Announcement

Collapse
No announcement yet.

Bugera T50 - Intermitent (mostly no) Sound at the Speaker Out, but signal at the FX Send

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by stoneattic View Post
    since I'm reading a bias of -51.7 VDC, which is in the range Enzo suggested I was assuming that the Infinium circuit was doing it's thing.
    As a general range, that's right. But he said that before you measured the -30V bias in the working unit.
    The computer that is running the bias circuit probably also shuts down the drive to the grids. Is the infinium circuit on a separate board that you can swap between the 2 amps?
    Even if not, swap the power tubes and see if the working amp now biases those tubes (from bad amp) at -52VDC.

    Originally posted by Enzo
    I have a sign in my shop that says, "Never think up reasons not to check something."


    Comment


    • #32
      Originally posted by stoneattic View Post

      I think you nailed it!
      The DMM I was using is this one: https://www.manualslib.com/manual/48...man-82139.html
      On page 6 of the manual it appear to list the AC Voltage as (40-400Hz)
      One of the others I tried was a HF like the one you linked. The other I tried was an old Heathkit that I can't find specs for but I'm guessing is topping out at 400-450Hz as well.
      I will try again at lower frequencies.
      Thanks!
      I set the signal generator to 380Hz so that theoretically my "better" DMM should be able to read the AC signal. Still no luck with that meter, but the freebie HF one did read, although I have my doubts on it's accuracy since I read over 800 VAC on a meter that is only supposed to be good to 750 VAC. Here's what I got:

      "Good" amp:
      Plate resistors
      R71: 814 VAC (closest to EL34s), 487 VAC (farthest from EL34s)
      R61: 812 VAC (closest to EL34s), 542 VAC (farthest from EL34s)

      Bias resistors
      R68: 631 VAC (closest to EL34s), 314 VAC (farthest from EL34s)
      R67: 632 VAC (closest to EL34s), 300 VAC (farthest from EL34s)

      Pin 5 on EL34s: 0 VAC


      "Bad" amp:
      Plate resistors
      R71: 821 VAC (closest to EL34s), 507 VAC (farthest from EL34s)
      R61: 825 VAC (closest to EL34s), 522 VAC (farthest from EL34s)

      Bias resistors
      R68: 629 VAC (closest to EL34s), 302 VAC (farthest from EL34s)
      R67: 630 VAC (closest to EL34s), 289 VAC (farthest from EL34s)

      Pin 5 on EL34s: 0 VAC

      "0" means the DMM actually read 0

      So from the above I would say things look okay there....but...see my next post.

      Comment


      • #33
        Originally posted by g1 View Post
        As a general range, that's right. But he said that before you measured the -30V bias in the working unit.
        The computer that is running the bias circuit probably also shuts down the drive to the grids. Is the infinium circuit on a separate board that you can swap between the 2 amps?
        Even if not, swap the power tubes and see if the working amp now biases those tubes (from bad amp) at -52VDC.
        I swapped power tubes as suggested and the bias on the bad amp measured -51 VDC. I guessing that does point to the magic Infinium circuit being the problem, or does that just mean something else is wrong and the Infinium circuit is reacting to that? My understand of the how the Infinium circuit is supposed to work is that the LEDs should light if it sees a problem.

        Unfortunately everything except the panel with the back panel jacks is one PCB. It looks like the Infinium is part of the main, except for the LEDs themselves.
        Click image for larger version  Name:	20200628_213903.jpg Views:	0 Size:	1.87 MB ID:	908639
        Last edited by stoneattic; 07-02-2020, 02:27 AM.

        Comment


        • #34
          Let's test your meters. Set the meters to AC volts. Get a 9v battery, and measure the "AC VOLTS" coming from it. Yes, I know it is DC. A good meter would show a quick volt or two spike and settle down to zero volts AC. If your meter reads 9v battery as something like 12vAC, then the meter will be confused in circuits trying to read audio.
          Education is what you're left with after you have forgotten what you have learned.

          Comment


          • #35
            Yay, the site's back up!

