Filament buss question

Discussion in 'Shock Brother's DIY Amps' started by moosie, Nov 23, 2017.

  1. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Check at the junction of r39 and r41, r41 and r40, the non-ground side of r33.

    Upstream of r34 should not be less than downstream, and the grid of V5b should be much more than 2mV signal.

    Discharge the caps, and check these points for DC resistance:

    Check resistance from V4 pin 2 to ground. It should read about 2M.

    Intensity pot wiper to ground, look for about 2.47M

    V4, pins 2 to 7, about 2M.

    You may also want to verify C25, C22, and C24 for value.

    The LFO signal is not making it to V5b in sufficient strength to allow that to half to work as designed, looks like. Now we try to find out why.
     
    Thin69 and moosie like this.
  2. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,586
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    DC voltage, to elevated ground:

    R39/R41 junction: 14 to 30 mv

    R41/R40 junction, and hot side of R33: Strange. Mostly hovering around 14.5 to 16.5mv, but then making occasional excursions either upward, to 22mv, and back to center, or downward, to 9mv, and back to center.

    R34 upstream side (V5b plate): 7 to 33 mv

    R34 downstream side (V4a grid): 14 to 26 mv

    V5-7: 15.5mv, not wavering


    Resistance:

    V4-2 to ground: 1.92M

    Intensity wiper to ground: 1.9M (keep in mind, with each round of testing, I reset the wiper in the center, and it's never in the exact same spot.)

    V4-2 to V4-7: 1.91M


    Capacitance:

    C25 .022uf

    C24 .048uf

    C22 is a single .27uf component on the schematic, but is built from a .22uf and .047uf, in parallel. I didn't measure this, because I would have had to lift leads to do so, and it's really tight down there with the socket in the way. I was afraid of doing more harm than good. I did measure each component immediately before connection each to the board, and they were in spec. If we really need this, I can try again.
     
  3. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Has something simple like swapping V1 and V5 been tried? We know both halves of V1 play nice, let's give that a whirl in the tremolo spot and see what happens. It usually takes a nice strong tube to actually drive the oscillator circuit properly.

    I'll look at data and the schematic again. All those numbers for component values seem to be right.

    I really wish I could get a scope on this. Being able to view amplitude and phase versus time on multiple points at once would be a huge help.
     
  4. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Last idea for today, since I'm heading to work. If swapping the tubes doesn't help, look closely at the V5b cathode circuit. Also, make sure V5 pin 7's connector is making solid contact with the tube.

    I'd also like to see the plate, grid, and cathode voltages of the four stages in question with everything at idle, tremolo off, if it's not a bother.
     
  5. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,586
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    I've had several tubes in the V5 spot, including the one that was in V1.

    I don't see anything amiss. Not sure if I'm checking correctly. I checked continuity with the pin socket gently poking the thin probe into the tube end. All pin sockets on V5 make contact with their intended target, and nothing else. It's VERY difficult to see pins 6, 7, 8, but I was able to confirm with a small dental mirror that nothing's leaning against anything else, and that all the solder joints look OK. Also checked continuity with the cathode resistor and cap to ground.

    No bother!

    V4:
    1: 272.5
    2: 18mv
    3: 2.8

    6: 265.7
    7: 20mv
    8: same as 3 (connected by wire, checked by meter anyway)

    V5:
    1: 136.4
    2: 0.1mv
    3: 1.35

    6: 187.8
    7: 14mv
    8: 1.28

    Not sure I'm ready to commit here, but point me to a cheap scope that will do the job...

    Sorry, I still don't get how this thing works.

    V5a creates the LFO (easy), but instead of getting immediately amplified like in the 5 triode method, the LFO signal goes through the intensity pot, and then splits one way to V4b grid (still unamplified?), and the other way to V5b grid, which amplifies that half, and inverts it. Then it heads over to V4a.

    Seems no wonder the LFO isn't strong enough when it gets to V5b. V5a isn't getting amplified, right?
     
  6. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,586
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Never mind about the 4 triode explanation. I just read your description in the "Exploring..." thread. Still don't understand, but I'll go read about that paraphase inverter.
     
