'88 Jubilee 2554, Why Plate Voltage Only 422v @ 70% Dissip.

Discussion in 'The Workbench' started by perkabrod, Nov 14, 2017 at 1:22 PM.

  1. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    I've just replaced all the pots, and reversed the mods that my 2554 had.
    I had a 2555 (4xEL34) before and plate voltage was around 460 V when biased at 36 mA.

    This combo 2554 I own has plate voltage in the 420-430 V range. It has 2xEL34 and a 100(?) ohm power resistor instead of a choke. The PT is the one typically used in European/Scandinavia JCM800/Jubilees etc, namely Drake 789-62, no center tap, hence the bridge rect. (That very same PT was on my 100W 2555 as well)

    What can cause low plate voltages, tube gain?
    Should I be worried about these plate voltages around the 420-430 V mark? At 70% I set bias to 40 mA.
    [​IMG]
    [​IMG]


    I have verified the screen and grid resistors being correct value. I have not tried with other tubes, do not know if the tube makes any difference, they're JJ tubes.

    (off topic, but here's a link showing the mods it had when I bought it and how they sound)
     
    Last edited: Nov 14, 2017 at 1:27 PM
  2. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    422 is low.
    Make sure nothing in the power supply is dropping AC from the wall including mains fuse/breaker.
    What is your measured wall VAC?
     
  3. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    Hmm okay, so 422 V is low you say :(

    Just caught up on some of Uncle Dougs videos and (re)learned that the OT influence the plate voltage of the output tubes.

    There's really not much in the path from PT to screen grid and from OT to anode of EL34.

    Mains (wall outlet) measures 233 V_RMS.
    I have another 2554 but it's waiting for tubes, unfortunately not the same transformers but I'll see what plate voltages I get with it...

    Given the mains voltage is absolutely normal, what's next?
    Should I check directly on the HV taps on the PT secondary side, standby as well as ON = play mode

    suggestions?
    Please post your plate voltages if you have 80's 2553/2554/2550/2558
     
  4. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    You can first check the power supply as mentioned. Measure voltages at the anode and cathodes of the power supply rectifier circuit.

    I have seen as high as 460 and as low as 430 volts on the plates but on average normally about 450. Power or mains voltage from the wall though can fluctuate from about 115 to 122 VAC with average 118 to 120.

    So when measuring things always note the wall voltage because it makes a difference.
     
  5. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    Mains voltage while amp powered up 231 VAC rms

    GND to first filter cap - is that what is called B+?
    425 VDC in playing position

    Heater voltage
    6.2 VAC rms

    PT secondary, between the two blue wires going into rectifier.
    333.3 VAC rms. OCD approved :)
    (No center tap on drake 793-62)



    The 100 ohm power resistor (instead of choke, 2553 and 2554 had resistors) measured 98.4 ohm. But the Vdrop from filter cap + side across the resistor was 0.6 VDC, is it realistic, just thought it would be higher...

    The diode (anode) in the bias circuit measured -43 VDC, referenced to chassis ground



    EDIT:
    Forgot to say that pin 3 on EL34 (anode) measured 421.5 VDC referenced to chassis ground.

    Also dug up this post by another one with very likely the same power transformer (the user who posted is also from Sweden)

    http://forum.metropoulos.net/viewtopic.php?f=4&t=19397&start=30#p275001
     
    Last edited: Nov 15, 2017 at 5:24 PM
  6. johan.b

    johan.b Well-Known Member

    Joined:
    Feb 21, 2015
    Messages:
    837
    Likes Received:
    466
    Location:
    Södertälje, Sweden
    333 volt AC should give you 466 volt rectified. How much ripple do you have on the first filter cap /B+. It's typically around 7 volt on that era Marshalls. .perhaps one of the diodes stopped conducting?...
    j
     
  7. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    333.3 VAC should give middle 440's and not 425 so something is dropping voltage a bit. Also the PT HT is a bit on the low tolerance side giving 333.3 VAC to the rectifier since your area is based on 230 mains at the wall.

    It could be the tubes but may also be the rectifying diodes or power supply capacitors causing issue.

    Again, measurments before and after each rectifying diode could give clues.
     
  8. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    That is to efficient of a value. I would not expect anything that high with 333.
     
  9. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    Appreciated your replies. I will make more measurements today. Can I pull the output tubes and safely have the amp in playing position?

    Also, when measuring rectifier, is it these 4 measurements I should make? see my sketch on picture from sluckeyamps
    Measure VDC or VAC
    [​IMG]
     
  10. ampmadscientist

    ampmadscientist Well-Known Member

    Joined:
    Jun 26, 2014
    Messages:
    5,231
    Likes Received:
    1,876
    Location:
    On the wrong side of the tracks.
    When you UN-modified this amp, there was some error.

    Go back in reverse order, and re-check every step you did.

    (possible broken circuit tracks?)

    OR...
    The normal voltage for an unmodified amp "is:" 420V. You never had "normal" until you UN-modified it.

    There is nothing wrong.

    Low gain: if you UN-modified it, you could expect the gain to be lower?

    But:
    Adjusting bias hotter causes the plate voltage to drop (this is normal.)
    Adjusting bias colder causes plate voltage to rise. This is also normal.

    At any point:
    Did you adjust the bias hotter - causing the plate voltage to drop?
    (drop to where it is now?)
     
    Last edited: Nov 16, 2017 at 9:23 AM
  11. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    Well yes and no.
    You can measure both AC and DC on both sides of each rectifier from chassis ground.
    But that is a bridge rectifier shown for a JCM900 whereas the 2554 has a regular full wave rectifier.
     