            Originally posted by Enzo View Post
            Let's test your meters. Set the meters to AC volts. Get a 9v battery, and measure the "AC VOLTS" coming from it. Yes, I know it is DC. A good meter would show a quick volt or two spike and settle down to zero volts AC. If your meter reads 9v battery as something like 12vAC, then the meter will be confused in circuits trying to read audio.
            I tested my various DMMs as suggested and got the following results:

            Craftsman 82139: 9v battery test shows ~120VAC for a split sec, rapidly drops to a couple of volts and eventually settles to 0. (doesn't appear to actually read audio though)
            Harbor Freight freebie: 9v battery test shows a steady 20VAC (read high VAC when trying to read audio)
            Heathkit ~1990 handheld DMM - 9v battery test shows nothing, just shows 0 (failed with audio as well)

            All three read wall outlet VAC, so they work for standard VAC, just not the audio (frequencies) I need them for.

            So it looks like none of my meters are capable of reading audio. The only one that actually APPEARED to read audio (the HF) read crazy high values and failed the 9v battery test.

            It looks like I might need a better DMM. Any recommendations for a reasonably price model that will read audio? I know "reasonably priced" is a relative term, but I'm curious as to what would be recommended and what out there.

            Thanks!



            Comment


            • #36
              Other than teh HArbor Freight, the other two would at least measure AC siting on DC.

              Drop your audio frequency. At least under 400Hz. I used 100Hz most of the time, because unlike a piercing 1kHz, I can listen to 100Hz for extended time. Once the system is working, I can play full band audio and listen with my ears.

              Test your meters again: turn on your audio generator to the highest level it makes, now measure its output as AC volts. Try each meter at 100Hz, 400Hz, 1000Hz (1kHz).. That will tell you if teh meter can handle audio signals. No meter will respond well to actual music signal.
              Education is what you're left with after you have forgotten what you have learned.

              Comment


              • #37
                That ^^^^. Or wire a 0.1µF/630V cap in series with the positive Harbour Freight meter lead to block DC while measuring ACV.
                - Own Opinions Only -

                Comment


                • #38
                  Originally posted by Enzo View Post
                  Other than teh HArbor Freight, the other two would at least measure AC siting on DC.

                  Drop your audio frequency. At least under 400Hz. I used 100Hz most of the time, because unlike a piercing 1kHz, I can listen to 100Hz for extended time. Once the system is working, I can play full band audio and listen with my ears.

                  Test your meters again: turn on your audio generator to the highest level it makes, now measure its output as AC volts. Try each meter at 100Hz, 400Hz, 1000Hz (1kHz).. That will tell you if teh meter can handle audio signals. No meter will respond well to actual music signal.
                  I tested using the above method and on the Craftman read .033 VAC at 100Hz and 400Hz. I read .031 VAC at 1kHz. The other 2 meters did not read anything.
                  Does the .033 seem like a legitimate value?

                  I didn't get home until late and did not have time to try the 0.1µF/630V cap yet.
                  Last edited by stoneattic; 07-10-2020, 12:11 AM.

                  Comment


                  • #39
                    I still need to trace the signal, but just received my socket savers that I needed to connect my bias probes. The sockets are PCB mounted and too deep into the chassis for my probes to fit. Anyway, here are the results:

                    Good Amp
                    Ik (bias current): 33.2/33.6
                    Vp (plate voltage): 355.6/355.6

                    Bad Amp
                    Ik (bias current): 0/0
                    Vp (plate voltage): 365.4/364.0

                    (using the same tubes in both amps)

                    So it looks like this confirms that there is no bias on the EL34s. Which I assume means that the Infinium circuit is not functioning. I will go ahead and try to trace the signal at a lower frequency just for completeness, but it looks like the problem is going to beyond what I will be able to do with the Infinium circuit. I'm thinking now my best bet is to pull the board out and look for obvious issues on the back side (cold solder, etc.) and if I don't find anything maybe look towards building a fixed bias circuit for it.

                    Comment


                    • #40
                      Originally posted by stoneattic View Post

                      So it looks like this confirms that there is no bias on the EL34s.
                      Maybe I'm missing something, but it seems like with -51V bias you should get some current.

                      Wait a minute, this is a class A amp? I can't think of a EL34 based class A amp to even get a frame of reference, much less a 50W amp. Wikipedia says the bias needs to be much higher than push pull, so maybe -51V would give zero current. Could it be a blown cathode resistor?
                      Last edited by glebert; 07-10-2020, 02:56 AM.

