Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups darrenriley.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com

JTM45 Sounds beautiful then Main Fuse blows!

Discussion in 'Amp Tech Center' started by koochems, Feb 19, 2018.

  1. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    Hi All, I'm new to the forum and have found it extremely valuable when building new gear! Thank you for that!

    Looking for some guidance/recommendations:
    I've completed my JTM45 build coupled with a dual greenbacks cab and am completely flummoxed at this point. I've been playing the amp at lower volumes consistently throughout the past 2-3 months since it has been built and the amp was singing! I felt confident enough to gig it. While trying to push it at performance levels with my accompanying pedals, the main fuse blew. This occurred when I was running a Zvex Mastotron which draws ~1mA. So naturally I replaced the main fuse 3A and tried again. Amp was singing beautifully (15 minutes at performance levels) until I ran the masty again. It sounded perfectly fine for ~3-4 minutes then the mains blew again. My measurements are fairly accurate to the metroamp layout so I'm not sure what the next step is. At this point, I bought another set of matched KT66's but am hesitant to replace the fuse and try again for fear of further damage to PT. HT fuse is in tact, fixed bias is around 45-48mA; there is some oxidation happening on power tube pins but nothing is out of the ordinary, and just to reiterate when the amp was on it sounded wonderful with no strange hisses, buzzes, cracks, pops, etc... Hoping to find some recommendations from the group if anyone has found themselves in a similar situation. I can send anymore information... otherwise, to the amp repair shop I go. JTM45 Guts.JPG JTM45 SB1.JPG Thanks in advance for any replies!
     
  2. Wally

    Wally Telefied Ad Free Member

    Mar 17, 2003
    Lubbock, TX
    Welcome. You don’t say what your plate voltage was at that 48ma of current draw. If it was around 450’ you are about 86% of max plate dissipation...kind of hot but it should be safe. Anytime an amp sounds great and then takes out a fuse one starts wondering about the output, ime. I would run the amp in a darkened room while watching the plates..and while the amp alone is being pushed. Redplates??? If not, then kick that. Zvex thing online and watch things related to tube plates some more.....

    That is a sharp looking amp you have there.
     
    Mr Ridesglide likes this.
  3. dan40

    dan40 Tele-Holic Silver Supporter

    943
    Aug 19, 2015
    Richmond Va
    Also look over every joint to be sure that there are no stray wire strands or anything else that could cause a short. It's likely one of the power tubes but I would still examine everything closely.
     
  4. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    I'm not mad... I'm just disappointed in myself! Thanks for the response all! I was shorting pins 2 (PT heater) & 3 (OT primary) on the V5 output tube. Of course something this simple would reveal itself when I reach out to the forum for the first time. I cleaned up the solder joint and gave them plenty of clearance. Remeasured all pins and everything checks out. I hope this solves the issue, and hope that it didn't lead to any other unforeseen damage. Thanks again.
    Pins2-3.JPG
     
  5. Wally

    Wally Telefied Ad Free Member

    Mar 17, 2003
    Lubbock, TX
  6. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    To avoid a hum loop? I see... This is my second DIY build, still getting acclimated to workmanship standards. Now that its open I'm gonna re-solder to reflect this. thanks for the information.
     
  7. Wally

    Wally Telefied Ad Free Member

    Mar 17, 2003
    Lubbock, TX
    I don't like the proximity of the heater filament wiring to other pins. I would suggest studying that area around the shorting very carefully to see what is happening where. Also, it is advisable to thoroughly clean all of the carbon tracing that we see there....that carbon residue can be conductive.
     
  8. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    soldered extensions to the heater wires and routed around the power tube sockets. Thanks again for the guidance! Will clean up the inside and give 'er another go! compareheaterwires.JPG
     
    Wally likes this.
  9. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    Hoping for more wise guidance as something is still not right with my JTM45! I've retested voltages again and again everything reads as expected (voltage chart attached with expected results in yellow), my B+ is ~465vdc, and biased 39.7mV (V5) and 35.7mV (V4). Added modifications such as, 5.6K swamp resistors in series with pins 5 on power tube sockets, replaced 470K/5W screen resistors with 1K/5W screen resistors. The amp sounds great but when pushed the KT66 power tube in V5 starts a light show, flashing. This happens when pedals are on, DL4 and overdrive. Clearly pushing the amp this hard was enough current to blow the high tension fuse. I'm thinking about buying new power tube sockets and replacing. Also, I have a new set of genalex KT66's but do not want to plug them into the amp as it stands to avoid further damage. Any thoughts? voltageChart-Actuals.JPG
     
  10. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    Your screen voltage is really high relative to plate voltage, and you've got a really odd voltage difference between pins 4 and 6. There should be a couple of volts of drop, and you've got a rise. Also, that screen voltage matches your B+ voltage which is another possible red flag.

    What voltage do you show at each power supply node, and what schematic are you built to?
     
  11. peteb

    peteb Friend of Leo's

    Apr 25, 2003
    Cascadia
    Good observations Clint.


    Also


    The 53v drop from the B+ to the plate, 465-412=53 volts is large for a bassman type of amp.


    Using ohms law and your numbers for current, the resistance of the OT primaries for v4 and v5 would be 1,485 and 1,335 ohms respectively. This too would be high for a bassman type of amp. Have you measured the resistance of the OT primary's and do you know what they ought to be?


