The Dynaco Tube Audio Forum

Dedicated to the restoration and preservation of all original Dynaco tube audio equipment - Customer support for Dynaco VTA tube amp kits, all Tubes4hifi.com products and all Dynakitparts.com products


    tube rectifier diode mod

    Share
    avatar
    ejc

    Posts : 32
    Join date : 2010-10-16
    Age : 68
    Location : Connecticut

    Re: tube rectifier diode mod

    Post by ejc on Sun Dec 04, 2016 2:55 pm

    Thanks
    avatar
    electricist

    Posts : 2
    Join date : 2019-01-12

    I don't see any reason for this modification with diodes - anybody can correct me?

    Post by electricist on Sat Jan 12, 2019 8:36 pm

    I don't see any reason for this modification with diodes.
    1 - that situation on-off-on is happening here 60 times per second, when positive 600 voltage wave on plate/anode replacing with negative 600 volts wave on plate/anode and back. SO, what is the problem if AC voltage disappears and in a few seconds appears again? Still the same situation. Arc? which arc? Arc with 1000v (+500v on capacitors and -500v from transformer) volts will penetrate  3 inches vacuum space inside tube - from plate/anode to cathode? you probably kidding. If arc can appear - why it does not appear 60 times per second?
    And as you of course well know arc can appear when flowing current through inductance is abruptly changing -here is no abrupt change : disappearance voltage in primary winding does not abruptly change current in secondary winding=inductance. Now voltage source in primary disappeared but there is still flowing current through secondary inductance(winding) (which is now a source) and it can cause (through transformation) arc in primary because there  -in primary winding- is abrupt change of current - voltage source disappearance, but not at secondary winding (that happens every time you turn off amp). And that arc in primary (or at least voltage caused by secondary winding if that voltage was not enough to fire arc) caused by still flowing current in secondary will try to support disappearing current here -in primary- in the same direction as before (similar to mechanical inertia).
    2 - So, as we just found out right after AC disappearance there -on one plate/anode of tube- will be still positive voltage because current in secondary  winding=inductance will continue flowing in the same direction (similar to mechanical inertia) - if at D-day moment there was positive half wave. And on other plate/anode of the same tube- there is negative voltage at that moment (with negative half wave) but there will NOT be ANY CURRENT, and that is why no arc is possible on plate/anode with previously negative voltage. But on other plate with positive voltage at this moment will be... - see 1- above.
    3 - when AC appears back again then 2 situations possible:
     I - when arc in primary (or at least voltage caused by secondary winding if that voltage was not enough to fire arc) did not decay yet
    II - when .................. already decayed.
    for each of I and II case there are possible 2 sub-cases:
     a - when new AC voltage in primary is supporting direction of arc in primary (or at least to a voltage caused by secondary winding if that voltage was not enough to fire arc)
     b - ........................................... is opposite to direction of arc in primary (or at least to a voltage caused by secondary winding if that voltage was not enough to fire arc)

    In case of the same direction there should not be any problems - just smoothly rising still existing (though decaying) present current direction in primary - so again, no arc conditions.
    In case of opposite direction - yes, then will be some abruptly change in primary current because new current abruptly compensate present arc/voltage and start flowing in opposite direction as usually. BUT, that abrupt change can not be high because source of present current in primary is flowing current of one of 2 tube's plate/anode in secondary - which is not very powerful itself. So, that spike of current in primary will be not much higher that usual turning-on current spike of cold amp. And at this moment voltage polarity on that plate/anode can reverse and be a little bit higher than usually when turning on amp. BUT AGAIN - THAT HIGHER STARTING NEGATIVE VOLTAGE ON ANODE IS NOT ARC, AND CAN NOT BE CONSIDERED AS ARC!! It will may be 20%, OK may be 30% (i doubt because tube's plate/anode current as source in second winding is just part of total transformer power, and can be just some percents of total power where there are also filament power, another anode/plate current, transformer's self consuming power). And that spike will be very very short. So, instead of 300v AC on tube's plate/anode there will be 400v or 500v AC. No big deal. But of course - not any arc. And on other plate/anode of tube where was negative voltage will appear little bit higher positive
    voltage - also not big deal, just 20-30% higher starting voltage/current.
     And if filament' temperature is still high there is no problem with suddenly again applied anode/plate voltage. If temperature is low it is like amp's cold turning on. But remember - when there is positive voltage at anode/plate then tube works in normal conducting mode. When negative: it does not conduct at all - capacitance's' voltage can not go through tube back to transformer winding without very very very high negative voltage on plate which is not possible.

