Announcement

Collapse
No announcement yet.

Low output from Eden WT800 in Bridge Mode

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

  • Low output from Eden WT800 in Bridge Mode

    Hello all.
    Okay, I've been lucky with trial, and error troubleshooting, and eventually repairing circuits. Don't want to waste anyone's time here with stuff I should already know. But I just can't figure this one out.
    Getting weirdness from bridge mode.
    First thing I did was replace the open 27K feedback resistor that had failed open.
    In bridge mode, I'm getting a larger signal on negative side (inverted right channel) than I'm getting from the positive side (non-inverted left channel).
    Both right, and left channels are putting out the correct wattage in stereo mode.
    Bridge switch seems fine. The R channel input is being muted by JFETS when in bridge mode. Getting signal from L channel output at 12K/27K junction.
    When scoping output negative terminal I'm getting 36V RMS before clipping
    Positive terminal is reading 22V rms.
    At this point, I can apply more signal and get 36V on Pos terminal as well, but the negative side is clipping.
    With the above 36/22V RMS scenario, I'm reading 58V across NEG and Positive with DVM.
    Obviously, far below the specified 800 watts for bridged into 8 ohm load.
    Using electric stove hotplates parallel to get a perfect 8 ohms.
    Hoping someone can see something obvious here.
    I appreciate, and humbly await any response to guide me.
    Thanks,
    Stan
    Attached Files

  • #2
    Do you have your scope ground disconnected when you measure with your DVM from neg to pos ?
    "Everything is better with a tube. I have a customer with an all-tube pacemaker. His heartbeat is steady, reassuring and dependable, not like a modern heartbeat. And if it goes wrong he can fix it himself. You can't do that with SMD." - Mick Bailey

    Comment


    • #3
      No grounds connected on both probes.
      Thanks for asking. Would upping the 27K resistor keep right channel from going into clipping? The circuit actually had a 27.4K on board.I'm thinking not?
      I really appreciate you taking the time to write back.
      Cheers!

      Comment


      • #4
        Not balanced

        Originally posted by Stan Keith View Post
        No grounds connected on both probes.
        Thanks for asking. Would upping the 27K resistor keep right channel from going into clipping? The circuit actually had a 27.4K on board.I'm thinking not?
        I really appreciate you taking the time to write back.
        Cheers!
        Schematic of more recent recent model attached, but same principle with the 12K/27K.
        Don't quite understand how this works, but could I balance neg/positive by tweaking that section?
        Thanks
        Attached Files

        Comment


        • #5
          I'm locked out of my shop at this time (in California), though I do have my database here. I forget what the power amp supply voltages are.....above +/-70V, less than +/- 80V if memory serves. The PCB layout on the power amp module is the same for the left Ch (right side viewed from the front) and right Ch (left side viewed from the front). You've already verified you're getting the same output level in Stereo Mode. The Bridge Mono mode is created by taking the output from the left Ch output, passing it thru the attenuator network which uses a 12k not on the PCB, and a 27k R16 on the Right Ch PCB, total 39k, which is the same value as the feedback resistor from the output stage to the Inverting input of the power amp diff pair. That makes the Right CH pwr amp an Inverting unity gain amp.

          Check what you get under No Load condition. Since you don't have a differential input Audio Analzyer (I presume), you still have your DMM in AC Mode. Set an arbitrary output level, of, say 10VAC on both channels in Stereo Mode. Switch to Bridge Mode, and check that you still have that same output level on both the Left and Right Outputs (relative to Ground). Should be the same. Connect the DMM between the Bridge Output binding posts (or Neutrik NL4, if you don't have the BP's). Should be 20V RMS. And, with scope, ground lead attached to spkr gnd, you should have the same level on both chennels, with the Right Ch opposite phase from the Left Ch in this mode. Increase the output until you see clipping in Stereo Mode again, still no load. Note your AC Output voltage. Switch to Bridge Mono, and see if your two channels are still the same, and if across the L-R outputs, you have twice the output voltage. If not, you can set up the Left Ch heat sink module so it's the same as the Right Ch, adding the Bridge Mode connection parts. I forget if that Bridge connection plugs into the amp board or into the power supply board, with the bridge input wire attached to the PCB. I didn't record close-up images of the Eden WT800 modules/chassis/wiring to give me those details.

