Announcement

Collapse
No announcement yet.

SVT-CL fault green/red blinking.

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

  • SVT-CL fault green/red blinking.

    Today I've played my SVT-CL, it started to crackle and lost low end & finally went to the fault mode (with red and green lights blinking).
    I've noticed two burnt screen resistors (it's already a version that has 220ohm/0,5W without diodes) - I checked all the preamp and power tubes and one KT88 had a short.
    (this is pretty new set, sad that one went bad after few hours of playing.. never mind)

    I have re-done all the 6 resistors tonew 220ohm/1W & checked solder joints. Put a spare KT88 to match all the rest 5 pcs, powered it on, but it will not bias (can't get green light, it just does not lit on both indicators) and when I try to raise the bias up, it does little click and puts the amp into the fault mode. I've already checked main board for burnt resistors and bad solder joints (nothing odd found). Also, nothing new is burned on power tube board. Everything seems to be working just fine. I have checked the same with borrowed set from SVT-VR and it's the same.

    Anything I could check? Am I missing something

    EDIT:
    1. Pulled tubes off: all the voltages on power tube sockets checked and they are the same. On driver tubes (12AU7) I have 390V on Pin1 and Pin6, but on the phase inverter tube I have 540V (in the same places) - is it OK?
    2. I can actually get a green bias light on the 2nd Bias Control, but I can't get it on the 1st B.C. No matter what tubes are there and no matter with 12AU7 swapping. The fault mode kicks in only when I break certain level of bias adjustment of 1st Bias Control (red light come on and a bit further, it's in fault mode)....

    like this: https://www.dropbox.com/s/4l4ooxglmsnjq9p/svt.mov?dl=0

    Last edited by boroman; 08-20-2020, 02:51 PM.

  • #2
    Where is your meter to set up the bias? Looks like one of the JJs is faulty from the sound. Looks like one of the JJs is faulty from the sound.
    Support for Fender, Laney, Marshall, Mesa, VOX and many more. https://jonsnell.co.uk
    If you can't fix it, I probably can.

    Comment


    • #3
      As the amp has already entered Fault condition repeatedly, I'd mark the power tubes per their positions, and leave them, along with the two 12AU7 driver tubes and the 12AX7 preamp/phase splitter tube out. First, the voltages you're reading on the driver tubes plate pins sounds reasonable, as there's no load on the supplies, other than the two preamp tubes. And, with the power tubes out, the plate voltage at pin 6 of the 12AX7 seems normal, since there's nothing to drop the supply voltages down in normal operation.

      And, we know you're NOT in Fault condition with the tubes removed, or there wouldn't be any High Voltages, as that power transformer's primary is connected by the relay on the AC Mains PCB, which is controlled by the current/voltage sense circuits on both the main PCB as well as in the Preamp section.

      First, I'd check to see what the bipolar supply voltages are, just to be sure we have both the +/- 15V for the LED Comparator Circuits and Fault Protection sense circuits. I normally have the power amp chasis and preamp chassis standing upright on the power xfmr end, with a small block underneath the main pwr xfmr to stabilize the chassis. The 'Garden Hose' containing the I/O wiring to the preamp will no doubt fight you with regards to where you want to place the preamp. I have a roadmap from the PCB layout and parts placement guides to help direct me while viewing everything from the bottom side of the main PCB.

      On the LED Comparator circuits, the window comparator voltages are nominally 0.147V and 0.294V, found on the inverting pins of the op amps used in the circuit. The nominal Fault buss,which is fed thru the six signal diodes D49 thru D54 is nominally 0.78VDC to 0.86VDC when a good set of power tubes are installed, nominal bias current around 23mA ea. The Fault Detection circuit is IC2B, whose output voltage is nominally +14V, and -14V under Fault conditions. There should be a 0.1uF cap in place between pins 7 & 6 to prevent false triggering of the detector.

      I've had issues with this circuit when Cap C13 has gone bad. The output of this Fault Detect circuit feeds the Relay circuit by way of D2. It leaves the main PCB at J12, and hits the Relay circuit on the AC Mains PCB at J35. Under normal conditions, the potential at J35 is a slow-rising voltage from R43/C12, then feeds J12 thru R44, and is only influenced by Fault conditions via diode D2, which can inhibit or shut down the relay control circuit. So, all of these circuits need to be healthy. I've had to spend time correcting the Fault Detector circuit in order to get perfectly good power tubes to work again. Not sure if that's what you're facing, as you've already swapped a known good set of tubes in from the SVT-VR. Though, that's no guarantee that the other tubes with the nominal bias settings used for your KT-88's will be correct for those.