  7. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    V5a is the oscillator, and it looks to be running like a champ from your earlier reports.

    That big oscillation gets sent through the Intensity pot, where it gets reduced to the signal that V4b sees. Looks like that part works too.

    What V5b does, is gets some of that signal applied to the grid. If you follow from downstream of the Intensity pot and the 470k resistor (r39) there's a branch that goes through a 1M resistor (r41) and past another 1M to ground (r33). That should cut the signal a fair bit and send reduced signal to the grid of V5b. V5b takes that signal and amplifies it to drive V4a, inverting it along the way.

    That's the essence of the paraphase. Part of the output of the first stage drives the input of the second stage. You should see gradually lowering AC signal from the V5b plate to the grid of V4a, as it passes through each resistor and leak path to ground.

    One more experiment with that side of the circuit. Max out the Intensity setting, and check a couple of points: Intensity wiper, V4 grids, V4 plates, V5b grid, and V5b plate. Let's see what the drive levels are with a really strong, robust signal.

    If you've built an audio probe, it's time to put it to use next. Next will be passing full range audio, and listening to the circuitry of V4 to see what's going on in the frequency splitter network and bias shift/cutoff.

    If something doesn't pop out soon here, it might behoove you to search out someone knowledgeable to scope out this circuit thoroughly. Depending on where you are, I can probably find info for a tech in the general vicinity if need be. This is such a complex circuit it's mighty tough to troubleshoot it thoroughly by keyboard.

    I will say I kind of want to build one now after following this thread. I love the sound of the demos, and have a deep fascination with that lovely warbling tremolo.
     
    moosie and King Fan like this.
  8. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,586
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    My work here is done, then. :twisted::lol:

    This whole thing has me wanting to a) get this working of course, and then b) build a five-triode version, to compare. (That's my excuse anyway. It might have to be housed in a 6G8, of course...)

    Thanks for yet more detailed explanation. I had figured out most of that, based on your earlier stuff, and Merlin. Great to have the confirmation, especially in context of this actual circuit.

    I didn't realize the voltage divider attenuation was happening on the way to V5b. I assumed (without really looking) that it happened afterward, on the way to V4a. Makes sense, though.

    Probe parts should arrive Tuesday, if the mail isn't too wonky around the holiday. After that, it looks like it'll only take about ten minutes to make.


    I have one question... and I'm nervous about asking. You mention the AC signal. All the measurements I've done thus far have been DC. Did you want AC, especially on the grids? There's no signal being generated as I'm testing...

    I figure with the probe I'll record a single guitar tone into the looper, and send it through. Do you want me to do that now, measuring AC with the meter?
     
  9. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Stick with the DC reading for now, and slow down the trem speed if you need to.

    The tremolo signal is technically AC, but the frequency is so low it's easier to watch the swings on the DC reading - plus at that low a frequency a multimeter may not view it as AC. The old school analog meters and VTVM are great for this.
     
    moosie likes this.
  10. King Fan

    King Fan Poster Extraordinaire

    Posts:
    5,615
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    This thread is kinda like a Christmas gift to me -- opened early, really enjoying it -- thank you, gentlemen!
     
    moosie likes this.
  11. keithb7

    keithb7 Friend of Leo's

    Posts:
    4,934
    Joined:
    Jan 9, 2010
    Location:
    Western Canada
    I’m a bit overwhelmed by all the intense info. Some great documents have been supplied here. Very impressive. Someday maybe I’ll understand what it all means.
     
    moosie likes this.
  12. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    It's definitely been a voyage of discovery so far. I totally see why this kit is rated so high on the difficulty scale; an oscilloscope and signal generator are almost a necessity for properly troubleshooting this when it's being difficult - a DMM doesn't give you that one critical piece of info, which is the phase of signals traveling through. 30 seconds to set a pair of probes on the two V4 grids would be very illuminating indeed.