          If you're NOT getting the same output voltage, UNLOADED, then yes, we're dealing with the external 12k resistor from the Left Ch output connector and the 27k resistor on the amp PCB for the Right Ch. If you are getting identical output levels at clip in both Stereo and Bridge Mono mode, and, twice that between the Left/Right outputs in Bridge, turn the output master down, and connector your load across Left/Right Output (only in Bridge Mono mode).

          If you're still getting the imbalance, that's when I'd swap the two amp modules in the chassis, taking care to connect the former Left Ch correctly for Bridge Mono with the chassis wiring. See if the problem moves under load. I'm assuming the problem occurs under load and NOT unloaded.
          Last edited by nevetslab; 03-26-2020, 06:00 AM.
          Logic is an organized way of going wrong with confidence

          Comment


          • #6
            Thanks for the detailed information, nevetslab.I really appreciate your time it took to post.
            It appears that the 27K I soldered in parallel with what I thought was an open resistor, was indeed good, and I was actually dealing with a resistor closer to 14 ohms at that point. OOPS
            I measured several times before making the call. Thought I had DVM probes on clean leads, but I think all the damn epoxy eden put on resistor still had some on where I was measuring.
            The whole time I thought there was something more to this simple way of inverting signal, but the schematic is so unreadable, I kept going down different rabbit holes.
            That said, how important is it that both +and- outputs are exactly the same?
            I was able to get them close to exact using a trimpot for the 27K. I'm getting about 36V RMS before clip from both +and- jacks. Using 2 ungrounded probes with a Tektronix scope This would equate to 648 watts into an 8 ohm load. Cheapish true RMS meter measures less across +and- jacks. So I'm thinking the scope should be trusted more?
            Is this an acceptable ballpark wattage output, or should I investigate further?
            Thanks again,
            Stan
            Last edited by Stan Keith; 03-26-2020, 01:40 AM. Reason: spelling grammer

            Comment


            • #7
              At this point I'm thinking your stove elements may not be acting the same with signal as without.
              Have you tried swapping them around to see if the imbalance moves to the other side?
              To verify their resistance under load you could measure current and voltage at the same time and calculate.
              "Everything is better with a tube. I have a customer with an all-tube pacemaker. His heartbeat is steady, reassuring and dependable, not like a modern heartbeat. And if it goes wrong he can fix it himself. You can't do that with SMD." - Mick Bailey

              Comment


              • #8
                Originally posted by Stan Keith View Post
                Thanks for the detailed information, nevetslab.I really appreciate your time it took to post.
                It appears that the 27K I soldered in parallel with what I thought was an open resistor, was indeed good, and I was actually dealing with a resistor closer to 14 ohms at that point. OOPS
                I measured several times before making the call. Thought I had DVM probes on clean leads, but I think all the damn epoxy eden put on resistor still had some on where I was measuring.
                The whole time I thought there was something more to this simple way of inverting signal, but the schematic is so unreadable, I kept going down different rabbit holes.
                That said, how important is it that both +and- outputs are exactly the same?
                I was able to get them close to exact using a trimpot for the 27K. I'm getting about 36V RMS before clip from both +and- jacks. Using 2 ungrounded probes with a Tektronix scope This would equate to 648 watts into an 8 ohm load. Cheapish true RMS meter measures less across +and- jacks. So I'm thinking the scope should be trusted more?
                Is this an acceptable ballpark wattage output, or should I investigate further?
                Thanks again,
                Stan
                Good find on the pair of 27k resistors. I've been down that road before. How important is it for the two power amps being equal? Mostly has to do with current each amp is contributing to the load. Considering we're really using this for amplifying music waveforms, which has NOTHING to do with symmetrical sine waveforms, the current being passed thru the load is always changing. It has more to do with headroom than anything else, I think. I'm sure more exacting answers to that will be forthcoming it this is way off base.