    The screen resistors on the schematic are 470 not 470k. I take it you replaced 470 with 1K?
     
  12. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    Thanks for the reply! I remeasured... again this morning due to your notes. I'm afraid the last voltage chart may have had inexplicable typos! Sorry for the confusion! I did replace the 470ohm with a 1K. The voltage drop between the screen and plate is ~1-2VDC, which I think is normal. The plate measurements also drop a couple VDC from the B+. I used the schematic attached. I measured between the CT located on a filter cap and the plate voltage to get the OT resistance. Also, took this measurement with the amp off. These measurements seem normal? Replace tube sockets?
    voltageChart-10MAR2018.JPG
    marshalljtm45 schematic.JPG
     
  13. Snfoilhat

    Snfoilhat Tele-Holic

    Age:
    37
    673
    Apr 8, 2016
    Oakland, CA
    Heater voltages appear to be sagged more than you'd expect. Which indicates either too much current draw or the wrong transformer turns ratio or wiring.

    And the B+ is higher even than the high AC wall voltage would account for, if that spreadsheet's 'expected' values are accurate.

    Are you using the 5AR4/GZ34 rectifier tube that the kit calls for, or some replacement?
     
  14. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    Yep, using 5AR4, 2 KT66's and 3 12AX7's. Also using the magnetic components 40-18033 PT and respective 40-18039 OT. I took all of those readings today, they should be as accurate as my DMM.
     
  15. peteb

    peteb Friend of Leo's

    Apr 25, 2003
    Cascadia
    Does 'light show' mean the screen?


    The measurements are giving a clearer picture of what is going on in the tubes, they are showing high screen current in V5.

    The measurements could be better, for the voltage drop across the OT you can measure it directly from plate to center tap and get a little more accurate than a whole number, i.e. 3 VDC and 2VDC. Also, since you have dedicated screen resistors you can easily measure screen current for each power tube, measure the voltage drop across each screen resistor and use ohms law.


    So far we have this

    V4:

    Plate current 26 mA
    Cathode current 28.4 mA
    Screen current 2.4 mA (inferred)

    V5:

    Plate current 19 mA
    Cathode current 34.6 mA
    Screen current 15.6 mA (inferred)


    Calcs for plate current:


    V4:

    I = V/R = (465-462)/115.6= 0.026 amps

    V5:

    I = V/R = (465-463)/105.5= 0.019 amps
     
  16. peteb

    peteb Friend of Leo's

    Apr 25, 2003
    Cascadia
    Expected screen current

    V4:

    The screen current is 8.5% of the cathode current*

    V5:

    The screen current is 45% of the cathode current**

    * good
    ** too high


    The expected range is 3 or 4 to 15, 16 or maybe a little higher, maybe 20, but not 45.
     
  17. koochems

    koochems TDPRI Member

    16
    Jan 19, 2016
    denver
    Yes, I believe it is the screen grid that's pulsing a brighter red when the amp is pushed - it is does not appear to be redplating, unless I'm confused about this definition. When I actuate pedals (i.e. DL4 delay draws more current than most on my board) the screen grid gets hotter and when pushed hard enough a white flash sparks inside the tube. This happened a couple times before eventually the HT fuse blew.
    My DMM resolution is 1 when measuring higher VDC, hence the whole number measurements. I'm going to borrow a more accurate and higher quality/resolution fluke from work tomorrow. This way I can get better numbers to calculate screen current. Per your observations, my plate current needs to be higher on V5 in order to lower the screen current?
     
  18. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    Now, I'm reading this as a roughly 2 to 3V drop from pin 6 to 4 on each power tube, across a 1k resistor. That's only a couple of mA across each one per Ohm's Law. That's not bad. Directly measuring the voltage drop across each resistor, along with accurate measurements of each resistor's value, will give you a better number.

    Likewise, using the transformer voltage drop method or a bias probe will give you the actual plate current for each tube.

    That screen flashover is an interesting sign. Something is going very awry when those tubes are driven hard. It could be that one tube has already suffered partial screen meltdown during the original problem, when the fuse blew.

    If it's not an inconvenience, would you mind getting a full set of readings from this set at idle, then swap to a new set and repeat?

    Last request: what OT are you using, with what speaker load and secondary winding? Using your numbers, I'd like to verify the power amp design is at least valid against the datasheet. Wouldn't be the first time a known design proved to be unstable in real life.
     
  19. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    Oh, quick question. Have you tried simply swapping the two current power tubes to see if it follows the tube or stays with the socket?
     
  20. peteb

    peteb Friend of Leo's

    Apr 25, 2003
    Cascadia
    You can measure the DC voltage drop across the OT and across the screen resistors with the low voltage setting because even though the voltages are high, the voltage drops are not they are just a few volts. Use the low voltage setting like you did on the cathode and get at least one decimal point.


    The cathode current and plate current are factors but the situation will reveal itself better when you measure the screen current. You won't need a Fluke to see this.



    Right, the cathode current is composed of the combined screen and plate currents. If the screen current gets more normal and is reduced it will show up as more plate current.

    At this point I would say whatever the plate is doing is less important than what the screen is doing. Get the screen under control first.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.