      You've already indicated not having both LED Bias circuits reading properly, so you may also have to troubleshoot that circuit. I assume you have all the service documents on this amp, though I've added some additional doc's that might be useful.


      Chassis Wiring Diagram.pdf

      Click image for larger version  Name:	SVT-CL Power Amp PCB Parts Location Bias Adj.jpg Views:	0 Size:	192.5 KB ID:	911555 Click image for larger version  Name:	Power Tube PCB Parts Location.jpg Views:	0 Size:	344.9 KB ID:	911557
      Attached Files
      Last edited by nevetslab; 08-20-2020, 06:20 PM.
      Logic is an organized way of going wrong with confidence

      Comment


      • #4
        I find this set-up orientation easiest when checking bias/plate current on the tubes. If the amp goes into protect mode, I'll remove the sextet of power tubes and try again. If that persists, I start looking at the power supply voltages from the bottom side of the PCB, while armed with the schematics, parts layout diagram, and begin probing, carefully. If I have to go digging deeper, then I leave this orientation to disassemble. I will have already removed the hold-down clamps, allowing me to check tubes in pairs, once I've marked them per their initial orientation.

        Click image for larger version  Name:	SVT-CL Tube Matching & Bias-1.jpg Views:	0 Size:	240.8 KB ID:	911576 Click image for larger version  Name:	SVT-CL Tube Matching & Bias-2.jpg Views:	0 Size:	198.8 KB ID:	911578 Click image for larger version  Name:	SVT-CL Tube Matching & Bias-4.jpg Views:	0 Size:	259.2 KB ID:	911580 Click image for larger version  Name:	SVT-CL Tube Matching & Bias-5.jpg Views:	0 Size:	262.0 KB ID:	911582
        Attached Files
        Logic is an organized way of going wrong with confidence

        Comment


        • #5
          Once again, I'm being plagued with a second set of files inside a box following the images I uploaded using the little picture icon above in the tool bar. WHAT am I doing wrong here???!! This drives me nuts being at the mercy of software shit.
          Logic is an organized way of going wrong with confidence

          Comment


          • #6
            The dual PS rail for the bias comparator is powered from one of the windings on the filament PT. (The main filament connections to the output tube board are connected with spade terminals that get oxidised over time and this creates a build-up of resistance. These terminals see just over 11A in normal operation. The startup surge (when the fils are cold) can jump many times higher when the terminal connectors are oxidised, causing arcing and flashing at these terminals on the output tube board - tell-tale sign if there is charring around these connectors). My hunch is the arcing can cause transients which get inducted back through the filament PT to the other windings (including the bias comparator supply). So check that the bias comparator power supply hasn't been compromised.
            Building a better world (one tube amp at a time)

            "I have never had to invoke a formula to fight oscillation in a guitar amp."- Enzo

            Comment


            • #7
              Originally posted by tubeswell View Post
              The dual PS rail for the bias comparator is powered from one of the windings on the filament PT. (The main filament connections to the output tube board are connected with spade terminals that get oxidised over time and this creates a build-up of resistance. These terminals see just over 11A in normal operation. The startup surge (when the fils are cold) can jump many times higher when the terminal connectors are oxidised, causing arcing and flashing at these terminals on the output tube board - tell-tale sign if there is charring around these connectors). My hunch is the arcing can cause transients which get inducted back through the filament PT to the other windings (including the bias comparator supply). So check that the bias comparator power supply hasn't been compromised.
              YUP! Seen that a good many times. And, of course, you don't see it unless you go digging. As he had already opened that board up to swap out the screen resistors, he might have noticed those connections if that condition was obvious.....thought the crackling could well have been from just that condition. Good point!
              Logic is an organized way of going wrong with confidence

              Comment


              • #8
                Originally posted by nevetslab View Post

                YUP! Seen that a good many times. And, of course, you don't see it unless you go digging. As he had already opened that board up to swap out the screen resistors, he might have noticed those connections if that condition was obvious.....thought the crackling could well have been from just that condition. Good point!
                Yup Click image for larger version

Name:	Power amp output board.JPG
Views:	426
Size:	4.84 MB
ID:	911624
                Building a better world (one tube amp at a time)

                "I have never had to invoke a formula to fight oscillation in a guitar amp."- Enzo

                Comment


                • #9
                  nevetslab Thanks so much for your time. I really appreciate that. I printed that post in the morning and went to check those voltages, but with a fresh set of eyes I saw R40 with defect (hole). It's hidden quite behind it and from the first look resistor does not look burned, but it measures 1M ohm (and going up!). Could it be the problem here? I don't have anything close to 10R/2W here, so will be ordering it soon. It seems that shorted tube made some mess not only on the tube board...

                  tubeswell Those places were checked. This is pretty clean amp, used in a A/C cotrolled room, so no corrosion and loose wires.

