what zener to use to drop 30v

[phsyconoodler]

Never done the zener diode voltage drop yet,but I'm needing to do that on a Trainwreck Express build.It has 437v on the plates and I want to shed 30v.
  What wattage of Zener should I use?It has a pair of 6v6's.

[jjasilli]

-30V I don't know, but per Dan Torres "Inside Tube Amps" you want a 50W reverse polarity zener to put in a 5/8" hole.  The manufacturer is NTE

NTE part # 5275AK drops 50V; K = reverse polarity
5279AK drops 60V; so a lower NTE part # may get you the desired -30V.  The actual drop is usually 5 - 10% less than "advertized", so a 50v drop seems acceptable.

[phsyconoodler]

I found a 1n2990B 33v/10 watt chassis mount zener. I'm thinking that would be ok,but not quite sure.If I get 20v drop I'd probably live with that.I could put two in series I guess.

[RicharD]

I wouldn't use a zener.  I'd calculate the current, resistance, ans wattage then slip a power resistor under the PT center tap w/ an appropriate bypass cap.  Is it silly con rectified and can you go to a tube rectifier?  The right PT would be the best solution IMO.

[phsyconoodler]

Could you walk me through that method?If it's that easy lots of people would have done that.

[tubenit]

Not trying to talk you out of using a zener at all .............

I ran JJ 6V6's for yrs at 476volts with no issue. I personally would not hesitate to use JJ's at 437 volt range at all.

With respect, Tubenit

[jjasilli]

I found a 1n2990B 33v/10 watt chassis mount zener. I'm thinking that would be ok,but not quite sure.If I get 20v drop I'd probably live with that.I could put two in series I guess.

Hard to put these in series - they genrate heat and are made to be bolted to the chassis.  So you'd need a non-gounded heat sink for the first rev. zener.  I have no experience, but some people feel that the reverse zeners are prone to failure.  

Alternativley you should be able to drop 30V with a power resistor in the B+ rail.  An inductive wire wound resistor will help simulate a tube rectifier -- it drops voltage and gives some reactance to imitate a tube's impedance.  There was a prior post on this.  

At idle a 30 volt drop @ about 60mA total amp draw = 500R resistor.  But at full signal the amp will draw more like 150mA.  500R X 150mA = a 75 volt drop.  W = V X I = 75 X 150mA = 11.25W X 2 for safety = 25W.  (Note that Torres recommended the 50W zener for big bottle tubes which draw more current.  One would think that a 50W rev. zener would survive in this set-up.  Still there are other options as pointed-out in this thread.  But 10W looks inadequate.  

[Willabe]

You can use a VVR/Power Scaling cir. to drop the voltage. It does'nt have to be adjustable, you can pre-set it for a certain voltage drop.


         Brad       

[RicharD]

I didn't think you were supposed to put a zener in series with the load like you would a resistor to drop voltage.  I thought you were supposed to put them in parallel with the load to regulate voltage.  Dropping 30 using a regulator = heat, and a lot of unnecessary taxation on your PT.  Even dropping 30 volts using a resistor = heat at this sorta current.  IIRC, a pair of 6V6's idles at about 70mA.  30V * 70mA = 2.1W.  By the time you're slamming away and with the addition of preamp tube loads, you're easily drawing 100mA which at 30V = 3W.  You're gonna want to use at least a 10W resistor.  Counting on thumbs, 30V @70mA = 420 dropping resistor.  That's probably gonna drop too much voltage, I'd try a 330 ohm 10W resistor first.  You can put this ahead of the plate tap (simply adding another power supply node) or you can insert it between the PT center tap and ground.  Either place you'll want to add an additional filter cap.  If you select the center tap, remember that ground is now more positive than the center tap so make sure you insert the filter cap the correct way (+ to ground).

Another thought, assuming this amp is silicon full wave rectified would be to series string 15 - 1N4007 on each leg of the full wave rectifier (30 total).  A 1N4007 drops about 1V.