                On your Tektronix scope, you can place the two channels into single ch differential mode (each scope channel's sensitivity needs to be the same, one set to invert, as all my Tek scopes have, along with most or all other two or more channel scopes). So, you CAN measure differentially across the Eden's bridge mono output, and use GND on one of the probes to the Eden's output Gnd. And, yes, it will have far greater bandwidth than your DMM's True RMS AC Mode. Enough so that you can sweep out with your oscillator/signal generator to at least 100kHz, though I wouldn't be doing that at full power. 1 to 10W level more than adequate to see the amp's bandwidth. Measuring accurately the Average AC voltage with the scope is not as accurate as with the DMM, since you're taking the p-p magnatude, and dividing that by square root of 2 (1.414). I'm guessing your DMM's signal bandwidth isn't real high. Easy enough to check though.

                So, 36V per amp channel, or 72V in Bridge.....that into an 8 ohm load is ~648W, while 80V out in Bridge is 800W. So, you're in the ballpark. If you're not using a Variac and AC Line Monitor to keep the AC mains at 120VAC at the power amp source, that could be why its a little low. So, it sounds like you're working ok.
                Last edited by nevetslab; 03-27-2020, 12:45 AM.
                Logic is an organized way of going wrong with confidence

                Comment


                • #9
                  Originally posted by g1 View Post
                  At this point I'm thinking your stove elements may not be acting the same with signal as without.
                  Have you tried swapping them around to see if the imbalance moves to the other side?
                  To verify their resistance under load you could measure current and voltage at the same time and calculate.
                  That is something I need to investigate.Thanks for that!

                  Comment


                  • #10
                    Originally posted by nevetslab View Post
                    Good find on the pair of 27k resistors. I've been down that road before. How important is it for the two power amps being equal? Mostly has to do with current each amp is contributing to the load. Considering we're really using this for amplifying music waveforms, which has NOTHING to do with symmetrical sine waveforms, the current being passed thru the load is always changing. It has more to do with headroom than anything else, I think. I'm sure more exacting answers to that will be forthcoming it this is way off base.

                    On your Tektronix scope, you can place the two channels into single ch differential mode (each scope channel's sensitivity needs to be the same, one set to invert, as all my Tek scopes have, along with most or all other two or more channel scopes). So, you CAN measure differentially across the Eden's bridge mono output, and use GND on one of the probes to the Eden's output Gnd. And, yes, it will have far greater bandwidth than your DMM's True RMS AC Mode. Enough so that you can sweep out with your oscillator/signal generator to at least 100kHz, though I wouldn't be doing that at full power. 1 to 10W level more than adequate to see the amp's bandwidth. Measuring accurately the Average AC voltage with the scope is not as accurate as with the DMM, since you're taking the p-p magnatude, and dividing that by square root of 2 (1.414). I'm guessing your DMM's signal bandwidth isn't real high. Easy enough to check though.

                    So, 36V per amp channel, or 72V in Bridge.....that into an 8 ohm load is ~720W, while 80V out in Bridge is 800W. So, you're in the ballpark. If you're not using a Variac and AC Line Monitor to keep the AC mains at 120VAC at the power amp source, that could be why its a little low. So, it sounds like you're working ok.
                    Thanks. Put amp back together. But I'm going to put back on bench, and adjust variac to stay at 120 V as I observe scope. BTW, I'm using the RMS feature on my Tektronix 2430A. I believe it has a bandwidth of 100MHz. This is not as good as a DMM for measuring RMS? Until now, I was just injecting a 1KHz sinewave, and not doing a full band sweep. I will do a sweep at 40Hz and up. Thanks guys for giving lots of food for thought. I'm starving! ha

