VOX V15 – Under The Hood

VOX V15 – Under The Hood

So, we’ve ascertained that there is power in to the VOX V15 and some valve lights but nothing audible except for a loud buzz. Time to take it apart and see what we have.

With the amp still plugged in and the mains switch turned off, we are earthed and so can begin poking about. Loosening the four  screws as shown means we can take the main board away from the cab and turn it over:

Unscrew the 4 screws holding the amp into the VOX V15 cab

Unscrew the 4 screws holding the amp into the VOX V15 cab

We can see 5 valves (one of which is a bit wonkily seated), what I take to be transformers, the circuit board printed on the back (nice touch!) and a mass of cobwebs:

VOX V15 Amplifier

VOX V15 Amplifier Board

Removing the cover from the leftmost valve in the picture – I’m guessing it’s a heat shield as it’s right next to the transformer – I switch on the amplifier and wait to see what happens.

VOX V15 Valves

VOX V15 Valves

The two larger valves and the two leftmost ones light up immediately, the one in the middle with the milky appearance doesn’t – looking closer we can see a hairline crack in the top. Uncle Google pointed me to this page which states:

  1. Valves can be quite dramatic if they blow with loud crackling noises and a very bright glow from the valve as the valve consumes itself.
  2. A valve can also blow with little or no visible effect.
  3. Sometimes, if the vacuum in the valve is lost through tiny hairline cracks, you will see that the silver coating on the inside top of the valve will disappear or go clear.
VOX V15 Cracked Valve

VOX V15 Cracked Valve

I can’t identify *anywhere* what kind of valves are recommended and so I  created a thread in the illustrious TDPRI forum for advice. Within 20 minutes I’ve got the following:

3 x 12AX7 and 2 x EL84 are what you need. This could very well be the source of the problems, but like anything in life there are no guarantees. Good luck with the resurrection!

———

Update: 13/4/2015

1 – centre tap

2 – red lead from winding to V4 tube

3 – blue lead from winding to V5 tube

Click for full size image

Click for full size image

Update: 21/4/2015

Click for full size image

Click for full size image

Update 29/04/2015

Click for full size image

Click for full size image

Click for full size image

Click for full size image

 

50 Responses

  1. Wyatt Noyes
    Wyatt Noyes at · Reply

    Phase inverter is ecc81 or 12at7.

  2. krrunch
    krrunch at · Reply

    Many thanks Wyatt! Sorry for the delay in replying, this has gone on a back burner as I am totally flummoxed as to where to go next .. valves have been replaced but no joy yet.

  3. Jon Ledwith
    Jon Ledwith at · Reply

    Okay. if you are brave enough, I can guide you through the process. You need a DVM at a bare minimum. An oscope is even better. The idea is to use a methodical approach to isolate the fault and not just shotgun components. Let me know if you are game.

    1. krrunch
      krrunch at · Reply

      Jon, that is a most amazing offer and I am definitely game! I do have a voltmeter, albeit a fairly budget one but if that gets me off to a start then brilliant! When you say brave, are you referring to the brick-sized capacitors or the learning curve? 🙂

  4. Wyatt Noyes
    Wyatt Noyes at · Reply

    Okay, so I know you have the schematic and a soldering iron. I would also recommend keeping the NFB hooked up until we sort out the hum. First step, remove all the preamp tubes but leave in the phase inverter. Turn the amp on with all controls set to zero. Does it still hum loudly?

    1. Wyatt Noyes
      Wyatt Noyes at · Reply

      Sorry, I forgot to clarify to leave in the power tubes.

  5. krrunch
    krrunch at · Reply

    Hi Jon, I’m away with work until Tuesday but will do this as soon as I get back. If I’ve understood correctly then I should remove the 2 x AX7 preamp tubes and leave in the Phase Inverter (AT7) and the power tubes (EL84). I’ll report back as soon as I’m home.

  6. Wyatt
    Wyatt at · Reply

    Yes. Don’t forget to turn the controls all the way counter clockwise. This will rule out the preamp stage if the hum is still there. No need to break out the soldering iron yet.

  7. krrunch
    krrunch at · Reply

    Finally got this done :/ I took out the 2 x AX7 and turned all knobs counter-clockwise to zero – the hum remains. There is a slight but discernable change in amplitude when I move the MASTER knob from 0 to about 1 but that’s it.