[phsyconoodler]

Thanks for all the replies guys!
  I like the idea of simply using a resistor between the center tap and ground.sounds ridiculously easy to me.What value filter cap should I use to bypass it?Will a 22uf/450v do it?

[RicharD]

Make it the same as your first filter cap.  Remember to pay attention to polarity, + to ground.  Here's a schematic where I used this trick.

[phsyconoodler]

Fabulous! Thank you very much!

[phsyconoodler]

Ok,I had a 1k/25 watt resistor laying around so I cautiously installed it and bypassed it with a 450v cap with + to ground.
  Now i know I only wanted to drop 30v but this effectively dropped the voltage down to 388v from 437v before. The resistor is mounted to the chassis with heat transfer grease and it doesn't seem to be getting hot.The amp is much more controllable now;before it was a wild thing that was almost unusable.
  I'm still not totally sure if that much voltage drop is going to be safe in the long run or should I just wait and get a more suitable sized resistor?The center tap voltage drop thing has me worrying.And you know what worry does to the lines of your face!

[RicharD]

437V-388V=49V
49V/1000R=49mA
49V*49mA=2.4W

I think you're fine for now.

[tubeswell]

I would use 2 x reverse-biased* 5W 15V axial zeners in series off the HT centre tap, (or the ground side of the a FW bridge if that is what your were using. I have done this on several amps and it works wothout a hitch. (Bruce Collins from Mission amps put me onto this trick a few years ago). Lots of people say "5W!, yikes, no way ,man!" but they are perfectly fine** used in this way. Mount the zeners on a terminal strip.

* i.e.: cathode (banded end) pointing to ground

** e.g.: say you are running a pair of 6L6s in PP on a tweed bassman. assuming 200mA maximum current draw on the High Tension winding the maximum dissipation for each 15V zener is: 15V x .2A = 3W, but I bet they don't even get that hot. (the HT current draw is more like 150mA on those amps, = 2.25W per zener  = absolutely fine)

[DummyLoad]

437V-388V=49V
49V/1000R=49mA
49V*49mA=2.4W

I think you're fine for now.

by a 10fold margin..

[phsyconoodler]

well this is a new way(new to me) to do things that seems to be just fine.
  Guess cracking the books once in a while can't hurt!

You guys are the best!

[John]

yet another thread bookmarked!  :smile:

[tonewood]

w/ an appropriate bypass cap. 

My brain must not be working well today. Can you explain why the bypass cap and why + to ground?

Thanks

[RicharD]

Typically we see a series dropping resistor tied to the positive end of the source.  In this case, we're tying the dropping resistor to the negative side.  The source is still making 473volts but we're dropping our 49 volts below ground.  Ground is relative to where we strap it.  Think about a virtual center tap on filaments.  Some older gear used a pot for this so you could swing the ground reference to the 2 supply extremes.  What phsyco is doing is referencing his ground 49 volts above the center tap's potential.  Now the center tap is more negative than the ground reference.  We want to filter ripple from this part of the source so we add a capacitor.  The polarity of a capacitor must be correct, so it's not so much that we're tying the positive to ground, but really more that this cap is in series with the rest of the filter caps.  Yes we are in effect creating a bipolar supply.  It's a pretty damn lopsided supply -49V 0V 388V, but bipolar none the less.  I suppose one could tap right off the center tap of the PT now for a negative bias supply, but I would not recommend it.  This negative voltage is going to fluctuate considerably under varying loads which makes for a less than ideal biasing situation. 

[Fresh_Start]

I know physco has his amp working, but I've got a conceptual problem with using a resistor under the PT center tap.  Buttery hit the nail on the head - we've now got a current-dependent "negative" voltage at the chassis.  The zener alternative with the cathode to the chassis puts the power transformer's center tap at a negative voltage relative to the chassis, but everything else can use the chassis as ground.

I've used a 50 volt zener on the center tap but in a small, single ended amp.  I think it was the NTE part jjasili referred to.