                    Comment


                    • #11
                      Originally posted by Stan Keith View Post
                      Thanks. Put amp back together. But I'm going to put back on bench, and adjust variac to stay at 120 V as I observe scope. BTW, I'm using the RMS feature on my Tektronix 2430A. I believe it has a bandwidth of 100MHz. This is not as good as a DMM for measuring RMS? Until now, I was just injecting a 1KHz sinewave, and not doing a full band sweep. I will do a sweep at 40Hz and up. Thanks guys for giving lots of food for thought. I'm starving! ha
                      I've never read the specs on their 2430A scope. The scope's vertical amplifiers have a BW of 100MHz, though I doubt if the RMS DMM in the system does. Should be in your Specs in your Owner's manual (wherever that is now). Having 100MHz BW on an RMS Multimeter.....that's in the expensive class of instruments. My Marconi 2610 has a 25MHz BW in the 200mV-2V range, drops down to 5MHz on 20V range, & 20kHz on the 200V range. My HP3400A I think is a 100MHz BW instrument, as was the Fluke 8920A that I once owned. Sold that. I'd guess your DMM of the 2430A is at least 20kHz.
                      Logic is an organized way of going wrong with confidence

                      Comment


                      • #12
                        As long as the signal is a pure sinewave there shouldn't be a difference between equivalent RMS and "True RMS". Extended bandwidth above the signal frequency is only required for the higher harmonics that don't exist in a sinewave.
                        - Own Opinions Only -

                        Comment


                        • #13
                          So, hooked up to variac with current meter. This thing draws about 14 amps. The sine waveform is starting to clip at 32V MS on each leg as the mains (AC monitored with DMM) drops to 113 V.
                          Unfortunately, I think I'm not getting good readings because the variac is rated for 5 amps ( I primarily do vacuum tube repairs). At least that's what the fuse is. Taking a chance when measuring just for a few seconds.
                          Was able to ride AC with variac to keep it at 120V. About the same wattage as before- I did gain a few watts at 1KHz, not so much with 40Hz. I believe 72x72/8 = 648 watts, right? The 72 figure is the bridged RMS voltage of +and- neg summed.
                          At 1KHz I'm getting more like 74x74/8 = 684 watts.
                          Now, I have to think about my equipment not up to par measuring high current amps like this.
                          I would think if it's quiet, and not distorting, the amp is probably okay at that point? Or could there be something more subtle that's sapping wattage?
                          Thanks again in advance for any info on this.
                          Stan

                          Comment


                          • #14
                            Originally posted by g1 View Post
                            At this point I'm thinking your stove elements may not be acting the same with signal as without.
                            Have you tried swapping them around to see if the imbalance moves to the other side?
                            To verify their resistance under load you could measure current and voltage at the same time and calculate.
                            Originally posted by g1 View Post
                            At this point I'm thinking your stove elements may not be acting the same with signal as without.
                            Have you tried swapping them around to see if the imbalance moves to the other side?
                            To verify their resistance under load you could measure current and voltage at the same time and calculate.
                            Please tell me if I have hooked this up incorrectly-
                            Hooked up DMM for AC amps in series with + side of amp to + terminal on dummy load. Brought up variac while observing both +and- RMS V. (inverted from each other) Both were at 22V, and I was drawing 6A AC current thru + leg..
                            However, when hooking up DMM exactly the same way on - side, the two signals are in phase. What the #&^?
                            I'm missing something fundamental here that would like to be schooled on.
                            Also, for the simple ohms law calculation, is the 6 amps I initially measured the total of both +and-outputs? So would it be 22RMS each side = 44V/6 amps= 7.3 ohms? Or would I double the current as well? 44/12 = 3.6 ohms?
                            if so, that would almost explain a few things.
                            But what of the two waveforms being in phase when hooking up DMM as described before?
                            I am a sponge. Please dip me in your knowledge.

                            Comment


                            • #15
                              However, when hooking up DMM exactly the same way on - side, the two signals are in phase. What the #&^?
                              It is not clear how you measured and what - side means. I suggest to post a drawing that shows how you connected the load and your meter.
                              - Own Opinions Only -

                              Comment

                              Working...
                              X