3", 0 db to 20 db looks like quite a change. Most change occurs within 10 db or less as long as I remember.

    Did you hear (subjectively) that this amp favored a certain group of headphones noticeably?
     
  5. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    The sound is interesting. Tonality is warm, not cold or lean as what people would have expected from KG design. In reality, this varies depending upon build.

    Personally, whole overall tonality was good, I got fatigued from the highs. It's not has bad as GSX2 on high gain, more like medium gain.

    The spitty sssssss, sssss, sssss got to me in about a minute despite the warmth in the lows. I got a headache afterwards. This was with RAD-0 and Sony MA-900. Maybe with stock HD650 or LCD2r1 it might work better.

    Clarity, plankton retrieval, and involvement are very very good; but's what's the point if sssssss, sssss, sssss. Again, I'm sensitive to stuff like this. Note that there's nothing in this set of measurements that would point this out.
     
    ultrabike, Lyander, Sherm and 5 others like this.
  6. Skyline

    Skyline Double-blindly done with this hobby

    Friend
    Joined:
    Sep 28, 2015
    Likes Received:
    2,526
    Dislikes Received:
    2
    Trophy Points:
    113
    I got no hint of the sssss effect with stock 650s for whatever it's worth. Very smooth and clean in my experience with it.
     
    Sqveak, StageOne and rtaylor76 like this.
  7. rtaylor76

    rtaylor76 Can't wipe his tag

    Friend
    Joined:
    Nov 9, 2015
    Likes Received:
    1,737
    Dislikes Received:
    83
    Trophy Points:
    93
    Location:
    Nashville, TN
    Home Page:
    This amp is sensitive to power supplies. I remember testing with two types and the general SMPS sounded dull and just wonky. The other Mean Well power supply sounded better. I have a linear power supply in my DIY build. Linear Power supplies might are harder because it requires +15/-15 rails.

    That rising THD curve in the high end might look like the sibilance you are hearing.

    The Gilmore is not that great with hard to drive planars or ribbons. A much better match with 600/650s or maybe some Focals. Gain was actually almost too much with my Denon AH-D2000's. That part is very correct.
     
    Last edited: Nov 29, 2019
    Sqveak likes this.
  8. A Child of the Jago

    A Child of the Jago Acquaintance

    Joined:
    Mar 24, 2019
    Likes Received:
    62
    Dislikes Received:
    0
    Trophy Points:
    13
    Location:
    London
    I lean warm in my tastes and a sibilant treble would annoy the heck out of me but I've had the Gilmore for the better part of a year - first with Ether 2s (which I would mark down as a pairing with real synergy if you lean warm - that chain was Tidal -> Eitr -> Modi Multibit -> Gilmore -> Ether 2), and then most recently with ZMF Verite.

    I've just switched the Gilmore back on having had the ECP DSHA-3F on for the better part of a few months and am surprised at how enjoyable it is (even after an exclusive diet of DSHA-3F). I also don't hear any sibilance (and can't ever recall even thinking that), and with the Ether 2 especially felt that the highs were hugely impressive.

    Chain at the moment is Tidal -> Eitr -> Onyx -> Gilmore -> Verite
     
    Last edited: Nov 30, 2019
    rtaylor76 likes this.
  9. Stuff Jones

    Stuff Jones Friend

    Friend
    Joined:
    Feb 3, 2016
    Likes Received:
    1,179
    Dislikes Received:
    13
    Trophy Points:
    93
    I'm running Bitfrost 2 > GL Mk2 > HD6xx without a hint a sibilance. I don't have much to compare with with the 6XX but for me it strikes a good balance between detailed and resolving but also pleasant and full bodied. The Senns may be doing the heavy lifting on the latter characteristics.

    As for not working well with planers, Mr Speakers sells them so there must be at least some good pairings.
     
  10. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    It's not outright silbilance per se. Hard to describe. But the effects are similar. It's more like solid-state fatigue, something I haven't heard in a long time. Remember, I'm the guy who had difficulty with "glare" on the RME ADI AIO, which few others seemed to hear.

    SMPS sucks. LOL, I didn't realize the included box was an SMPS (just checked). I thought it was an LPS.
    Another consideration: the GL Lite mk2 might not be exactly a DIY Dynalo, there could be differences in design. As I've said, never had issues with DIY Dynalos or Dynahi.

    As far as rising distortion with more power output as seen in the plot, that's normal for discrete designs, and wouldn't be the cause of the fatigue. There's usually a sweet spot where distortion hits a low point, and then from there it goes up. Some chip designs tend to have wider sweet spots, and then things suddenly turn into shit.

    I also did a distortion vs frequency sweep. THD+N is stable across the frequencies. I didn't post this.
     
  11. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Friend MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    9,296
    Dislikes Received:
    30
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    I actually spent some time looking up if this came with an SMPS or LPSU because of the measurements. Everything seemed to indicate the stock PSU is a linear one...

    Is that not actually the case?
     
