electronics 101 question.

[seavote]

still not confident with electronics.  can i use 3 12V zener diodes in parrallel in place of a 33V zener diode?

 

[livingnote]

In series, and then you get 36V. If you parallel 12V zeners you get 12V.

 

[seavote]

ok i see how 12V zeners in parrallel will give me 12V, not sure what i was thinking or mixing up. but wont 12V zeners in series also give me 12Vs?  i can get 12Vs from radioshack so i was hoping to avoid waiting till i did another mailorder.

 

[livingnote]

Well, with diodes, think valve. You know, like you have in your bicycle. You have to apply a certain pressure for it to open, and what you see on the other end of the valve is the pressure you applied with your pump MINUS the pressure you need to keep the valve open. Now think three of those valves right after each other. Apply enough pressure for one valve to open, you still don't have enough to open the next. So of course you need enough pressure to keep all three open.

Zener diodes are like a two-way valve that needs the standard valve pressure one way, but can open backwards with a specified pressure too, this is called the "Zener effect". Now you can either suck with your pump (=neg voltage) or you can install the valve backwards, which is what is the fate of the Zener per se.

If you have 3 of these in parallel, all that will happen is that you will need the standard pressure a single one needs to open all 3, but they will let more air (=current) through, depending of course on what resistance (little bag o'sand) you have set after them. If you tie their other ends together and give them a single resistor to ground, what would theoretically happen is that the current distributes itself evenly among all three at a third of what would flow through one if you only used one.

Not so in practice: These little suckers have tolerances, so the one with the least voltage needed of course opens first and thereby effectively sets the voltage to below what the others would need, making them an open circuit while hogging all the current. This is the reason btw why you can't paralleletize diodes as you would resistors, and why you should never tie a string of LEDs in parallel through a common R which is calculated for the current of all of them. Blows the works.

 

[livingnote]

Ah but if you're in a pinch and a little masochistic or whatever...if your standard diode has a voltage drop of .6V, take two Zeners in series and then serialize 15 standard diodes in there with 'em, that'll give you your 33V  8)

Just make sure the standards are cathode to gnd and the zeners cathode to +. And measure that sucker out, could be the standards have .7V.

 

[seavote]

i'm not that desparate  i'll wait till i place an order with mouser. just have wait till i need enough of an order to make the shipping price worth  while. thanks

 

[PRR]

How exactly do you need your 33V?

Two 12V Zeners plus one BJT B-E junction will give 31V.

> wont 12V zeners in series also give me 12Vs?

Steal the batteries from every car on your block.

Three 12V batts in parallel is still 12V.

Three 12V batts in series is 36V.

If lugging lead is too much work, you can prove the concept with 1.5V or 9V batts.

I have 12" logs. If I lay them side by side, I have a 12" thick floor frame. If I stack three up, I have a 36" pile. (Actually this place musta had a lot of 6" logs; the house floor is framed 6" thick, but in the barn they stacked two 6" beams to make a 12" sill.)

 

[seavote]

i wasn't very specific was i? i need to drop 80V down to 34V. -80 to -34V actually. this is the bias supply for the grid of a 6v6 tube in a fender amp.

"Three 12V batts in series is 36V"
  i mis understood how a zener works. i thought a 12V zener would only allow 12V to pass through it. so the first zener in series would drop 80V to 12V. the second,third. add infinite would also only allow 12V to pass through also. i guess you could say i thought there was a cap as to the voltage they would pass.
this is wrong?
and i must admit i had a hard time following the valve analogy. sorry living note

 

[PRR]

> this is the bias supply for the grid of a 6v6 tube in a fender amp.

Well, that's different.

G1 bias is VERY low-current.

G1 bias must track G2 voltage. And since no Fender has regulated G2 supply, you would never regulate (Zener) the G1 supply.

You use a resistor divider. The general impedance around 100K. In this case one resistor must drop (80V-34V)= 46V, the other one drops 34V. It will work if you use a 46K and a 34K; it will probably work just fine with standard-value 47K and 33K. Although since tubes vary and I doubt "34V" is really what you want, I would say a 47K fixed and a 50K variable.

Where is this 80V coming from? A 50VAC bias-tap?

I can't think of any rational way a 6V6 needs -34V in audio.

 

[seavote]

i  belive the 34v does get dropped further along in the circuit. here is a schematic.

http://www.schematicheaven.com/fenderamps/princeton_aa964_schem.pdf

  the amp was modded about 25 years ago and the tech put in a much larger non stock transformer and redid the -V supply using a diode bridge. the -80 is coming off the diode bridge which is located just befor the intensity control. where there is a single diode on the schematic. from what i can remeber the mod sounded great . my brother(his amp) is not interested in bringing this "frankenstein" princeton back to stock. at the moment he would just like to get it working again the way he remembers it. 

 

[PRR]

> get it working again

It's not working?

 

[livingnote]

no prob, the simplest concepts in electronics can be quite deceptively so...don't know about your amp, but basically you have the diode logic backwards: A diode doesn't let the specified voltage through, it lets the entire voltage minus its specified voltage through.

