Tube Rectifier Diode mod

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by hdahs143, Jan 17, 2015.

  1. hdahs143

    hdahs143 Supporting Member

    Messages:
    1,156
    Joined:
    Jul 24, 2007
    Location:
    Virginia Beach, VA
  2. Peteyvee

    Peteyvee Premium Platinum Member

    Messages:
    52,706
    Joined:
    Mar 11, 2010
    Location:
    Just East of Venice Beach
  3. trobbins

    trobbins Member

    Messages:
    311
    Joined:
    Mar 16, 2010
    Location:
    Melbourne, OZ
    I skimmed through that linked forum thread and couldn't identify one poster who knew why the mod was an appropriate mod. The mod is there as a belts-and-braces extra layer of protection to the power transformer should a diode, or diodes, in the valve rectifier start to fail and conduct when it shouldn't be conducting. The ss diode has negligible change to the normal performance of the amp (it adds less than 1V drop to an already large peak drop across a valve diode), and doesn't come in to play with reverse recovery.

    The voltage rating of each added diode needs to cope, so in some larger amps it should be 2x 1N4007 in series.

    If a valve diode does go bad, then hopefully you hear some added hum, and then get the amp serviced, otherwise you may end blowing the first filter cap from higher ripple in the long term.
     
  4. Ampedusa

    Ampedusa Member

    Messages:
    4,933
    Joined:
    Mar 28, 2009
    Location:
    Austin, TX
  5. pdf64

    pdf64 Member

    Messages:
    5,192
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    It seems pretty sensible to eliminate the potential for a common failure mode to subject the B+ winding from fault current.
    Especially given that guitar amps often have no fuse on the B+ and employ a hot switching of a heavily capacitive load arrangement with the standby (thereby exposing the rectifier tube to surge current).
    Note that many current tube rectified Fenders include Si diodes in series with the rectifier plates, presumably due to their inclusion increasing the amp's resilience.
     
  6. trobbins

    trobbins Member

    Messages:
    311
    Joined:
    Mar 16, 2010
    Location:
    Melbourne, OZ
    If the valve diodes are the common arrangement of a dual diode valve, with one diode per winding, then a failing valve is likely to cause both anodes to conduct (eg. gassy failure). As RG's article indicates, the only apparent change is an increase in HT. That may not be noticeable, and may lead to hotter bias conditions on the output valves, and may stress the filter caps.

    If the valve diodes are the less common arrangement of say two dual diode valves, with one valve's diode per winding, then a failing valve is likely to cause only one 'diode' to always conduct. That may possibly be noticeable as a hum from increase fundamental ripple.

    Either way, it may not be easy to recognise a failed valve. A simple LED-resistor across a valve diode would be an indicator, but only of good operation, not of failure. Hmmm.
     
  7. Jackie Treehorn

    Jackie Treehorn Member

    Messages:
    1,829
    Joined:
    Jan 20, 2004
    Location:
    New Orleans, LA
    I lost multiple jj gz34's in my st70 before this mod. Doesn't it split the current capacity between the tube and silicon?

    Changes the sound, though. I don't do it in guitar amps.
     
  8. Diablo1

    Diablo1 Member

    Messages:
    622
    Joined:
    Nov 22, 2013
    It doesn't change the sound to have another diode in series with the tube rectifier - the voltage decrease is less than one volt. All it does is protect the power transformer in the case of a tube rectifier internal short. It doesn't split the current capacity since the solid state and tube diodes are in series. It's a protection device for your amp, and that's a good thing, and cheap.
     
  9. Trout

    Trout Member

    Messages:
    7,576
    Joined:
    Jun 23, 2006
    Location:
    Illinois/Far West Burbia
    It is indeed a good(not perfect) way to protect the rectifier tube, more so than the transformer.

    Some rectifier tubes are a bit more vulnerable than others.
    Tubes with a limited input capacitance like 5Y3's often benefit from this.
    I put the diodes in as standard practice on anything I build just for that extra layer of protection.

    One of the reasoning's is that it helps prevent arcing out the rectifier tube during switching in and out of standby mode.

    This is also somewhat transformer dependent as well.
    I feel it sometimes has to do with the type of standby used, lifting the PT CT being one example where this setup really works well.

    Checking with a storage scope reveals some serious transients generated form the collapsing magnetic field in the power transformer during switching and spikes of over 3KV have been seen on some Hammond and other brand transformers.

    Fender even included avalanche rated diodes in their re-issue amps (Tweed Deluxe & others)

    These are not especially fast switching diodes, they are a softer recovery though.
    Fender had used BYD33 diodes in the 5E3 re-issue.

    Personally, I tested the (hear the difference) aspect on several guitar amp circuits with a diode bypass switch. The ONLY difference I heard was a slightly reduced noise floor.
     