  12. rtaylor76

    rtaylor76 Can't wipe his tag

    Friend
    Joined:
    Nov 9, 2015
    Likes Received:
    1,737
    Dislikes Received:
    83
    Trophy Points:
    93
    Location:
    Nashville, TN
    Home Page:
    It IS in fact a cable wart SMPS. One that likely has filters, like some of the Mean Well ones. But let's also remember the Liquid Spark is also powered by a SMPS.

    There was talk of HeadAmp having some sort of upgraded power supply for the Lite MKII, but never happened.
     
  13. jowls

    jowls Never shitposts (please) - Friend

    Friend
    Joined:
    Oct 3, 2015
    Likes Received:
    3,985
    Dislikes Received:
    1
    Trophy Points:
    113
    Location:
    43°S
    According to Justin it’s still happening. Only after he has fulfilled orders for GSX Mini.
     
    Sqveak and rtaylor76 like this.
  14. jowls

    jowls Never shitposts (please) - Friend

    Friend
    Joined:
    Oct 3, 2015
    Likes Received:
    3,985
    Dislikes Received:
    1
    Trophy Points:
    113
    Location:
    43°S
    I found the top end of mk2 to be pretty non fatiguing. The supplied SMPS was changed from voltage specific to universal earlier this year. I wonder if that could have some part to play.
     
    Sqveak likes this.
  15. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    What DAC are you guys running?

    I think I may have figured it out. Switched the Modi 3 to a Modi Multibit. Also tried with a tweaked RAD-0 (stock has some minor peakage) and Verum 1. No more issues. Very good, but a bit dull in terms of dynamics and slam.

    I can't help but feel the SMPS is limiting the GL mk 2 and maybe unnaturally sharpening the transients a bit. Lot of potential that might be missed.

    More later...
     
  16. Skyline

    Skyline Double-blindly done with this hobby

    Friend
    Joined:
    Sep 28, 2015
    Likes Received:
    2,526
    Dislikes Received:
    2
    Trophy Points:
    113
    I was running BF1 (Multibit) at the time.
     
  17. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,602
    Dislikes Received:
    16
    Trophy Points:
    113
    Location:
    Irvine CA
    @purr1n, do you know what is the output impedance?

    I believe a relatively high output impedance can result in higher THD+N for 300 ohms vs 32 ohms due to the voltage divider between the amp load and the test load (if the amplifier is operating in the linear region at 32 ohms).

    This happens with the 2i2 gen 2 (10 ohm output impedance).

    A higher Vin will be needed to drive the 32 ohm load vs the 300 ohm load for the same Vout, because more power will be dissipated across the amplifier's output impedance in the 32 ohm load case vs the 300 ohm load case.

    This means the amp will push more signal power for 32 ohms vs 300 ohms, for the same Vout across the test load, and therefore more SINAD (unless bordering the non-linear region). How much depends on the amps output impedance.
     
    Last edited: Dec 2, 2019
  18. Stuff Jones

    Stuff Jones Friend

    Friend
    Joined:
    Feb 3, 2016
    Likes Received:
    1,179
    Dislikes Received:
    13
    Trophy Points:
    93
    BF2. My only comparison was GOV2A which sounded brighter and thinner.
     
  19. rtaylor76

    rtaylor76 Can't wipe his tag

    Friend
    Joined:
    Nov 9, 2015
    Likes Received:
    1,737
    Dislikes Received:
    83
    Trophy Points:
    93
    Location:
    Nashville, TN
    Home Page:
    Tried it with a Bifrost Multibit 1 and with the SMSL Sanskrit 10th, which also uses a AKM4490 like Modi 3. The SMSL was close to the BF, but had just a hint of mid-grain.

    Also, I will say it does have good lows, it did bring out some upper end with HD650's, but in a good way - lifting the veil if you will. Also remember, the amps and driver are bipolar transistors (BJT), although the input is all FET. Fet always has a slightly sweeter and sharpness, but to me is balanced out by the bipolar amp itself. Like I said before, I compared it to a good Hafler or Bryston speaker amp.

    At some point, I am going to be working on separating my LPS and amp board into two boxes and using the same 5-pin DIN connector for the Lite MKII. Just not enough time during the day/week/month/year/life.
     
  20. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    50,968
    Dislikes Received:
    84
    Trophy Points:
    113
    Location:
    Antarctica
    I believe it's less than 1-ohm, but I have not measured. Either way, the distortion is so low as to not matter. As for IEMs, I don't think this amp was ever intended for IEMs with its gain.

    OK, this makes better sense now and why I've had more success with MM as a DAC and headphones with smoother response in the highs. I'll give the RDAC a try too. The amp may sound tipped up a bit in the highs, so it will be a matter of synergy.

    I was going to dig up an LPS and try, but with the 5-pin, it might be too much of a pain in the ass (I lack time) to bother. Do you know the pinouts? Or maybe @Justin can chime in? These discrete designs usually benefit enormously from LPS over cheapo SMPS. I actually wonder if the AC mains grass in the measurements (purely academic - this stuff theoretically according to science should not be audible) is a result of the SMPS.

    @Justin: Do you have an LPS that you could ship to us so we can try?
     

Share This Page