    One thing I noted though I’m not sure if it means anything: the power tubes light up happily but there is no glow from the PI AT7, I also tried an old one in case the newer one was bust but no light from that either. The real question is: should there be?

    Many thanks,
    K

    PS I have this on my workbench now so no more delays as opposed to previously where it was half way across the city in a garage ..

    1. Wyatt
      Wyatt at · Reply

      Excellent progress. In mine the AT7 pi glows. You should check to see if the heater voltage is present at the tube. The heater should be wired in parallel. You should see around 6.3vac between pins 9 and 4 and the same between pins 9 and 5 at the socket. While you are at it you could do a dcr check on the tube to make sure you see resistance between the same terminals on the tube. Pins 4 and 5 in the amplifier should be tied together. It is quite unusual that the heater would be bad in the tube. If the voltage is present at the tube socket, you may need to “re-tension” the socket. One thing in our favor is that I have the same amplifier here. While you have the tube out, could you check to see if the hum is present without the pi? Mine is silent with it removed.

  8. krrunch
    krrunch at · Reply

    OK, I broke out the multimeter:

    DCR: (calibration set to 200 Ohm)
    pins 4 + 5 on AT7 tube – 18.2 Ohm
    pins 5 + 9 on AT7 tube – 10 Ohm
    pins 4 + 9 on AT7 tube – 10 Ohm

    pins 4 + 5 on socket – 0.5 Ohm (so as good as tied together?)
    pins 4 + 9 on socket – 0.7 Ohm (so as good as tied together?)
    pins 5 + 9 on socket – 0.7 Ohm (so as good as tied together?)

    AC voltage:
    pins 4 + 9 on socket – 6.5
    pins 5 + 9 on socket – 6.5

    With the AT7 PI removed completely, the hum is still there.

    I have the old AT7 and the new one I ordered when I thought it may be bust – the new one is marked OTK1 which is apparently the same as a 6DJ8 (Russian new old stock) and which I was told is a direct replacement for the AT7.

    I first tried the old AT7 valve – I popped it half out (to rule socket tensioning out) and measured the voltage across pins 9/5 and 9/4 – both are around 6.3vac and the tube *does* get hot though there is no glow.

    I then tried the same with the OTK1 valve – again the voltage across the valve pins was 6.3 vac but the tube stays cold. I guess this means that my original tube is still OK but the new one has a busted heater?

    Thanks
    K

  9. Wyatt
    Wyatt at · Reply

    So… In this circuit a 6dj8 will not work as a substitute. The 6dj8 requires 6.3vac between pins 4 and 5. Your old tube is not an option as the white color is indicative of a breach in the vacuum. When we resolve the source of the hum, you can use a 12ax7 in the pi position while you are sourcing a 12at7. I have found the EH and fender branded ones work fine if you don’t want to pay big bucks for an nos variety.

  10. Wyatt
    Wyatt at · Reply

    So at this point you should unplug the amplifier, remove the chassis from the back panel, and flip it over. Keep it unplugged and check a few things. Make certain the electrolytic capacitors are fully discharged. Use the diode check function in your DMM and check the four diodes that form the full bridge (~.4V-.6V forward bias, open circuit reverse bias). If they check out as good, with only the power tubes installed plug in the amplifier, let it warm up for a minute, and check the voltages on all sides of the 390 ohm power supply filter resistors. You may want to use a clip lead from the negative probe of your meter to a chassis ground so that you only need to hold one probe. Do not touch the chassis with you hand while taking measurements. From the tube side to the front panel, here is what I get on the leads of the 390 ohm resistors:
    331 VDC, 3.95 VAC
    327 VDC, 0.4 VAC
    332 VDC, 0.05 VAC

    From your description of the initial problem, I suspected a bad electrolytic capacitor or a bad diode but it is always best to start with the least intrusive methods.

    My voltages may vary from yours as my amplifier is a UK model run off of a step-up transformer but your AC voltages should be similar to mine.

    What voltages do you come up with?

    1. krrunch
      krrunch at · Reply

      Putting an ax7 into the at7 slot definitely lights up and the buzz/hum doubles in volume! I’ve ordered a correct at7 replacement which should arrive next week.