                  Attached Files

                  Comment


                  • #10
                    Originally posted by tubeswell View Post
                    The dual PS rail for the bias comparator is powered from one of the windings on the filament PT. (The main filament connections to the output tube board are connected with spade terminals that get oxidised over time and this creates a build-up of resistance. These terminals see just over 11A in normal operation. The startup surge (when the fils are cold) can jump many times higher when the terminal connectors are oxidised, causing arcing and flashing at these terminals on the output tube board - tell-tale sign if there is charring around these connectors
                    That's an excellent observation that needs to be noted by anyone who works on this series of amps, also the Fender/Sunn Bassman 300. Another tell-tale is the plastic covers over the spade connectors turn brown then black from the heat. Sometimes they melt right off. Snip off the spade connectors, strip the filament wires back maybe a cm or so (3/8 inch in USA) cover the insulation near the ends with an extra layer of heat shrink, then solder the filament leads directly to the PC board lands. Bingo - no more contact resistance problems, yay!

                    This isn't the future I signed up for.

                    Comment


                    • #11
                      Originally posted by boroman View Post
                      nevetslab Thanks so much for your time. I really appreciate that. I printed that post in the morning and went to check those voltages, but with a fresh set of eyes I saw R40 with defect (hole). It's hidden quite behind it and from the first look resistor does not look burned, but it measures 1M ohm (and going up!). Could it be the problem here? I don't have anything close to 10R/2W here, so will be ordering it soon. It seems that shorted tube made some mess not only on the tube board...

                      tubeswell Those places were checked. This is pretty clean amp, used in a A/C cotrolled room, so no corrosion and loose wires.
                      The last source I had on those was Digi-Key https://www.digikey.com/products/en?...s=RSB-10RCT-ND

                      Previous source, also thru them was a part made by Huntington Electric, 3W 10 Ohm 3%

                      Logic is an organized way of going wrong with confidence

                      Comment


                      • #12
                        Originally posted by Leo_Gnardo View Post

                        That's an excellent observation that needs to be noted by anyone who works on this series of amps, also the Fender/Sunn Bassman 300. Another tell-tale is the plastic covers over the spade connectors turn brown then black from the heat. Sometimes they melt right off. Snip off the spade connectors, strip the filament wires back maybe a cm or so (3/8 inch in USA) cover the insulation near the ends with an extra layer of heat shrink, then solder the filament leads directly to the PC board lands. Bingo - no more contact resistance problems, yay!
                        While I've also done that direct soldering down to the base of where the Male Fast-On terminals are pressed in/soldered, it now 'tethers' that PCB to the chassis when you gotta go digging further. I never did invest in a hand tool that can crimp the right angle 1/4" insulated Female Fast-ons of either type. The current version of the PCB connection is with dedicated solder pads for those heavy wires to go into. I found out the hard way while desoldering them and attempting to slip the wires out of the holes, it instead lifted those huge wide traces off the PCB!! Cheap PCB material from Asia strikes again.
                        Logic is an organized way of going wrong with confidence

                        Comment


                        • #13
                          Okay... I have replaced bad cathode resistor and now it does not go to the fault mode with Bias #1 set after specific threshold. But still, all I could get from Bias#1 is no light or red light. I don't suppose it's just a fault of bad green diode, because no matter how I try to set it, it's humming (hum like unbalanced bias), so the problem is related to somehwere else

                          I have found similar problem over the internet with some tips and diagrams (below/attachments).

                          "Measure your "+15V" rail (should be 15V); also IC4 pin 6 and IC6 pin 6 (should be 0.294V, call it 0.3V).
                          A red LED comes on when the drop across any 10 Ohm cathode resistor exceeds that 0.294V (say 30mA). If ALL are going on at the same time, it suggests your "+15V" is out-o-whack.
                          Why the "+15V" is sick: we start with raw +20V and drop it on a 30 Ohm resistor and a 15V Zener. Many 1990s-era amps have trouble here. The hot resistors buried in a hot tube amp cook their PCB solder joints, and fail. Inspect the area around R60 R61 D27 D28, repair toasted PCB as needed, and replace generously with up-sized parts mounted for better air cooling.
                          This also has CB1 CB2 "THRMSTR"s into the +/-20V rectifier. I don't like that. Use both probes on ACV to read the voltage drop across each CB when running. Anything more than a Volt is probably wrong, the thermistor has drifted beyond spec, and is strangulating the poor +/-20V supply. I'd just jumper them, though a couple 1A SB fuses may be a wiser choice."


