1. Win a Broadcaster or one of 3 Teles! The annual Supporting Member Giveaway is on. To enter Click Here. To see all the prizes and full details Click Here. To view the thread about the giveaway Click Here.

Exploring Fender's Harmonic Tremolo

Discussion in 'Amp Tech Center' started by moosie, Dec 23, 2017.

  1. LudwigvonBirk

    LudwigvonBirk Tele-Holic

    Age:
    118
    Posts:
    933
    Joined:
    Aug 26, 2017
    Location:
    Madison
    How about subharmonics, couldn't those also enter in, and be heard?

    There's a lot of "outside the string frequencies" stuff going on when I plot my 1950 D28 and my 17" home brew archtop. Yes it is quieter but it is there.
     
  2. RLee77

    RLee77 Friend of Leo's

    Posts:
    4,314
    Joined:
    May 15, 2016
    Location:
    Nevada
    Sure, there’s a lot of low freq content below the low E 82 hz freq. Percussive effects, palm muting, string noise, etc.
     
  3. jhawk

    jhawk Tele-Meister

    Posts:
    111
    Joined:
    Feb 8, 2014
    Location:
    Logan, UT
    Yes, I found this alarming and it prompted me to question my assessment of the filters. In regards to the “indirect low-pass”, I figured maybe this was only a feature of the 4-triode version (to which the patent corresponds, I believe?). However the -A 5-triode version functions, it seems overly complicated.

    I think it would be of worth to create a streamlined version of the harmonic trem circuit! I’ve already hijacked this thread somewhat, and maybe further discussion on improving the circuit would be best as it’s own thread? Thoughts on this idea?
     
  4. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,593
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    I say go for it. Designing an improved harmonic trem seems to be firmly on topic, and in the spirit of this thread. Even though there's a lot of respect for Leo's work, this thread isn't meant to be a museum.

    When an actual build starts, then it needs it's own thread. Until then, with such a niche technical subject, it's nice to have a lot of relevant information in one place.


    That's my 2 cents, anyway. Others may feel differently.
     
    Snfoilhat, FenderLover, jhawk and 2 others like this.
  5. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,528
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    Here translated and paraphrased from German, an except from Manfred Zollner's excellent, very comprehensive, huge, and highly recommended book from 2009 Physik der Elektrogitarre
    Part of which can be viewed here...From which I hope he doesn't mind me presenting this small but fascinating bit. ( the chapter continues with an in depth look at the more complex Vox AC30 vibrato circuit )
    The 1960 Vibrasonic used a circuit combining vibrato - frequency modulation (FM), and tremolo - amplitude modulation (AM). The amplified guitar signal passes through a crossover network comprising of a low-pass and a high-pass split branch. Strictly speaking, it is a high pass and bandpass, but the bandpass has a centre frequency of 60 Hz, which is very low. upload_2018-4-23_22-35-53.png
    Both signal branches are amplitude modulated in opposing directions by the Low Frequency Oscillator (LFO), with the amplification alternating between the high and low pass.
    This alters the volume balance of the signals harmonic content, and also, to a lesser extent, produces a phase-shift which through inter-modulation results in pitch changes.
    The instantaneous angular frequency ω is the time differential of the phase angle φ. Because the first-order high pass, rotates the signal phase by up to 90°, and a first-order low-pass up to -90°, the swing from the high to the rotated Low-pass filtered Phase, leads to the Fender circuit typically producing a rotation maximum of about 120°.
    This can cause frequency modulations of up to approximately ± 10 Hz of frequency sweep, which would be quite noticeable pitch changes, as the threshold for perceptible FM changes is about ± 2 Hz.
    Figures show the magnitude and phase responses of the 5G13 Vibrasonic circuit.
    The 6G13-A eliminated the resistive Voltage divider in the high-pass branch, adding a gain of about 7dB.



    upload_2018-4-23_22-31-15.png
    Level and phase response of the crossover network.

    upload_2018-4-23_22-33-6.png

    Level and phase response of the overall system.



    .
     
    Last edited: Apr 23, 2018
    jhawk and RLee77 like this.
  6. LudwigvonBirk

    LudwigvonBirk Tele-Holic

    Age:
    118
    Posts:
    933
    Joined:
    Aug 26, 2017
    Location:
    Madison
    Who has peer-reviwed Manfred Zollner's work and do you trust it?

