Massdrop Cavalli Tube Hybrid amp technical measurements

Discussion in 'Headphone Amplifier Measurements' started by atomicbob, Mar 8, 2018.

  1. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    There are some amplifiers that have 0.1% 2nd harmonic distortion that people find sound quite good. There are some amps where even 0.0001% 4th and 5th harmonic distortion are intolerable.

    Regarding tubes, there can be quite a variance from the same manufacturer and same tube model. Ask someone with a tube tester about this.
     
  2. GanGreinke

    GanGreinke Friend

    Pyrate Contributor
    Joined:
    Mar 29, 2016
    Likes Received:
    512
    Trophy Points:
    93
    Location:
    New York, NY
    @Junki, in case you aren't aware, @atomicbob has had problems in the past where other people have used his measurements to try and make arguments that aren't actually supported by his measurements. These people didn't take the time to try and understand what the measurements did and did not represent, how the measurements were done, and what factors could have caused variability in the measurements (i.e. different measurement scales, differences in tubes, even those from the same manufacturer, etc.). Instead of trying to learn this information, they decided to jump to conclusions (usually calling a product crap, calling the manufacturers liars, calling people who like the products shills or blind sheep, etc.). Whether you did so intentionally or not, you fit into this profile.

    I get it, there is a lot of misinformation and snake oil in the audio industry. There is a lot of overpriced stuff, and manufacturers like to make a lot of claims that might be laughable. But you'll also miss out on a lot of great gear if you try and make conclusions based on an incomplete understanding of the data. Don't be afraid to ask if you don't understand something; at worst, you'll be called a noob and given a link to a reputable source by someone with a better understanding of it.
     
  3. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Thank-you @GanGreinke . I was reacting as if I was yet again in conversation with another of a particular net personality's underlings. However I still get irritated when people won't do the homework requested. Next post will illuminate.
     
  4. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet

    Aspersions have been cast on my mcth measurements for potentially corrupt or unethical behavior due to disparities depicted in the mcth measurements. It is clear that the person making such statements neither understands how the measurements are derived nor what information they convey, which would have been evident if the requested homework stated at the beginning of the technical measurements post had been performed. Let me reiterate that homework assignment and general statement:

    If you are unfamiliar with audio measurements please use a search engine with the query:
    "audio measurements" or "audio measurement handbook"
    Look for publications by Richard C. Cabot and also by Bob Metzler, both from Audio Precision. There are other useful publications as well. These will provide basic knowledge.


    Interpretation of the following measurements is beyond the scope of this post.

    So a request for interpretation was made by the individual being discussed. Having noted the lack of understanding I chose to suggest the following course of action hoping it might prove enlightening:

    To focus on a single attribute without context of the others will mislead you. Can you really hear 78 dB below main signal? To test yourself perform a listening experiment with the wavefile found here. You should also read the first several posts in that thread carefully. You should then seek the references provided at the beginning of the first post in this thread. Those will help add perspective.

    What was the point of my response?

    1. Note that a single measurement out of context is misleading as will be demonstrated below
    2. Ask oneself how this particular measurement fits within actual audible experience
    3. Read about measurements and how they may be manipulated
    4. Re-request the individual to perform the homework necessary to understand measurement derivation and interpretation

    Those who have followed my technical measurement series will have read my repeated loathing of single number specifications which are derived from complex measurements. The following discussion will demonstrate why I am constantly harping on this topic with a comparison of numbers to graphs.


    20180307 mcth A04 THD+N  THD  nth-HD 30R - annotated.png
    CTA = Continuous Time-domain Analyzer measurements
    FFT = Fast Fourier Transform measurements

    This is the numbers only chart. It is a highly condensed set of numbers to represent complex frequency dependent behaviors.

    An observation was made that THD+N % (CTA readings outlined with a blue box) is much larger than THD % (FFT detector readings outlined with a red box) in these measurements for mcth. The disparity appears approximately a factor of 23 to 24x. How can this be? The answer is available in the graphs that were posted and by applying knowledge gained from doing the homework requested at the beginning of the technical measurements post (in italics above.) In particular, Bob Metzler while at Audio Precision wrote “Audio Measurement Handbook”. How a THD+N measurement is obtained may be found on page 30. Further, how a THD measurement is obtained may be found on page 29. The next several graphs will provide a visual tutorial aiding the understanding of what Bob Metzler wrote.