    And more important - if that mysterious arc in secondary as you say is still possible (i still can't imagine how because voltage should go from capacitance through tube vacuum space to transformer where that arc should appear and 1000v diode will stop it? even if "yes?" - where that negative voltage behind diode can come from? - mistery!), and if it will penetrate 3 inches vacuum space inside tube then diodes' 1000v withstanding voltage is a joke for that voltage, because voltage penetrating 3 inches vacuum is at least 10 times higher than diode's 1000v you calculated based on normal voltage of secondary winding.  Voltage of arcs is very very high, and it will penetrate your diodes and winding's isolation too!

    So, please, tell me where i am wrong.
    Thanks.
    tubes4hifi
    tubes4hifi
    Admin

    Posts : 1526
    Join date : 2008-11-30

    Re: tube rectifier diode mod

    Post by tubes4hifi on Sat Jan 12, 2019 9:09 pm

    Hi electricist,
    nice thinking, but yeah, there are some points that aren't obvious, there is more to it than that.
    First thing is that the tube filaments are already hot, not cold, so rather than presenting a load of maybe 100 ohms, they are now practically a dead short (5v/2a = 2.5 ohms).
    2nd thing is that the capacitors are nearly full charged and so also a low impedance and current surge.
    3rd, even if I'm wrong (along with LOTS of others) once you blow out a rectifier tube or two, then try this circuit, and rarely blow out a rectifier tube, then you'll be convinced.
    The old saying, better safe than sorry !! It's two diodes, about 20c worth, can save a $30 rectifier tube, well worth it !!
    avatar
    electricist

    Posts : 2
    Join date : 2019-01-12

    Re: tube rectifier diode mod

    Post by electricist on Sun Jan 13, 2019 12:59 am

    tubes4hifi wrote:Hi electricist,
    nice thinking, but yeah, there are some points that aren't obvious, there is more to it than that.
    First thing is that the tube filaments are already hot, not cold, so rather than presenting a load of maybe 100 ohms, they are now practically a dead short (5v/2a = 2.5 ohms).
    2nd thing is that the capacitors are nearly full charged and so also a low impedance and current surge.
    3rd, even if I'm wrong (along with LOTS of others) once you blow out a rectifier tube or two, then try this circuit, and rarely blow out a rectifier tube, then you'll be convinced.
    The old saying, better safe than sorry !!   It's two diodes, about 20c worth, can save a $30 rectifier tube, well worth it !!


    Sorry, I may be did not get your general idea, but here are some correction to your note:
    1 - when filament wires are hot they have higher resistance than in cold condition, so they are "not dead short" but opposite. They are -in your term "dead short"- when they are cold, for example may be 1.8 Ohms or even less.
    2 - when capacitors are fully charged the charging current is low, right, but equivalent resistance is high, not low.
    All this (high impedances) means that at AC re-appearing moment AC current was not high at all, was less than at usual cold turning on.
    3 - if rectifier tubes were blown then may be there is other reason than mentioned here. And, by the way, those blown rectifier tubes were new or used? what was exact damage to each blown tube (open filament wire, or what)? Were all they damaged at the same situation - exactly after AC voltage re-appearance? After how many times of disappearance-appearance?
    What is the normal AC voltage in that apartment/house because if regular AC voltage is already higher than normal then it is different story - spike voltage before disappearing could be so high that voltage of transformer's plate/anode winding damaged amp even BEFORE AC re-appearing and for owner it looked like re-appeared AC damaged amp (by the way, that super high spike could be the reason why power station cut the power).
    Remember: that winding has about 2.5-3 times higher voltage than line AC voltage and capacitors probably could not hold voltage higher than regular DC: they started leaking/warming/more leaking/more warming and current became too high for tube and tube was damaged. So, not any arc voltage on tube, but not enough high capacitors working voltage was a reason. But owner decided when tube became bad then capacitor also went bad, though more likely opposite - capacitor went bad first because of high voltage spike, because of not high enough cap's working voltage for high spikes in that area.