    (My dad is a 2nd gen prussian doctor from a lonnnng chain of scientists from there fwiw. I'm maybe a "black sheep" but am not an automatic skeptic wrt translated and paraphrased German science papers fwiw)
     
    Last edited: Apr 24, 2018
  7. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,528
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    I more than trust it! The translating I did myself, and the paraphrasing was to avoid translating the whole chapter.
    (Google translate turns the German "Endröhren" into "Tailpipe" it should be "Output Tubes")

    Professor of electronic accoustics, M.Zollner's book is 1290pages packed with reasearch gathered together by himself and his students over many years at the extreemly well equipped technical college in Regensburg...and it is well presented and written.
    His reputation is well respected by his peers, and the seminars he runs for hobbyists and enthusiasts are well reviewed. The book has good write-ups as far as I have seen.
    Although costly to buy, he has had the good will to set the whole of the tomb online for anyone to read....here it is gitec-knowledgebase/physik-der-elektrogitarre-buch/
    Even if you don't understand German, take a random look in it - Part 1 - Guitars, Part 2 - Amps, and you may want to start learning....es ist nicht so schwierig.

    :cry:The only critic one might have against him is the detailed measurements and hearing test analysis he has made to lead him to the ridiculous that the wood you use to build an electric guitar has no effect on the amplified sound:eek::twisted:
     
    Last edited: Apr 24, 2018
    SSL9000J, Snfoilhat and FenderLover like this.
  8. LudwigvonBirk

    LudwigvonBirk Tele-Holic

    Age:
    118
    Posts:
    933
    Joined:
    Aug 26, 2017
    Location:
    Madison
    Last edited: Apr 24, 2018
  9. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,528
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    Well spotted....I never bothered to look if there was an English part to the website, I've always just read the German, (which is much more extensive)

    I see the part I translated is also in English....the whole chapter on tremolo...so I will put a link direct to that chapter here....and I will check for errors in my translation:oops:

    https://gitec-forum.de/wp/wp-content/uploads/poteg-10-08-02-tremolo-vibrato.pdf
     
    Paul-T likes this.
  10. LudwigvonBirk

    LudwigvonBirk Tele-Holic

    Age:
    118
    Posts:
    933
    Joined:
    Aug 26, 2017
    Location:
    Madison
    Some of us are breathlessly awaiting the precise translation of sections 10.1 to 10.7.6 :)
     
  11. FenderLover

    FenderLover Poster Extraordinaire

    Posts:
    5,701
    Joined:
    Jun 11, 2009
    Location:
    Minnesota
    That's validating. Interesting how the low pass is again described as a band pass, and the description in the Fender patent is not really applicable to the operation of the actual circuit.

    Thanks for the links!
     
    Last edited: Apr 24, 2018
  12. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,528
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    Yes, as I say, it's good stuff, I've often dipped into it over the last 7 years. The only real critic I have come across for the book was that it spends so much time :)discussing and comparing the workings, sounds and:Ddistortion characteristics of 1960's and 70's amps from Marshall,:D:lol: Fender and :lol:Vox rather than looking at more modern amps.:lol::lol::lol:

    Shame,shame,shame.....
     
    Last edited: Apr 24, 2018
  13. jhawk

    jhawk Tele-Meister

    Posts:
    111
    Joined:
    Feb 8, 2014
    Location:
    Logan, UT
    A More Streamlined Harmonic Tremolo

    To get the ball rolling, I've started revising the harmonic trem circuit. I'm modifying the -A version of the circuit. Perhaps I'm wrong, but there seems to be a consensus that this 5-triode version is a little better than the 4-triode (albeit a little bloated), so this will be my starting point:
    harmonic tremolo.jpg


    As for the goals of this simplification endeavor, FenderLover has made a number of excellent points on what Leo could have changed:


    With this in mind, here is the aim of this redesign:

    1. Simplify the high and low pass filters, while minimizing the "notch" or "scoop" created by the filters

    2. Find a way to eliminate the cathode follower, and therefore a triode

    3. Replace the tube LFO and phase inverter with solid state devices (Keen's ideas will be put to use, no doubt! http://www.geofex.com/Article_Folders/sstremolo/sstremolo.htm)


    Here is the first revision, with simplified high and low pass filters. The frequency cutoff for both the high and low pass filters is ~320Hz. This roughly bisects the notes on a guitar, making the effect quite usable anywhere you play, and also minimizes the mid-scoop present in the old arrangement. (Component value changes in red, addition or deletion of components in green)
    harmonic tremolo r1.jpg

    I will do more revising as I have time to figure it out! Any feedback is very welcome.
     