    20180307 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF.png
    Here is a screenshot of THD+N, THD and FFT measurement for the mcth. First thing to note is the numbers do not match exactly those of the previous screenshot. The number of harmonics included in the THD calculation differs. The numbers only chart calculates with 2nd through 10th. The graphic above is calculating only 2nd through 5th harmonics.

    Ever wonder why the manufacturers specifications are usually stated as THD+N < x.xxx % ? Measurements vary between acquisitions, sometimes by 6 to 12 dB or more. This may be caused by several sources of which the following are examples:

    1. Frequency at which the measurement is performed.
    2. Output level at which the measurement is performed.
    3. How the data is acquired at any point in time, which is beyond the scope of this discussion.
    eg: FFT sample trigger, window function, etc.
    4. If tubes are involved, microphonics and temperature variance may influence low level data in the acquisition

    In the graphs left channel data is colored blue, right channel data is in red. Let’s focus on the left channel for the remainder of this discussion.


    20180307 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF - stimulus notched.png
    THD+N is a relatively easy measurement to acquire. A stimulus is provided to the device under test at a known level and frequency. In this case 0 dBu (0.775 Vrms) and 1 KHz. The stimulus frequency is removed from the DUT output in the analyzer with a notch filter. What signal remains is THD+N for the DUT. In the graph above, 1KHz has been removed as seen by the black space where the 0 dBu, 1 KHz component was in the FFT. The rest of the blue signal is used for the THD+N calculation. THD+N measurements came about as easy to implement in days of analog only capability.


    20180307 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF - dominant THD+N mains noise.png
    Now concentrate on the area outlined in Red on the left channel. These are the components that dominate the THD+N measurement. They are AC power supply mains related, fundamental and harmonics above. This accounts for the high number observed in the mcth THD+N% measurement. It is likely that the tube in my mcth had a higher than average residual hum contribution. Subsequent to this measurement the tube developed some static noise in the left channel and I replaced it with a new one which has no static.


    20180307 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF- THD pct annotation.png
    THD is a more difficult measurement to perform in the analog domain. With capable computers and FFT analysis techniques THD has become a common measurement. In this measurement frequency domain windows are created to acquire only the nth harmonic bins from the FFT. Above we see all signal eliminated except 2nd, 3rd, 4th, 5th and 6th harmonics in the red outline box. THD shown in this measurement only incorporates 2nd through 5th harmonics. In the first, numerical only screenshot above THD was calculated using 2nd through 10th and has a higher value than this one using only 2nd through 5th harmonics.

    The answer to the THD+N vs THD disparity was available to the reader in the THD+N, THD nth harmonic + FFT graph all along.

    From the preceding we observe that THD is indeed low but THD+N is power supply mains dominated in the mcth. The likely culprits are SMPS and any particular tube sensitivity to mains noise. Tubes vary in their sensitivity to noise, even among tubes from the same manufacturer and model. This led me to explore replacing the SMPS with an LPS. The results are shown again below.


    20180307-04 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF - LPS.png
    Substituting an LPS for the SMPS a factor of 5x reduction occurs in THD+N while THD is approximately the same, considering measurement variance discussed above and the original tube used for these tests. Before anyone asks, no I haven’t retested with the replacement 6922. It may require several hours to setup and verify proper test conditions. The actual data acquisition takes only minutes. Then there is the time to document and validate data acquired. Finding a four hour time slot is not as easy as it sounds in a busy lab.

    Other notes.
    1. Massdrop didn’t specify THD+N. Only THD. THD+N will always be at least as large a value as THD, but usually larger due to the incorporation of all noise.
    2. My tests are at 0 dBu (0.775 Vrms) while their tests are performed at 2.21 dBu (1.0 Vrms). There will be some variance in the results between test run at the two differing levels.
     