    Another question: after Modification which tube was installed - new or used? How many times that tube hold the same kind of AC disappearance-appearance which damaged old tube? How high was AC voltage spike right before disappearing and at AC re-appearinhg in both cases: when damage was done and now after Modification? So, only knowing answers to all above mentioned questions with fully detailed data we will have some statistics and can say what was the real reason of damaging tube and if this Modification really helps. For me now it looks very trivial: real reason is not any arc, but not enough high capacitors working voltage for high spikes in that area. If you replace capacitors with twice higher working voltage than B+, then i believe there will not be such problems. Besides, if your diodes help - why do you need tubes? Remove the rectifier tubes and leave only those diodes - it will be fine and less power consuming, higher B+, higher output power - paradise for all parts and never damaged rectifier tubes again. Though if I am right, sooner or later your diodes will be damaged too.






    avatar
    rjpjnk

    Posts : 120
    Join date : 2018-07-18

    Re: tube rectifier diode mod

    Post by rjpjnk on Mon Jan 14, 2019 2:01 pm

    I may be wrong, but here is my thinking.

    The specified peak reverse voltage of a high quality vintage 5AR4 is 1500v,

    https://drtube.com/datasheets/5ar4-amperex1958.pdf

    but I've heard it said that the new copies by Chinese and Sovetek might have much lover PIV limits before they self destruct.

    It seems that in normal operation the tube will have  PIV equal to the difference between the B+ DC in the filter caps and the negative peaks in the transformer secondary. So this is about 800v right? I guess the thinking is that some of the newer tubes even 800v might be unhealthy.

    The addition of the diodes removes the negative from the plates so the tubes don't see a high peak inverse voltage.

    Yes, the diodes alone would work just fine. The tube is mostly there to slow the rise time of the B+.
    deepee99
    deepee99

    Posts : 2076
    Join date : 2012-05-23
    Location : Wallace, Idaho

    Re: tube rectifier diode mod

    Post by deepee99 on Mon Jan 14, 2019 8:00 pm

    Just break down and buy a genuine, NOS Mullard GZ-33. Your problems will be over. Did everything I could to ruin mine; nothing worked.
    avatar
    rjpjnk

    Posts : 120
    Join date : 2018-07-18

    Re: tube rectifier diode mod

    Post by rjpjnk on Tue Jan 15, 2019 1:15 pm

    Is the GZ-33 a drop-in substitution for 5AR4?
    deepee99
    deepee99

    Posts : 2076
    Join date : 2012-05-23
    Location : Wallace, Idaho

    Re: tube rectifier diode mod

    Post by deepee99 on Tue Jan 15, 2019 1:21 pm

    rjpjnk wrote:Is the GZ-33 a drop-in substitution for 5AR4?
    http://antiqueradios.com/forums/viewtopic.php?f=13&t=276422
    Peter W.
    Peter W.

    Posts : 1220
    Join date : 2016-08-07
    Location : Melrose Park, PA

    Re: tube rectifier diode mod

    Post by Peter W. on Tue Jan 15, 2019 3:19 pm

    I have been sitting on my fingers, and I will continue to do so. But the nature of filaments, hot and cold, clearly needs to be better understood.
    Tubes4ever
    Tubes4ever

    Posts : 128
    Join date : 2015-07-14
    Location : Star, Idaho

    Re: tube rectifier diode mod

    Post by Tubes4ever on Wed Jan 16, 2019 8:47 pm

    rjpjnk wrote:Is the GZ-33 a drop-in substitution for 5AR4?

    I discovered that it is. I have an ST70. I substituted a NOS GZ-33 in place of my 5AR4 and I didn't even need to rebias the tubes! That's how close the voltage drops of both tubes are. The GZ-33 is definitely a much more robust tube and looks much better!
    deepee99
    deepee99

    Posts : 2076
    Join date : 2012-05-23
    Location : Wallace, Idaho

    Re: tube rectifier diode mod

    Post by deepee99 on Wed Jan 16, 2019 10:48 pm

    Tubes4ever wrote:
    rjpjnk wrote:Is the GZ-33 a drop-in substitution for 5AR4?

    I discovered that it is.  I have an ST70.  I substituted a NOS GZ-33 in place of my 5AR4 and I didn't even need to rebias the tubes! That's how close the voltage drops of both tubes are.  The GZ-33 is definitely a much more robust tube and looks much better!
    It's the "studliest" tube for audio applications since the Great War! And for our purposes, the pin-out's the same as a 5AR4. Light a pair of genuine GZ-33s up in a dark room and pretty and otherwise reserved women will disrobe at the sight.

    Sponsored content

    Re: tube rectifier diode mod

    Post by Sponsored content


      Current date/time is Sat Mar 23, 2019 3:36 am