Hope that helps,

Chip

[jjasilli]

There are old threads on the center tap resistor issue.  It's late in NYC; just back from a gig; work in the morning.  I'm burnt out.  Where is the Keeper Of Old Knowledge.   

[RicharD]

>we've now got a current-dependent "negative" voltage at the chassis.

No.  The chassis is still earthed. The PT center tap is now negative.  Exactly how negative is current dependent to a certain extent, but any unregulated DC power supply is current dependent to a certain extent.  I suppose you could do the same thang with a zener.  I've no experience using a zener in this fashion so I must plead the 5th.

[Fresh_Start]

One thing I forgot to mention:  with the Zener approach, you need to ground the first ("reservoir") filter cap with the power transformer's center tap - not on the chassis.  Ground the rest of the power amp as close to the Zener as possible, but NOT with the center tap.

When I used this approach, I originally grounded the entire power amp on the anode of the Zener.  Voltages were weird (don't remember exactly how) and the amp was noisy.  Grounding as described above fixed the problems.

HTH

Chip

[DummyLoad]

it seems to me that the RC network or zener on the CT creates a DC current to flow through the secondary winding. PTs aren't gapped to tolerate a DC currents. to me it seems like the transformer would run much warmer and possibly run into core saturation than with the usual CT grounded method. hopefully someone with more savvy will chime in.

--DL

[Geezer]

I used 5x of these 10v/5W zeners ( http://www.mouser.com/Search/ProductDetail.aspx?R=CZ5347Bvirtualkey61000000virtualkey610-CZ5347B ) to drop ~~50v from an ~~120mA cathode biased circuit.
I also bought some of the 14v & 18v versions, but felt more "comfortable" dropping a smaller voltage per a larger # of units to spread out the heat a bit.

I strung them in series off the CT....they work perfectly & don't even hardly get warm. I had some extra space on the turret board, so they are mounted right there @ the 1st filter cap end of the board.
When I first installed them, I left the ground connection unhooked & used a clip lead to try different points of ground......none made any difference at all in either voltage drop or noise, so I simply grounded them along with everything else in the power amp.

Given the excellent results I've experienced, I will not hesitate in the future to use these (when needed) to bring an over-zealous B+ into line. And @ $0.52 each, IMO they can't be beat!

G

[phsyconoodler]

The resistor/cap idea also works pretty flawlessly.So far so good.
  I amp still in awe of the knowledge of some of you guys!

[jjasilli]

Geezer:  I'm wonering if you notice that zeners have any effect on tone.  Zeners will regulate the power supply.  A regulated power supply in a guitar amp is not what most guys would probably be looking for on this Forum.  But the voltage drop is relatively small, so the amount of regulation may be insignifigant. 

OTOH, a resistor -- whether in the B+ rail or in the CT path (I think) will contibute to dynamic sag, by dropping more voltage with more current draw.  Any comments on that?

[phsyconoodler]

This mp does not sag much at all with the resistor in the center tap,so I think that isn't an issue here.I know it does in the B+ line as I've used those many times.
  I'm curious how long this one will last? If there are any problems with this method or if it's just a really simple,easy and cheap way to do it.So far it does exactly what buttery says it does and I hear no tonal drawbacks.

[jjasilli]

Maybe my question to Geezer is moot:  if there's so much B+ voltage that "gymnastics" are needed to drop it, then the tubes may not get starved of voltage enough to sag anyway.  Still practice is more better than theory!   

[tubeswell]

I used 5x of these 10v/5W zeners ( http://www.mouser.com/Search/ProductDetail.aspx?R=CZ5347Bvirtualkey61000000virtualkey610-CZ5347B ) to drop ~~50v from an ~~120mA cathode biased circuit.
I also bought some of the 14v & 18v versions, but felt more "comfortable" dropping a smaller voltage per a larger # of units to spread out the heat a bit.