  5. Ringingears

    Ringingears Honorary BFF

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    3,659
    Trophy Points:
    113
    Location:
    Northern Californium Valley
    Thanks @atomicbob. Personally I don’t feel you needed to reply to any of these accusations on your character. You already do so much for this community, and your measurements and reading suggestions have helped me to improve my understanding of measurements and what they tell us and what they don’t. It irritates me that you had to use more of your valuable time to respond to this. Again much thanks for all you do.
     
  6. Junki

    Junki Facebook Friend

    Joined:
    Aug 13, 2017
    Likes Received:
    102
    Trophy Points:
    33
    Location:
    United States
    First of all, I want to thank the moderators of SBAF for, at least at this point of time, not censoring my posts from this thread as @atomicbob had requested. They serve as the best source of supporting evidence for the rest of this post as well as to just how truthful statements like the following are.
    (Btw, you can say my name, rather than "this person.")

    I'd like to bring the parts of this thread concerning the issue I raised back into focus.
    The discrepancy I raised was never been about THD or THD's relationship with THD+N%. My original post:
    The discrepancy I raised was about the measured noise level of @atomicbob's stock MCTH. It's clear from the above that my bringing up the difference between THD and THD+N was merely to underscore that THD isn't a problem, which I also explicitly stated was great. The problem I brought up was specifically noise and its discrepancy with what is published. This becomes even more clear when I move on to SNR, which has nothing to do with THD or an active signal. So to paraphrase my original concern and to include some unstated transitions from metric to metric that I seemed to wrongly assume would be obvious:
    • The noise level measured is really high.
    • (Unstated in my head) I'm well aware Massdrop doesn't publish THD+N, which is unfortunate because I want the N part, which is the part that would tell me the ballpark noise level this device is supposed to perform at, but they do publish SNR. I also know how noise and SNR are simply related both conceptually and mathematically, so let's look at that instead.
    • (Back to typing) The measured noise level results in a measured SNR of -77.9dB. The advertised SNR is -97 dB. Could you, with your amount of experience with equipment and their measurements, shed some light as to why this is? I'd very much appreciate it because I can't and this discrepancy is preventing me from pulling the trigger on this thing even though I want to because people whose opinions I respect subjectively like it.
    There has been a handful of posts since I originally brought up the concern, the most informative being @atomicbob 's most recent. One common theme in all of them is condescension towards me. The other is that the stock MCTH's noise issues are due to PSU (which I thought I made clear that I was aware of in my first post in this thread) and tube variance (the only fresh signal, or the only new piece of info, for me in all of this).

    What hasn't been addressed or asked about in anyone's post besides my own, even after I brought it up in my very first post, is this:

    If we need to use a LPS with the MCTH to even get SNR into the -90 dB range, is it okay that Massdrop advertises that the SNR for this amp, which I remind everyone comes with a specific SMPS power supply that @atomicbob measured with, at -97 dB? When, as purchased, it measured at -78 dB SNR?

    This is the issue. The fact is, no one is talking about this! To this very post! Why? If it's because everyone here already knows the reason and that reason is acceptable, why not simply tell me?

    About the inconsistent assertions of my unethical or corrupt allegations. I don't think anyone is corrupt, especially not @CEE TEE or Massdrop. When I brought up the issue originally, I thought it'd be cleared up with a simple answer. Something like, "My tube was defective" or "my tube connection was defective" or "I don't know but I'll look into it when I get around to it." A show of willingness to acknowledge the discrepancy between what was measured and what is advertised at least, even if there was a good reason for it.

    Instead, I get responses from many people that completely ignore this discrepancy, and precisely this is what feels unethical to me. Yes, your LPS measurements demonstrate that power is a contributing cause to the measured noise, but the entire endeavor and explanations of it are investigative. By investigative, I mean it's only useful for informing hypotheses like "maybe the SMPS power supply I got was defective, or maybe the AC power thingies in the board was defective."
     