      I’ve located the rectifier diodes I think – tucked along the side of the green 32uF capacitor. When you say to check that all the electrolytic capacitors are discharged, do you mean the large green one and the two somewhat smaller black ones @ 47uF? Or are there more I need to take into consideration? I also need to make a discharging cable but only have 68k resistors so may need to go shopping .. Research time!

  11. krrunch
    krrunch at · Reply

    Right, a little bit of research later:

    After checking the caps with the multimeter, I saw that the DC voltage was hovering between 2 and 5 so no need to go down the route of using a bleed resistor. All 4 diodes are weighing in between 502 and 530 so this is a pass.

    Using a clip for the common wire to the chassis, I checked the voltage around the (in-series) 2 x 390 Ohm resistors, though I’m not sure I’ve done it correctly as my AC results seem to be wildly different(!):

    From the tube side up towards the control panel:

    177.4 VDC, 246 VAC
    174.7 VDC, 243 VAC
    170.5 VDC, 237 VAC

  12. Wyatt
    Wyatt at · Reply

    Okay, so it sounds like the large electrolytic filter capacitors may be bad. Before you order replacements, let’s check a couple of other things. With the amplifier unplugged, measure the do resistance of the primary side of the output transformer. The center tap is tied to the post on the tube side near the 390 ohm resistor and diodes. The transformer has a split winding. Measure between the center tap and where either winding connects to the el84 tubes. I am still concerned one of the diodes may be bad. Did you verify they do not conduct in reverse polarity? I will look at my amp to verify if there is an easy way to check that the negative terminal of the capacitors is connected properly. Did the amp ever work while you had it?

  13. krrunch
    krrunch at · Reply

    The amp did indeed use to work, I remember the day it died about 25 years ago when my teenage world fell apart!

    I have added a photo to the bottom of the article above and marked where (I believe) the centre tap is and the 2 cables from the windings that connect to the tubes.

    Measuring from 1 to 2 (V4) gives me 108 Ohms and a crackle from the speaker.
    Measuring from 1 to 3 (V5) gives me 290 Ohms and a crackle from the speaker.

    I checked the diodes in both directions again to make sure – one way gives between 503 and 530 – the other way yields nothing.

    I’d just like to say thanks again for your time!

  14. Wyatt
    Wyatt at · Reply

    Excellent. It is my pleasure to help. You are correct about the center tap. It is actually two separate windings tied together in series with the middle connection to form the center tap. So I get 254 Ohms from 1 to 2 and 231 Ohms from 1 to 3 on my transformer. With the amp unplugged, pull the power tubes and measure the transformer primary resistance again to verify. Make certain these leads are not in contact with anything else such as an adjacent tube terminal, etc. Also, measure from 1, 2, and 3 to the chassis. Has the o/p tranny ever been rewired or replaced before? It is becoming highly suspect…

  15. krrunch
    krrunch at · Reply

    Hi Jon, I had already pulled the power tubes when taking the measurements yesterday so I put them back in and can confirm the same readings with or without them:

    * point 1 – 2 = 108 Ohms
    * point 1 – 3 = 290 Ohms
    * point 1 to chassis = no conductivity
    * point 2 to chassis = no conductivity
    * point 3 to chassis = no conductivity

    Nothing was ever fixed or replaced on the amp in the time I had it, it must have been only a couple of years old when I got it.

  16. Wyatt
    Wyatt at · Reply

    Great. Assuming the secondary is properly grounded, there is no short between the primary and secondary windings of the output transformer. Measure from point 2 to point 3. If your result is ~400 ohms, plug the amp in without tubes and turn it on. Measure the DC and AC voltages between the center tap of the transformer (point 1) and the chassis ground. Let me know what your results are. We are making some headway 🙂

  17. krrunch
    krrunch at · Reply

    Exciting! Headway sounds good 🙂

    So with the tubes out, point 2 -> 3 measures 394 Ohms 🙂
    (Does this mean that half of the OP transformer is ok?!)