I strung them in series off the CT....they work perfectly & don't even hardly get warm. I had some extra space on the turret board, so they are mounted right there @ the 1st filter cap end of the board.
When I first installed them, I left the ground connection unhooked & used a clip lead to try different points of ground......none made any difference at all in either voltage drop or noise, so I simply grounded them along with everything else in the power amp.

Given the excellent results I've experienced, I will not hesitate in the future to use these (when needed) to bring an over-zealous B+ into line. And @ $0.52 each, IMO they can't be beat!

G

That's what I've always been sayin

[J Rindt]

It would be awesome if one of you guys that have graphics/computer skills, could get this info visible. I appreciate how much effort that might take, and I wish I had the knowledge/software to do it myself. But if some of these PS Voltage Dropping techniques could be preserved in schematic/lay-out form, it would be a huge assistance for future builds, builders, discussions.
Thank You

[RicharD]

it seems to me that the RC network or zener on the CT creates a DC current to flow through the secondary winding. PTs aren't gapped to tolerate a DC currents. to me it seems like the transformer would run much warmer and possibly run into core saturation than with the usual CT grounded method. hopefully someone with more savvy will chime in.

--DL

I kinda see this & kinda don't.  Seems like if this was a problem then voltage doubler circuits would be bad for  power transformers.  Obviously I didn't come up with this idea on my own.  I actually remember from where I hijacked it.  (see attached)  My lack of the gift for the written word is preventing me from trying to explain why I don't there there is DC current flowing through the PT secondary, but in a nutshell aren't the diodes and polarized capacitors preventing this?

[RicharD]

Above is a diagram showing a full wave rectifier w/o any filter caps or a reference to ground.  Now imagine Rl as 2 resistors in series.  Next reference ground to the point where the 2 resistors tie together.  This is basically what Phsyco is doing.
http://www.electronics-tutorials.ws/diode/diode_6.html

[Geezer]

It would be awesome if one of you guys that have graphics/computer skills, could get this info visible. I appreciate how much effort that might take, and I wish I had the knowledge/software to do it myself. But if some of these PS Voltage Dropping techniques could be preserved in schematic/lay-out form, it would be a huge assistance for future builds, builders, discussions.
Thank You

This is what I have done.

The zeners can be a combination of 10v, 14v 18v (or whatever value you wish) to acheive the desired voltage drop. The zeners shown give an approximate 50v reduction.

[J Rindt]

Perfect.......
Thank You

[VMS]

Hey guys,

is this the same thing as "back bias" but without the bias part?

http://aikenamps.com/BackBiasing.html

[DummyLoad]

I don't there there is DC current flowing through the PT secondary,

yes there is... your clue is the polarized electrolytic cap across the CT resistor - polarized caps do not like reverse currents. period. they get unhappy quickly.

you have DC flowing through the PT core and this is not good for a transformer that isn't gapped handle it and PTs aren't gapped to handle DC, however, since the power demand is relatively low, it works. this plan (resistor, zener, etc.) will cause some abnormal rise in the core temperature and if the transformer is of marginal spec. in the application, i believe that it can shorten the life of the PT.

respectfully, and still mostly clueless...

--DL

[darryl]

it seems to me that the RC network or zener on the CT creates a DC current to flow through the secondary winding

The zeners or resistors only change the ground reference of the transformer secondary. The rectification is still full-wave, so there is no residual unbalanced DC in the transformer secondary.

[Fresh_Start]

What is the theory/purpose of the cap in parallel with the zener(s) between the CT and "ground"?

Thanks,

Chip

[plexi50]

I have seen threads on doing this for some time now but untill now have not fully understood this. Geezers diagram makes it self explainitory. Pretty cool and useful tool. 

[J Rindt]

I have seen threads on doing this for some time now but untill now have not fully understood this. Geezers diagram makes it self explainitory. Pretty cool and useful tool. 

See Geezer......Told you it would be helpful. Didn't take long, did it......?

[NEjoe]

Well I'm glad everything seems to be solved.  Definitely more than one way to go about making 30 Volts of B+ disappear.