    Last edited: Mar 31, 2018
  7. k4rstar

    k4rstar Britney fan club president

    Pyrate BWC
    Joined:
    Jun 11, 2016
    Likes Received:
    6,941
    Trophy Points:
    113
    Location:
    Toronto, Canada
    Yes. I tried to find a list of other users who were as upset as you about this issue, and this is what I came up with:

    [​IMG]

    If you care about a ~20dB discrepancy in SNR you should purchase a better measuring amp that likely sounds worse. At least then you can stop shitting up this thread by pretending to be very smart.
     
  8. Ringingears

    Ringingears Honorary BFF

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    3,659
    Trophy Points:
    113
    Location:
    Northern Californium Valley
    @Junki. Your fitting in well here. Read more. If you are a noob to measurements, learn more before you post. You seem to want to win an argument more than anything or you would let it drop. Which I suggest you do. This is a waste of time.

    I am one of the last people who would say this on this forum, but you really don’t get this place yet it seems. Measurement discrepancies are one of the first things to learn about measurements. Massdrop might actually use his measurements to improve the unit. Knowing CEE-TEE personally, that’s what he will do. I won’t post anything more in this thread.
     
    Last edited: Mar 31, 2018
  9. msommers

    msommers High on Epipens

    Pyrate
    Joined:
    Nov 6, 2016
    Likes Received:
    2,750
    Trophy Points:
    113
    Location:
    Calgary, Alberta
    Home Page:
    Thank you for all your hard work, Bob. I’d like to emphasize the FREE WORK you do for all of us despite your crazy schedule.
     
  10. Junki

    Junki Facebook Friend

    Joined:
    Aug 13, 2017
    Likes Received:
    102
    Trophy Points:
    33
    Location:
    United States
    That's reassuring. Thanks for trying to be helpful and talking about my original concern.
     
  11. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    @Junki

    1. Most likely, you did NOT read Bob Metzler's "Audio Measurement Handbook", or not as thoroughly as atomicbob wanted. Metzler discussed some concerns. You mentioned or cited NONE of them.

    2. You had a strong belief that "measurements don't lie". I had that belief until very recent (I was also advised to do 1. when I raised my question to atomicbob. haha). Read 1. and read others' measurements and try to learn from their comments. Mathematical definition of each term is the easiest part. Imo measuring is on whole different level.

    3. While I can be easily convinced between -77.9db and -97db ("advertised") of SNRs, it's ok that you may pursue more explanations (mostly obtainable from 1. and 2. though). Even if so, why do you write your concerns HERE (in very ignorant shapes)? Of course first question may be OK. Atomicbob gave you some answers (something to read) which you DID NOT read. He is NOT in charge of customer services in Massdrop. Email to Massdrop to ask those if needed.

    4. Do you perhaps think -10db at -97dbFS is as equivalently definite as -10db at 0dbFS? You like to convert scales from log to linear or oppositely. Please do that and think again. Hope you have a good sense of randomness and errors.

    5. Different companies take various positions about specifications. Conservatively (with worst conditions) - very rare, or optimistically (with ideal conditions) - quite prevalent (I guess). Different people measure differently, different purposes may choose among measured results differently, and so forth. Even laptop battery time measurements (one of VERY MAJOR MARKETS) vary a lot across different reviews. We are living in the REAL world. Don't be too naive.

    Except your first post in this thread, every other reply you made looks like NOISE, worsening signal-to-noise of this thread. Feels very ironic.
     
  12. Ringingears

    Ringingears Honorary BFF

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    3,659
    Trophy Points:
    113
    Location:
    Northern Californium Valley
    @Junki. Not sure got it. You might want to quote the entire post. Do hope this can stop now. Apologies to all for not keeping out this.
     
  13. maverickronin

    maverickronin Friend

    Pyrate
    Joined:
    Mar 17, 2018
    Likes Received:
    313
    Trophy Points:
    63
    Location:
    Midwest, USA
    I bet it's a typo in the specs on Massdrop. 97bB SNR is probably what the MCTH will do at full output, not 1Vrms as listed on the site.

    If I'm reading AtomicBob right, he measured 78dB SNR into a 30 ohm load at 0 dBu output on the MCTH.