    Measuring the centre tap -> chassis:

    VDC: 414
    VAC: 913

    (Which is exactly the same as the VDC/ADC in the cap next to it)

  18. Wyatt
    Wyatt at · Reply

    Yes, my current belief is that half of the output transformer is comprised. Before you order one, we need to resolve the power supply issues. 913 VAC is … really high. I could see 913 mV but 913 is high. The mains power supply ties to the chassis through 2 of the diodes. Can you verify there is essentially a dead short (0 ohms) between one of the terminals on 2 of the diodes to the chassis?

  19. krrunch
    krrunch at · Reply

    Hmmm OK, just to clarify in case I’ve made a mistake (which I may well have):

    * with my DMM measuring DCV and scale set to 1000, I get 414 – 416
    * with my DMM measuring ACV and scale set to 750, I get 913

    I thought that if the scale is set to 750 then this is the *maximum* it can measure, therefore I’m not sure what that measurement is ..

    Ooops. I just fried my multimeter. With power off, common cable to the chassis, I touched it to one of the diodes, there was a little noise from the meter and now it doesn’t work any more! I’ll have to get another ..

    Pretty sure it was the 2nd one counting up from the centre tap.

  20. Wyatt
    Wyatt at · Reply

    I am sorry you fried your meter. 🙁 It may just need a fuse. I should have instructed you to discharge the electrolytics in the power supply prior to measuring ohms where there would be residual voltage. I will check my amplifier tonight to determine which diode junctions are tied to ground.

  21. krrunch
    krrunch at · Reply

    No worries, it’s all a learning experience 🙂 I’ll pick one up at lunchtime today. What do you think re my measurements though that I get 913 when the scale is set to 750?

  22. Wyatt
    Wyatt at · Reply

    It would be better to start with the scale set to minimum and have the meter say “OL” and work your way up. Is there a decimal point missing? It should be more like .915 VAC as there is no load so the filter caps should charge up and have virtually zero ripple. This is why I wonder about the neg terminal of the cap or lack of the connection to the chassis ground. When I check my diodes from the tube side to the faceplate I have a connection to chassis (~ 0 ohms) on the second terminal of the second diode and the second terminal of the forth diode.

  23. Wyatt
    Wyatt at · Reply

    Another option is to measure the voltages (DC and AC) as referenced to the chassis ground at all of the legs of the diodes. Call them positions 1-8 starting from the tube side. I can check mine to determine which legs are tied together to form the bridge as their physical placement is not indicative of how they are connected to each other schematically. I am confident we will fix your amplifier.

  24. Wyatt
    Wyatt at · Reply

    So my thought about 913 vac is this. It is possible the o/t internally arcing causing a flyback condition that would generate the high voltage. Since we have determined that half of the primary is bad, you could disconnect the two wires that form the center tap of the o/t from the post near the resistors and diodes. If you now get a clean dc measurement and very low ac measurement at the post you removed the center tap from, that would explain the high ac voltage you observed.

  25. krrunch
    krrunch at · Reply

    OK! Armed with my spanky new multimeter I went over everything again.

    The diodes are still good though measuring in the 600s instead of the 500s now. They’re just getting better all the time!

    Measuring the centre tap -> chassis:

    VDC: 414
    VAC: 913 (setting scale to 200 gives overload so looks as though it really is 913v)

    The second terminal of diodes 2 & 4 counting from tubes to faceplate is not quite 0 Ohms, but 3 Ohms and 6.5 Ohms respectively.

    From points 2 to 3, resistance is still 401 Ohms.

    Between centre tap and chassis:

    Diode Voltages:
    Pos ACV DCV
    1 924 417
    2 308 141
    3 308 141
    4 0 0
    5 923 418
    6 311 143
    7 311 143
    8 0 0

    Just for my own understanding (and from my very limited knowledge), if the diodes are arranged in a “diamond” configuration, does that mean there are 2 paths the current can flow through to give DC at the other end?

    Also, if positions 1 and 5 were the diode terminals that are directly connected to the input transformer then I guess this would be normal but the question is then why there is AC measurable in the output transformer?

  26. krrunch
    krrunch at · Reply

    Ah, just saw your comment after I posted mine! I’ll make sure the caps are discharged and then get the soldering iron out .. I’m away for the next couple of days but will update on Sunday, travel time permitting 🙂

    1. Wyatt
      Wyatt at · Reply

      So, there should be a chassis ground lug near the faceplate on the mains side of the pcb. Measure the resistance between the diode legs (4 and 8) and this lug. There should be a wire from the pcb to this lug. 3 and 6.5 ohms is on the high side. It should be much closer to 0. For grins, what do the probes measure when you touch the together? Yes, the diamond creates two paths for the current to be sourced from. This will result in the ripple frequency to be twice the line frequency. I will post a link that de-mystifies the arrangement.