Trainwreck in 6V6......usually 5 triodes at about 1.2mA each for the preamp and PI plus about 45mA times two tubes for a grand total of 96mA....we can call it 100mA for ease of calculation.  However since this is at 400V instead of 200V,....probably 20mA/tube would be more accurate.  Didn't old Fender princetons and harvard's run two 6V6's with a 70mA power transformer?? Anyway dropping 30 Volts time 100mA means shedding about 3 watts off the B+ (probably only 1.5 to 2 Watts in reality).

I suppose the 470 ohm ten watt resistor would probably do the job, but that's current dependent and a little too....uncontrolled...for me.

I would have done it by using an inexpensive MOSFET and three resistors, ( Total cost about ONE dollar.......and 7 bucks for shipping).  The MOSFET I would pick is the STP2NK60Z and costs 72 cents at mouser. I bought ten so the cost was only 54 cents each.  The part is total overkill (rated at 45Watts), but the price is right.

It's just a basic voltage follower (aka MOSFET common drain).

I thought about using great big zener for this type situation and then I got to looking at the prices.  HOLY SH.....  20 + bucks!!!!  

Anyway, for those who don't deal with FETS here's the very simplified overview, there's three pins on a FET (gate, source and drain).  When used as a voltage follower whatever voltage signal is fed to the gate pin will then be followed by the source pin provided that the drain pin is at a higher voltage.  

Virtually no current flows from the gate (perhaps microamps on a bad day) to the source pin or drain pin.

However my analog friends tell me that oscillation could occur unless there is a resistor on the gate to act as a stopper (otherwise I'd be done in two resistors).  Also some people feel you must use zener diodes to protect the FET from overvoltage in Vgs or Vsg, but this part has the protection built in already.

So we have a voltage divider of R1 and R2 taking 30.51 Volts off a 437V signal and feeding 406V to the gate pin.  

If you want to get slick you could throw a potentiometer into the R1 spot and make it variable, but then you have to figure out where to mount the pot.  
For less than 3 watts I'm not sure I would bother with a heat sink for the mosfet unless you already have something lying around.

Oh, I need to make this a jpeg or gif to see it here.

[J Rindt]

Awesome!
Thanks to you as well........

[phsyconoodler]

Nejoe,
   Not being familiar with mosfets,could you draw that into the transformer schematic so we could see exactly where to use this?
It looks like it's in the B+ line to me and not the center tap.Just a voltage regulator right?

[Willabe]

Not being familiar with mosfets,could you draw that into the transformer schematic so we could see exactly where to use this?
It looks like it's in the B+ line to me and not the center tap.Just a voltage regulator right?

Yes, it's a VVR thats set for a voltage with a voltage divider, instead of an adjustable pot. I think it still needs the 12v zener across the GS, just like in the VVR.


             Brad         


                     

[Geezer]

Not being familiar with mosfets,could you draw that into the transformer schematic so we could see exactly where to use this?
It looks like it's in the B+ line to me and not the center tap.Just a voltage regulator right?

Yes, it's a VVR thats set for a voltage with a voltage divider, instead of an adjustable pot. I think it still needs the 12v zener across the GS, just like in the VVR.


             Brad         


                     

Not if you use the STP2NK60Z....it has the zener protection built in>>

http://www.st.com/stonline/books/pdf/docs/10066.pdf

[Willabe]

Ahh, even better than. Here's a drawing. Depending on the mosfet you may or may not need the zener across the GS.
Thanks Geezer for settin me straight.

          Brad        

[VMS]

At the end of this page is an example where the mosfet is on the CT:

http://geofex.com/Article_Folders/mosfet_folly/mosfetfolly.htm

[NEjoe]

OK here's a drawing.  R1, C1 and C2 are just filtering to make a smoother DC, R2 and R3 are just bleeder resistors for the caps.

You could put multiple MOSFETs into a design; but I've read that 12ax7s tend to be pretty constant at 1.1mA per stage so there might not be a point to it.

Just ignore the thing in the upper right, it was part of another circuit.