                      Comment


                      • #41
                        Originally posted by stoneattic View Post
                        ...Good Amp
                        Ik (bias current): 33.2/33.6
                        Vp (plate voltage): 355.6/355.6

                        Bad Amp
                        Ik (bias current): 0/0
                        Vp (plate voltage): 365.4/364.0

                        (using the same tubes in both amps)

                        So it looks like this confirms that there is no bias on the EL34s...
                        Bias is a voltage between cathode and grid, negative in this field.
                        The 'Ik' reading will probably be cathode current at idle, which is controlled by the bias (voltage) but it is not, per se, bias. Ik might be referred to as, idle / static / quiescent (cathode) current

                        Previously, about -50Vdc was measured on the 'bad amp' EL34 control grids; that's the bias, it's probably excessive, hence this is not a 'no bias' scenario.

                        Sorry for being a pedant but using terminology incorrectly won't be helping communication or analysis here.

                        Originally posted by glebert View Post

                        Maybe I'm missing something, but it seems like with -51V bias you should get some current...
                        The 350V triode plot on p8 of https://tubedata.altanatubes.com.br/...129/e/EL34.pdf indicates about 10mA.

                        Originally posted by glebert View Post
                        ...Wait a minute, this is a class A amp?...
                        If the 'good amp' is only drawing about 30mA idle cathode current, I don't think it's likely to be intended to be class A.
                        Last edited by pdf64; 07-10-2020, 10:38 AM.
                        My band:- http://www.youtube.com/user/RedwingBand

                        Comment


                        • #42
                          Originally posted by pdf64 View Post

                          If the 'good amp' is only drawing about 30mA idle cathode current, I don't think it's likely to be intended to be class A.
                          On some websites it is labelled a class A, but this is how Bugera describes it. "Switchable Class-A/AB operation for ultimate power amp voicing: Class-A for classic warmth and Class-AB for raw, high-speed power" Maybe the mode switch is a place to look for a problem? I would try to do a resistance measurement from cathode (pin 8) to ground on a cold amp.
                          Last edited by glebert; 07-10-2020, 11:46 AM.

                          Comment


                          • #43
                            Originally posted by glebert View Post

                            Maybe I'm missing something, but it seems like with -51V bias you should get some current.

                            Wait a minute, this is a class A amp? I can't think of a EL34 based class A amp to even get a frame of reference, much less a 50W amp. Wikipedia says the bias needs to be much higher than push pull, so maybe -51V would give zero current. Could it be a blown cathode resistor?
                            It doesn't make sense to me either. I'm wondering if I screwed up the current reading doing something stupid like not taking it off standby, but I would think if that were the case I would not have read plate voltage. Doesn't standby typically mean only the heaters are on? I'm going to measure again when I get home tonight.

                            Comment


                            • #44
                              Originally posted by pdf64 View Post

                              Bias is a voltage between cathode and grid, negative in this field.
                              The 'Ik' reading will probably be cathode current at idle, which is controlled by the bias (voltage) but it is not, per se, bias. Ik might be referred to as, idle / static / quiescent (cathode) current

                              Previously, about -50Vdc was measured on the 'bad amp' EL34 control grids; that's the bias, it's probably excessive, hence this is not a 'no bias' scenario.

                              Sorry for being a pedant but using terminology incorrectly won't be helping communication or analysis here.


                              The 350V triode plot on p8 of https://tubedata.altanatubes.com.br/...129/e/EL34.pdf indicates about 10mA.


                              If the 'good amp' is only drawing about 30mA idle cathode current, I don't think it's likely to be intended to be class A.
                              Thanks for the clarification on the terminology. I guess since the probe is sold as a bias probe I assumed that the current reading is the "bias".

                              Comment


                              • #45
                                Originally posted by glebert View Post

                                On some websites it is labelled a class A, but this is how Bugera describes it. "Switchable Class-A/AB operation for ultimate power amp voicing: Class-A for classic warmth and Class-AB for raw, high-speed power" Maybe the mode switch is a place to look for a problem? I would try to do a resistance measurement from cathode (pin 8) to ground on a cold amp.
                                Yes, there is an A/AB switch. I'm not sure what it really does since I don't have schematic. But I was set to AB in both amps. I like the way it sounds there better.
                                I will measure the switch and cathode to ground on both amps when I get home tonight.

                                Comment

                                Working...
                                X