  12. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    I actually believe that is just the way my amp is set up if you study the photo in the first post. Europe (and possibly Canada) used these transformers for jcm 800, 2550 etc, I don't see any center tap, also two of the diodes are bridged with a jumper, and a cap in the rectifier. To me, it is just like the attached sketch. That's why I'm not sure where/how(AC/DC) and in reference to what.

    Forgot who posted this question:
    Did you adjust the bias hotter - causing the plate voltage to drop?
    (drop to where it is now?)

    — I sadly did not measure plate voltage prior to "restoring", though I have had a 2555 head in the past that was at around 460 V, 36 mA when 70% dissipation.

    The amp in this thread was at 430 V, 25 mA after I restored it. I dialed it in to 70% dissipation and ended up at ~420 V and 40 mA
     
  13. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    Here's the measurements.

    AC ripple on first filter cap 1.8 VAC rms. DC voltage around 423 V

    I swapped to some other EL34 tubes that I have in my 2550, some minor difference in voltage and plate current but still in the 420/430 region. Though I was surpriser to discover that these EL34 caused a 50 Hz hum in play mode, constant hum loudness no matter the output master volume...

    Now I know tubes can produce hum! Swapped back to the JJ EL34

    I did not measure voltsgr between individual diodes, though I did remember to use the diode checker on all 4 rectifier diodes, every diode tested around 0.55 V drop. To me that seem normal.

    I think I need to fire up my 2550 and see what values I have on it, the only difference is that it has a choke instead of a 100 ohm resistor.

    Comments?
    Starting to suspect everything is normal
     
  14. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    Well a power transformer can be out as much as 20% but other things can pull the voltage down. That is what we are trying to determine.
    460 x .8 = 368 but that would kind of suck.

    I see a full wave rectifier in the picture of your first post just like every other 2554. The black wire looks like a center tap off of the transformer to me.

    By the way, biasing hotter will not drop the voltage that low.
     
  15. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    I estimate that the bias pot range makes a difference of 10-15 V on plate voltage, went from ~20 mA to ~40 mA last time I biased it.

    About the center tap, not sure I follow, I thought mine didn't have it. Here's a schematic of the rectifier in my amp, not like the 2550 psu circuits online but just like every other amp stamped SEM on the white sticker.

    [​IMG]

    Here's a photo I put together from browsing others jubilee photos. My amp in the bottom and a US amp on top (maybe with a center tap)? Don't mind the white and red wire near R23, it's just hanging there while I was replacing pots

    [​IMG]
     
    Last edited: Nov 17, 2017 at 12:29 AM
  16. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    16,433
    Likes Received:
    5,659
    Location:
    US of A
    Can you provide a closeup of the rectifier circuit on your board? It is organized/laid out the same but maybe has different traces.

    The PT transformer has a black wire hitting chassis ground in your picture. It should be HT center or heater center or both. Give a close up picture of that as well.
     
  17. johan.b

    johan.b Well-Known Member

    Joined:
    Feb 21, 2015
    Messages:
    837
    Likes Received:
    466
    Location:
    Södertälje, Sweden
    Scandinavian (and some other European markets) amps don't have the Centre tap. That's why you see the modified bias circuit ( with X class 0.047 cap). I don't know why people refuse to accept this. It allows the use of a single ~330 volt winding instead of a ~330-0-330 (660volt) winding.
    My guess it's a "Semko" thing ( Swedish elektric safety..similar to CSA in Canada or UL in USA), but that's just a guess.. Scandinavian amps also had the autofuse I see in your amp.
    The black wire in the picture is Centre tap for heaters. If you zoom in you see the heavy wire winding on the tap.
    J
     
  18. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    Black wire going to L1 is grounded at 1 st filter cap/chassis bolt

    [​IMG]
     
  19. perkabrod

    perkabrod Member

    Joined:
    Apr 18, 2012
    Messages:
    41
    Likes Received:
    1
    Yes I have only seen this type. Living in Sweden. This amp says SEMKO but my other 2554 and 2550 and also the 2555 I had in the past said just SEM. Looks the same to me though.

    Will get around to measure my other amps for reference this weekend. But this 2554 just seem to be on the low side on plate voltage compared to others. Just extra suspicious because it had been modified...

    EDIT:
    330 VAC rms * sqrt(2) = 467 V theoretically max B+

    Different output tubes didn't make a difference. 1st filter cap ripple 1.8 VAC.
    Diode drop approx 0.55 V per diode. Where does the rest of the voltage go in my amp, to end up on around 425 V. Have not measured across the .047 uF cap and resistor R23, maybe there's some 20'ish V drop there?
     
    Last edited: Nov 17, 2017 at 2:08 AM
  20. RickyLee

    RickyLee Well-Known Member VIP Member

    Joined:
    May 22, 2010
    Messages:
    8,091
    Likes Received:
    2,055
    Location:
    Riverside CA U.S.A.
    I am figuring your power transformer is expecting 240VAC and then your wall voltage is closer to 230VAC. Your 6.2VAC heaters is a bit lower than all my Jubilee amps, where mine are usually 6.7VAC. My U.S. wall voltage is 123VAC - 125VAC usually. And then my plate voltage runs 460VDC - 470VDC on my '87 Silver 2555 and my '89 2550.

    The schematic for the 100W 2555 amps shows the bias supply voltage has its own dedicated winding on the PT secondary. Then the schematic for the 50W models 2550, 2553, 2554 and 2558 have their bias supply voltage taken off or fed/supplied from one leg of the PT secondary HT supply. I have seen this latter bias circuit pull down the power tubes plate voltage, and then that could vary even more with your power tubes transconductance, or how much control grid negative voltage (bias supply voltage to EL34's pins 5) your EL34's need to bias up properly.
     

Share This Page