    The rest of Massdrop's specs say it will do 1W in 50 ohms which is ~7Vrms which is ~19dBu

    Increase the signal by 19dB, noise stays pretty much the same, SNR is now 97dB.

    Not really sure why no one seems to care though. Regardless of the rest of the sound, noise floor will limit the sensitivity of what headphones you can use without unacceptable hiss.
     
  14. Elnrik

    Elnrik Super Friendly

    Pyrate
    Joined:
    Jan 10, 2017
    Likes Received:
    8,973
    Trophy Points:
    113
    Location:
    Denver CO
    Home Page:
    So, I listened to this amp and it sounds great. Maybe we should do more of that.

    (You know, instead of arguing over measurements we don't understand and raising atomicbob's blood pressure.)
     
  15. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    Ok, let's try this again.

    Look at your assertion that the noise measurement is two orders of magnitude apart. Since you kept referencing noise and a factor of 10^2. I thought you must be referring the THD+N (which is a noise related measurement) and THD. They differ by a factor of approximately 25 and the only difference coming close to your factor of 100. Now I see you don't understand how SNR is measured for audio devices.

    SNR is a voltage measurement in audio, not a power measurement. As such your assertion is in error. The difference between the massdrop spec and my measurement is onlly approximately one order of magnitude. It is actually less. More on that later.

    Here is the calcuation for SNR in audio:
    Audio SNR calculation.png

    All audio analyzers make the measurement this way. It is not calculated as a power using 10 * log (Ps/Pn) as seen on wikipedia.

    Here is some suggested reading on the topic:
    Texas Instruments Audio Power Amplifier guidelines - see page 13 equation 5
    http://www.ti.com/lit/an/sloa068/sloa068.pdf

    Intersil application note - see equation 1
    https://www.intersil.com/content/dam/Intersil/documents/an97/an9789.pdf

    Analog Devices training seminar - see equation 3
    http://www.analog.com/media/en/training-seminars/tutorials/MT-003.pdf

    Rane Note 145 - Audio Specifications
    http://www.rane.com/note145.html


    First point, I left your name out as the post was copied directly from part of a paper I have prepared for lecture at local AES section and in corporate training. I didn't feel it appropriate to include your screen name in such public lectures.

    Second point - why the difference in SNR measurements:
    Ok, answer still remains look at the spectrum, not the numerical measurement for enlightenment.

    20180307 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF - stimulus notched.png

    SNR uses a very similar FFT spectrum measurement with the additional constraints that 0 dBr (relative) has been set to 0dBu, then signal dropped -60dB relative to assure a quality measurement. In analysis a notch filter is again applied to the amplifier output as shown above, bandwidth set for noise voltage measurement. In my measurement noise bandwidth is set to 22 Hz to 22 KHz as stated in the SNR measurement notes. Now the 20 log (Vs/Vn) is calculated.

    Look at the power supply contribution once again which will dominate the SNR noise measurement:
    20180307 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF - dominant THD+N mains noise.png
    This suggests the power supply and tube combination is contributing to the residual noise.

    As stated above, later the tube supplied with the mcth developed audible static noise in the left channel. Prior to that my listening tests using HD800, HD650 let me hear only the faintest mains noise which became totally inaudible with the LPS. Please note my listening environment is below NC-15. This is an acoustic lab. Normal environments have much higher ambient sound levels and what I could faintly hear will be totally masked by normal listening spaces. So I didn't get excited about this discrepancy in SNR given the mains noise as major source of the higher measurement.

    Let's look at some measurement differences.

    Massdrop measured at 1.0Vrms and I measured at 0.775Vrms. My measurement was 2.21 dB lower. So it is likely that raising my measurement level would have changed my result of -78 dB to a little below -80 dB. The absolute residual noise won't change. My bandwidth filter was 2 KHz wider than that used for the Massdrop measurement. That could account for another 1 ~ 2 dB. Now we may have lowered my reading to -82 dB. So the difference is now narrowed potentially to 15 dB, a factor of 5.6x. this is far below your initial assertion of a 100x disparity of which you became so agitated as to make some pretty strong statements inferring impropriety. Fact of the matter, you didn't understand the measurement methods and results calculation. the suggested reading would help enhance your understanding for the future.