                          I have measured those voltages at IC4 and IC6 and they are OK.
                          +15/-15V rail too.
                          Checked R60/R61 = OK

                          Still, don't know where to dig next...?
                          Attached Files

                          Comment


                          • #14
                            IC5A, IC5B and IC6A are the window comparator stages for the GRN LED on the lower half of the output stage V4-V6. You might need to replace Quad op amp IC5 (TL074) and see if that restores the GRN bias LED. When I'm setting up an SVT after power tube failure, I will have removed the hold-down clamps so they're out of the way. I'll have all six tubes removed, then pre-set the bias voltage on the driver tubes Pin 8 for a nominal voltage of around -45VDC. I'll install one pair of power tubes....usually positions V3 & V4, and, measuring across the cathode resistors for those two tubes (R36 & R38), see if I have around 230mV (23mA). If its way high or way low, I'll remove the power tubes, and reset the bias on both pots until I'm in that 'ballpark' figure. Then, I'll go thru and check the rest of the tubes, in pairs, still using positions V3 & V4. I write down the cathode voltage for each tube (which I had previously marked with a Red or Blk Sharpie per tube position V1 thru V6 when I first removed them. 230 mV across 10 ohms = 23mA

                            During this procedure, having preset the bias voltage the same for the upper half and lower half, now you're getting a look at the nominal cathode current of each tube, and it will allow you to 'juggle' the six tubes for a best match. If I had to move tubes around for that as a new starting point, THEN I'll install all six tubes and measure the cathode voltage for each tube, listing each on a chart, and will then will tweak the two bias pots, seeking an average current of 23mA. The spread may be 19mA thru 26mA, or even worse, as often happens over time. I'll often have to swap tubes around again until I get as close as possible. I don't like having 26mA on any one side, as you'll end up seeing the RED LED eventually turn on after it sits and idles a while. I'd shift the average current for those down to keep out of the RED zone.

                            The GRN LED circuit will ONLY light when there are three tubes installed in the upper & lower halves. During the initial measurements, only the RED LED might light if it has a lot higher transconductance than the other tubes.

                            I normally have a 4 ohm dummy load connected on the output. Once I have a decent plate current balance on all six tubes, I'll let it idle for a while....20-30 minutes and see if the balance has remained stable. If so, I'll power down, remove the dummy load and connect the speaker to see if there's hum or not. Usually, if I have good plate current balance, hum is minimal. Sometimes I'll have to tweak the bias to minimize it, and re-measure the resulting plate current per tube. Then, I'll move my cooling fan into place to cool the power tubes down. Then, once cool, I'll remark those tubes that have been moved to new positions. Remove the six tubes, and put the hold-down clamps back into place, then put the power tubes back in. I won't reassemble the amp until I've gotten thru all of the bias adjustments, as well as closely scrutinized all of the PCB's for solder joint fractures....particularly all of the rear panel parts....bias pots, Preamp Output/Power Amp Input jacks, XLR connector, as well as the power supply harness connector headers...those 4mm spaced headers are known to develop solder joint fractures.

                            If you've had to lift out the main PCB at any time, that's when you want to tighten the Output Xfmr's mounting screws, as two of them are inaccessible when the main PCB is in place. There's one screw that's critical....the left front corner on the Output Xfmr side that mounts the main PCB to the chassis standoff. The main PCB's ground to chassis is thru this path. I often find the standoff itself gets loose, so it HAS to be tight, as does the screw into it thru the PCB. The Power and Output xfmrs get loose in these amps all too easily.

                            I'd also recommend pulling the preamp apart to check for solder joint fractures. The sheer weight of the amp in transit over time shakes everything, so with so many parts being the 'supporting' mounting members of the PCB (pot solder terminals, jack terminals)....solder joint fractures are an ongoing issue with these amps.
                            Last edited by nevetslab; 08-26-2020, 07:14 PM.
                            Logic is an organized way of going wrong with confidence

                            Comment


                            • #15
                              nevetslab Yesterday I put all of this amp tubes to another CL (Heritage) and running sweet and stable, so in this case, tube mismatch (and in general, tube fault) is deafinitely not the case. They are all good.

                              I'll try to look at IC5 & IC6 - but wouldn't it be off at 12AU7 drivers then (if one of IC's is bad)? Voltages seems OK at the drivers. And about the preamp. I might check and swap the preamps from CL's but from what I see, speaking shortly, preamp delivers one stage of voltages, and then its kind of separated down in the main PCB (main chassis). That would mean, the problem still lays in the main PCB (sorry for lack of professional language, I'm not so experienced when talking bout that stuff)

                              I remember reading about 9V battery into some of the solder joints to check if the green light is OK.
                              Do you remember where I should connect it?

                              I'll deafinitely check the bias on the cathode resistors today. We'll see what are the readings...

                              Comment

                              Working...
                              X