    And I still wanna build one one day! :D
     
    moosie likes this.
  13. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,586
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    You never did answer my question.... how cheap can I get away with, and have a decent (new) scope? Now I'm really wanting to see this thing through, but I can't spend a grand on it. $250? Possible?
     
  14. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Oh, no I didn't, did I? Sorry.

    A modern digital like my Rigol is about 350, and is an almost absurd amount of scope for this work. A used Tektronix in working order from eBay is very adequate, and will last decades with proper care. They're industry standard scopes, and built like tanks.

    You want 2 channels minimum, a reasonable sample frequency (50MHz is common for modern digital scopes, but well above what is needed for audio amp work), and it should have a pair of 1x/10x selectable probes or a pair of each type.

    If you want to pick out a couple and ask if they're adequate, feel free. If you have a ham radio club around, those guys will often have used analog scopes for sale as they upgrade to digital.
     
    moosie likes this.
  15. keithb7

    keithb7 Friend of Leo's

    Posts:
    4,934
    Joined:
    Jan 9, 2010
    Location:
    Western Canada
    moosie likes this.
  16. keithb7

    keithb7 Friend of Leo's

    Posts:
    4,934
    Joined:
    Jan 9, 2010
    Location:
    Western Canada
    I should hook my Tektronix up to my 6G8A. That would be cool! If y'all would walk me through it. I'm Clueless in Kamloops over here. I could post comparative findings for Moosie?
     
    moosie likes this.
  17. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    That would be awesome!

    Some points of interest would be each triode grid, but the real money spot would be the grids of the two triodes that handle the high/low frequencies at the same time - one probe on each grid, with the tremolo on. You should see two opposite waveforms overlaid. V4a and V4b on moosie's schematic, for reference.
     
  18. keithb7

    keithb7 Friend of Leo's

    Posts:
    4,934
    Joined:
    Jan 9, 2010
    Location:
    Western Canada
    Ok I’d like do do this. There’s this Christmas thing going on. I keep sneaking away from the inlaws when I can to follow this thread! My lovely timid wife will beat me with a turkey baster if I haul an amp chassis and my scope out, over the next 24 hours.

    I could do somtething after tomorrow probably.
     
    moosie likes this.
  19. clintj

    clintj Friend of Leo's

    Posts:
    4,957
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Put down the scope! Back slowly away from the scope! Lololol. :)

    No hurry, man. I've got my bench cleaned off for the holidays, too. No amp work for me til after the 1st, and I'm kind of enjoying it after the rush to get a bunch of Christmas work out the door.

    These threads keep my mind gainfully employed, are great for passing time on the bus to work, and since we're on curtailment at work I have ample free time to check in to get updates, think over the new data, and plot the next step. Far more enthralling than sudoku or crosswords. Family time is family time, though.

    Merry Christmas/Happy Holidays/your joyous greeting of choice, etc to you guys!
     
    King Fan and moosie like this.
  20. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    63
    Posts:
    17,586
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    I'm looking at a couple scopes. Both new, more or less, from Amazon.

    It might help to explain why I'm doing this. First, I'd really like to see this one through. To fix this thing on my own ... with uh, massive help from you all. But without handing it to a tech.

    Second, and more importantly, I see one need for a tool, I figure more scenarios will arise. At the very least, it'll be a learning tool.

    So...

    Siglent SDS1052DL for $215 (open box)

    Rigol DS1054Z for $349

    Far as I can tell, both come with basic probes.

    The Rigol is 4 channel. Does that mean, in my scenario, I could look at all four grids simultaneously? Or both grid and plate, on a pair of power tubes? That seems pretty cool... :D

    I looked at eBay briefly. I usually only feel comfortable buying used when I fully understand the product. Like guitars and amps. Also, the pricing, after shipping, and buying probes separately... well, the Siglent rivaled any of those prices.



    Quality concerns with either? Lack of features? I understand each of these must be fairly near the absolute bottom of the market.

    Anything to consider using these with tube amps, i.e. a high voltage probe? For example, will I be able to peek at my power supply, for the learning experience?

    @clintj, since I seem to be spending about what you did, which one - even one not listed - would you get?
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.