    What could cause a 15 dB or factor of 5.6x difference? The sensitivity of a given tube to noise. Certainly reducing the power mains noise caused such comparable drop. However I have yet to retest the mcth with the replacement tube and original SMPS for reasons stated previously.

    For further reading about variance in tube characteristics these might provide a good start:

    Tube Characteristics - page 2 - bogie tube facts - What then is a Bogie tube today
    http://www.alltubetesters.com/articles/bogey_tube_facts.pdf

    Tube Heater Hum
    http://www.valvewizard.co.uk/heater.html
     
  16. Junki

    Junki Facebook Friend

    Joined:
    Aug 13, 2017
    Likes Received:
    102
    Trophy Points:
    33
    Location:
    United States
  17. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    MCTH measurement update

    MCTH measurements presented on 20180307 were acquired with a tube that subsequently developed static on the left channel. Since then I have replaced the supplied EH 6922 with an EH 6922 from Tube Depot selected for low microphonics and gain matching between the two triodes. I don't have time today to post updates for all of the data that is affected, but here are the most important ones to consider. 300 ohm behaves substantially similar to the 30 ohm presented below.

    As can be seen, SNR is even lower than the Massdrop specification of -97 dB. THD+N is lowered substantially as this tube does not exhibit the power supply noise sensitivity. Gain linearity improves below -80 dB. However the residual noise, which now measures low with the SMPS, still improves further with the LPS.

    30 ohm load update 20180402

    MCTH signal to noise (A07) 30R load
    01 20180402 mcth A07 SNR 30R.png

    MCTH THD+N THD nth harmonic distortion (A04) 30R load
    02 20180402 mcth A04 THD+N  THD  nth-HD 30R.png

    MCTH THD+N THD nth harmonic distortion (A04) FFT 0dBu 30R load
    03 20180402 mcth A04 THD+N  THD  nth-HD FFT 30R 4XF.png

    MCTH THD THD+N vs Freq 0dBu 30R
    04 20180402 mcth THD THD+N vs Freq 30R 4XF.png

    MCTH Gain vs Amplitude Left 30R load (G2_L1T3_1_A)
    05 G2_L1T3_1_B.png

    MCTH Gain vs Amplitude Right 30R load (G2_L1T3_1_B)
    06 G2_L1T3_1_A.png

    MCTH residual noise spectrum Left and Right A=0dB 30R load
    07 20180402 mcth residual noise A=0dB 30R 4XF.png

    MCTH residual noise spectrum Left and Right A=0dB 30R load with Acopian A28MT210M
    08 20180402 mcth residual noise A=0dB 30R 4XF - LPS.png

    So there you have it. Less than expected SNR and THD+N differing considerably from THD caused by a slightly marginal tube with a specific deficiency. It happens. This is life in the Tube Lane.
     
  18. Junki

    Junki Facebook Friend

    Joined:
    Aug 13, 2017
    Likes Received:
    102
    Trophy Points:
    33
    Location:
    United States
    Nice. Look at that linearity. Buying one.

    Thanks for taking the time to investigate and re-measure @atomicbob
     
  19. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,653
    Trophy Points:
    113
    Location:
    On planet
    MCTH connector and output polarity.

    MCTH 4-pin XLR and TRS connectors are wired in parallel as verified with DMM. 4-pin XLR pins 2 and 4 are tied together to shield of TRS connector.

    Output polarity is inverted from input on both channels (picotech 5243b oscillosope).

    Input is 20 Hz sine 2Vpp.
    Trace colors
    Blue Input
    Red Output

    MCTH left channel with 300R load

    20180601-01 mcth polarity L ch 300R 4XF.png

    MCTH right channel with 300R load
    20180601-02 mcth polarity R ch 300R 4XF.png
     
    • Like Like x 5
    • Epic Epic x 1
    • List

Share This Page