    Last edited: Apr 24, 2018
  14. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,528
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    Looking at the original filter frequencies, it looks like they crossover at about 425Hz, and when both phases at at the same amplitude, the very soft and wide-banded mid scoop (4 octaves, so the complete guitar range) has only about a -3dB attenuation; compared to the tone stack that is in the amp, this is very linear.
    Surely the moving together of the HP & LP network frequencies as you plan will lead by comparison to a cumulative cross-over mid range boost. Would not changing the amps tone-stack to be more linear be a better option than first cutting, and then reboosting the mids?
    Or do you think the greater common share of frequencies in the two phases will have a positive effect to the intermodulation between them, and hence a greater FM effect?

    upload_2018-4-25_14-32-40.png
     
  15. jhawk

    jhawk Tele-Meister

    Posts:
    111
    Joined:
    Feb 8, 2014
    Location:
    Logan, UT
    When the effect is off, is that soft middle curve the frequency response we get? I had come to the conclusion that when the effect is off, we get a more severe mid-scoop.

    I wondered about the side effects that moving the filters together would have. I figured we might get a more pronounced phaser-like effect...which I assume to be the FM we are talking about.

    I suppose my goal here is to minimize the impact of the trem circuit when added to any amp. Since the effect is not bypassed when off, it will change the frequency response, but I guess I’m not absolutely sure how severe this change is.
     
    Last edited: Apr 25, 2018
  16. FenderLover

    FenderLover Poster Extraordinaire

    Posts:
    5,701
    Joined:
    Jun 11, 2009
    Location:
    Minnesota
    I was assuming that moving the bands closer together would have less effect when bypassed. Realizing there are only a few parts in the filter section, analysis might cause more paralysis than just trying it to see. Greater leaps can be made in the non-audio portion of the circuit, IMO. Here is a snip from my circuit musings - it is just an exercise at this point - UNVERIFIED:
    Vibe.png
     
    jhawk likes this.
  17. jhawk

    jhawk Tele-Meister

    Posts:
    111
    Joined:
    Feb 8, 2014
    Location:
    Logan, UT
    Could you help me understand a few of the changes you’ve made?

    Smaller mixing resistors - preserve high frequencies?
    Using an LED to bias the oscillating triode?
     
  18. FenderLover

    FenderLover Poster Extraordinaire

    Posts:
    5,701
    Joined:
    Jun 11, 2009
    Location:
    Minnesota
    The LED is a fairly common biasing technique that removes the need for a bypass cap. The mixing resistors are a guess that it wouldn't hurt to use a smaller value. Power is fed through an inductor from the first PS node, and is a good method to decouple any LFO noises from the audio section. Fixed bias cathodyne just because it's easy, large cap values just because we're dealing with 0.5-10Hz LFO.

    If you compare this to the -A version you may wonder, as I did, why so many parts were used. I feel there is enough here to function and shouldn't be difficult to optimize, even to the point of adding all the 'extra' parts back in.
     
    jhawk and Bendyha like this.
  19. SSL9000J

    SSL9000J Tele-Meister

    Age:
    49
    Posts:
    232
    Joined:
    Dec 18, 2016
    Location:
    Atlanta, GA
    They are no longer "tubes," or "valves." Henceforth I am calling them tailpipes. Because it sounds awesome.
     
    robrob and Bendyha like this.
  20. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,528
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    I posted earlier on in the thread, the articles writen by R.H.Dorf which seems to have been the inspiration for not only Fender, but more so Vox, in there designing Harmonic F.M. Vibrato's.
    The copy I posted was so ratty, that one could hardly read it, yet alone make out resistor and cap numbers.......so I rescanned the articles, removed the non-article bits, and sent them through a text bridge. So now the are a treat to read, and I think I can finally understand the workings of the circuit now:D.....still have to read-up on LFO's a bit though:oops:
    Basically the two articles are much the same...but there are some differences to the schematics.
    Interesting that he refers to the cathodyne/split-load PI as a long-tailed PI:confused:
    I have also added the magazine article on "Measuring phase angle shift in A.F" that Dorf refers to in the "Universal" article.
     

    Attached Files:

    FenderLover likes this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.