  27. krrunch
    krrunch at · Reply

    Resistance when touching the pins is 0 Ohms 🙂

    With the centre tap removed from the post:

    Measuring the centre tap post -> chassis:

    VDC: 414
    VAC: 918

    so no change :/

    I’ve put a picture at the bottom of the post above – is this the lug you mean? (circled in yellow) It’s sitting above the resistor, beside the top diode.

  28. Wyatt
    Wyatt at · Reply

    I have to check my amp to be sure that is the correct “chassis ground lug”. I believe there is a lug that is screwed to the chassis near the area you pointed out but to the right of the pub, not on the pcb. Something is not adding up. There should not be >900 vac present.

  29. Wyatt
    Wyatt at · Reply

    All righty then, so I checked my amp and the ground lug I was describing is actually on the chassis right next to where the earth from the AC line comes in and is tied to the chassis. There should be a black wire tied to a pin on th pcb just to the upper right of the one where the orange wires tie to the pin you circled. one thing that would be of value would be to measure the ACV of the secondary of the mains. This would be the ac voltage between the two blue wires that connect to the circuit board near the diodes. You could also measure the dc voltage between each of these and the chassis ground lug as described above.

  30. Wyatt
    Wyatt at · Reply

    Prior to taking any measurements could you verify that the AC plug is wired correctly. Make certain earth goes the chassis and the neutral and hot go to their respective places between the plug and the amplifier. Verify by checking continuity. I would hate for us to overlook something fundamental.

  31. Wyatt
    Wyatt at · Reply

    Great. Thanks for the heads up. You may also want to consider some solder wick while you are at it.

  32. krrunch
    krrunch at · Reply

    Hi Jon, sorry, it took me a little time to get back into the swing!

    With no tubes inserted, the orange wire disconnected and taped up, measuring from centre tap to the black wire’s ground lug on the top right of the PCB:

    ACV: 925
    DCV: 419

    Is this to be expected?
    Thanks!

    1. Wyatt Noyes
      Wyatt Noyes at · Reply

      So could you disconnect the transformer center tap wires from the post and measure the voltages from the center tap post to the ground lug?

  33. krrunch
    krrunch at · Reply

    The values from the centre tap to the black ground lug are exactly the same:

    ACV: 925
    DCV: 419

    What I really don’t understand is how I can get a reading of over 600 when 600 is the maximum range for that multimeter :/

    However I’ve tested with the large capacitor and it shoots up to 921 vAC too and then slowly discharges to around 3 so it must be right ..

    1. Wyatt noyes
      Wyatt noyes at · Reply

      So I thought about your readings and suspect there is one or more bad diodes in the bridge. I used a circuit simulator called LT spice and a diode blown as a short circuit will cause the same condition you observe. Sometimes a bad diode will test fine with a meter but exhibit Zener breakdown in reverse bias and look like a short. I suggest either replacing all of the diodes since they are inexpensive or desolder at least one leg of each to measure them out of circuit. Even if they all test good, one may in fact be bad. Be careful de soldering the diodes so that you do not de-laminate the copper foil from the pcb. Use solder wick to remove solder blocking the holes prior to pushing the component legs through to ensure the copper foil is not compromised. 1N4007 diodes should be readily available. If you trim the legs after soldering it is recommended the solder joints be reflowed.

      Let me know your thoughts.

      1. Wyatt Noyes
        Wyatt Noyes at · Reply

        So, let me know what you measure for voltages after you work out the diodes. Then you will be ready for the next steps. I suspect the meter is capable of measuring voltages as high as you measured but the insulation rating of the wires, probes, etc. is 600.

        1. krrunch
          krrunch at · Reply

          That makes sense! Diodes have been ordered and I’ll report back as soon as I have them replaced 🙂

          1. Wyatt
            Wyatt at ·

            Hi, I was wondering if you have made any progress? I am here to help.

Leave a Reply