  10. Jackie Treehorn

    Jackie Treehorn Member

    Messages:
    1,829
    Joined:
    Jan 20, 2004
    Location:
    New Orleans, LA
    So you heard my st70 before and after? Maybe you can't hear it in a guitar amp, but it definitely deepened the low end on my hifi amp.

    However, it has increased the longevity of the gz34 significantly.
     
  11. Trout

    Trout Member

    Messages:
    7,576
    Joined:
    Jun 23, 2006
    Location:
    Illinois/Far West Burbia
    Nope,, you never invited me over to hear it ;)

    I never tried it in a HiFi amp, just posting my experiences with guitar amps.
    But I do agree on longer rectifier tube life.
     
  12. Jackie Treehorn

    Jackie Treehorn Member

    Messages:
    1,829
    Joined:
    Jan 20, 2004
    Location:
    New Orleans, LA
    Well, I'd invite you over but the CE replacement can cap has now failed!
     
  13. Trout

    Trout Member

    Messages:
    7,576
    Joined:
    Jun 23, 2006
    Location:
    Illinois/Far West Burbia

    OUCH!!

    Last time I serviced the caps in my McIntosh MC-240 I made phenolic adapter plates and used JJ's. About 6 years in they are still chugging along fine(knock on wood)

    Ever consider getting an aftermarket cap/supply board ?
     
  14. Diablo1

    Diablo1 Member

    Messages:
    622
    Joined:
    Nov 22, 2013
    Do you have a plausible theory why a pair of SS diodes in series with a rectifier tube would change the sound? The purpose of the rectifier is just to charge up the first filter capacitor with DC voltage. The amp runs off the DC in the filter cap, and I can't understand why it should care what happens at the rectifier. The series resistance isn't going to change more than a smidgen, and the voltage on the B+ should be within 1 volt whether you run SS diodes or not.
     
  15. pdf64

    pdf64 Member

    Messages:
    5,192
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    From the above, can we be sure it's the same mod?
    I don't see that it would stop the rectifier tube from failing or reduce the current load on it, though as trobbins points out, it acts to hide the failure?

    Perhaps it is the same mod, and the rectifier tube had completely failed, and the amp has been running on the Si diodes, which should act to reduce B+ sag, thereby helping the amp's bottom end?
     
  16. Diablo1

    Diablo1 Member

    Messages:
    622
    Joined:
    Nov 22, 2013
    Since new production tube rectifiers are such a crapshoot, it makes sense to add the SS diodes to protect the PT. As RG Keen pointed out, if the tube rectifier shorts, that puts AC on the first filter cap, which then also fails by shorting. This puts a big current draw on the PT secondary, and may destroy the PT. And the PT is the most expensive component in the amp. The SS diodes prevent AC from reaching the tube rectifier.
     
  17. trobbins

    trobbins Member

    Messages:
    311
    Joined:
    Mar 16, 2010
    Location:
    Melbourne, OZ
    Using a standby switch in the CT, like the Fender 57 re-issue, certainly could cause a power transformer secondary transient if one of the secondaries was conducting at the time. The worst-case situation would be switching at the peak of the diode current waveform, which can get quite high, especially if the amp is in overdrive at the time, or the standby switch is toggled quickly on then off such that the main filter capacitor is charging up, or if the switch contact bounces at that time.

    Leakage inductance in the secondary winding would force power to continue somehow - probably via arcing of the switch contact, or via the two 5Y3GT anodes in series. Otherwise it will just ring around the diode stray capacitances and discharge out via the primary, and dampen via the secondary DCR.

    I'd be retrofitting a MOV across each HT secondary half-winding !
     
  18. Jackie Treehorn

    Jackie Treehorn Member

    Messages:
    1,829
    Joined:
    Jan 20, 2004
    Location:
    New Orleans, LA
    Yes, I looked into a couple. I've already got the VTA board which has a bunch of filtering/individual bias, etc. I think those other boards struck me as somewhat redundant, bu they may be worth a second look, as another CE cap probably isn't the wisest choice. JJ doesn't seem to have a can with more than 2 sections.

    There is an interesting thread on this forum comparing a copper cap rectifier to a real GZ34. The tube's sag was less with higher loads, so it was not behaving like a simple resistance. Maybe putting diodes in front has an effect on the sag.
     
  19. mark norwine

    mark norwine Member

    Messages:
    16,191
    Joined:
    Jul 25, 2005
    Location:
    NJ...GSP135
    No more calls folks....we have our winner.

    The diodes protect the PT, not the Rec tube.
     
  20. cap47

    cap47 Member

    Messages:
    2,243
    Joined:
    Sep 16, 2009
    Location:
    Bushnell, FL
    I have had this mod on my 5e3 for about 6 months now. I never noticed a difference in tone or volume. Basically I moved the incoming wires to an unused pin and bridged the 1N4007s across to the pins in prior use.
     

Share This Page