 

[seavote]

heres the tale of the amp. about 25 years ago my brother had a local tech do a "distortion" or boost mod to his stock fender princeton. it worked fine for about 6 -8 months. one day while playing it it "started to smoke". he shut it off and its been in his closet since. recently he heard the two mics and mic pre i made and asks if i could look at his amp. when i opened it here is what i found.
here is the stock layout:


http://www.kbapps.com/audio/schematics/tubeamps/fender/vibroluxreverbaa964layout.html


the 4x  20uf cannister type capacitor had exploded(replaced). the 1W, 1k and 18K resistors off these caps were fried(replaced). on the separate smaller "circuit board" (the - power supply?) the 25uf electrolytic was replaced with a 50uf and the single diode was replaced with a diode bridge. two of the diodes were open  so i replaced the two faulty ones. (hmmmm , just hit me... do all 4 diodes need to be the exact same specs?) the original power transformer was replaced with something 3-4 times the physical size of the stock transformer. the 6v6 tubes both tested bad. i replaced them. and the intensity pot looked fried. replaced that also. after i repaired all i could see "component wise" all the B+voltages to the tubes read a little high but nothing crazy. the 2 junctions, on the - PS diode bridge, where the tranformer is connected to the bridge read 38V and 42V. where is connected to ground it reads 0 , but  -85V coming out of the supply where the schematic calls for -38.
when i tested for a signal i got nothing. used a known working speaker.  im thinking the output transformer  as the connections on the output jack look rock solid. this is where i stand with the amp now. i know some of you amp techs are cringing that i'm not bringing the amp back to stock (or having a pro work on it)but my brother is not looking to sell it . he wants to make music with it. any help or suggestions are appreciated.  i'll get this working yet

 

[PRR]

> 25 years...it "started to smoke"
....capacitor had exploded... resistors off these caps were fried.... two of the diodes were open .... original power transformer was replaced with something 3-4 times the physical size...the 6v6 tubes both tested bad....the intensity pot looked fried.

That is a LOT of damage. It is remotely possible one 6V6 shorted, and all else cascaded from there... but it seems extreme.

And I thought you'd said the PT was replaced 25 years back, but your newest post implies you replaced it? Or am I mis-reading the passage?

> the single diode was replaced with a diode bridge

Why?

And how? You don't just "replace" a 2-wire diode with a 4-wire bridge. And the way this diode is used must be redesigned if you change anything. (It would be an interesting thing to put on a senior EE student's final exams.)

Yes, yes, half-wave rectification sucks. But here the suckage does not matter. Current is low, the cost of filtering is small, and most residual buzz cancels in the push-pull 6V6es.

And the "half-wave sag" is key to getting the correct voltage. We need to sag from 340VAC to 34VDC... "full wave efficiency" is NOT what we want.

> tested for a signal i got nothing. ...thinking the output transformer as the connections on the output jack look rock solid.

An amplifier is a chain of weak links. Just cuz the OT-jack joint is solid is no proof the OT is bad; it could be ANY stage before that.

> -85V coming out of the supply where the schematic calls for -38. ... got nothing

Not surprising. Bias sets the tube part-ON so it will respond to signal. 6V6 with 400V on G2 will cut-OFF if G1 bias is more than 40V negative. You got twice 40V. The 6V6es are cut OFF. If you drove it VERY hard you would get some blits, but the stage before it can't make that much drive.

There may be other problems. But you must get the bias in the 30V-40V zone to have "life" in the 6V6es.

> tech do a "distortion" or boost mod .... fine for about 6 -8 months ... it "started to smoke".

Were you able to identify what the tech did for this "boost mod"?

And obviously I'm thinking: if I put a dragster supercharger on my Honda, it might work for 6 months before blowing its guts all over the pavement. Which is remarkably close to the tale of blown parts you describe.

> cringing that i'm not bringing the amp back to stock

I'm old enough to call that a minor Fender, junk for playing small gigs. I suppose it is very valuable today.

But old man Leo was no fool. Doing it "his way" is not wrong, even if only to get a "working base" to re-apply whatever "boost mod" your brother loved so much.

Wire the bias supply exactly as shown in the Fender drawing!! Use 1N4007 diode and 2W resistor at the 100K point. Remove 6V6es. Power-up. Carefully measure voltage at the Intensity pot. Unless the modded PT is stupidly over-blown, you should get a "good" bias voltage. Testing without power tubes, perhaps -38V. Or actually: 1/12th of what you get at the main B+ node. The tap from main PT HV winding, and the ratio 100K:27K, are critical.

 

[seavote]

>"I'm old enough to call that a minor Fender, junk for playing small gigs. I suppose it is very valuable today. "

would be if it wasnt messed with years ago.

>"I thought you'd said the PT was replaced 25 years back, but your newest post implies you replaced it? Or am I mis-reading the passage? "

yes  its a little unclear the way its written. the transformer was changed years ago.

>"the single diode was replaced with a diode bridge.

Why?

i dont know myself

>"And how? You don't just "replace" a 2-wire diode with a 4-wire bridge.

there are other  wiring changes that i can see. i'll take a good look and try to understand them. i imagine if i dont have the "boost" channel on i can hook up the  - power supply to stock and give the amp a try.
thanks

 

[livingnote]

>And the "half-wave sag" is key to getting the correct voltage. We need to sag from 340VAC to 34VDC... "full wave efficiency" is NOT what we want.

This makes me curious, I'm fiddling with a guitar amp right here, too...I always thought it would be "best" to keep the plate voltage as straight as possible, i.e. install big enough filter caps so there's no residual ripple on the plate. Is that not right in that case, or are we dealing with such a drop that any ripple gets filtered out anyway? (In this setup there is a 33µ/350V, then a 10K R and then a 10µ/350V cap).

 

[PRR]

> I always thought it would be "best" to keep the plate voltage as straight as possible

Different question.

PLATE voltage should generally be "solid". Hi-Fi amps usually aim for pretty-steady. Guitar amps may or may not; there are various opinions about "sag" and also this anomalous result about ripple. (You may be able to read the original paper.)

seavote is asking about the bias supply, the small negative voltage used in deep AB "fixed bias" mode. It should generally track the plate and screen voltages when you move from 110V rooms to 125V rooms (or your 210V-240V variations), so it should come from the same transformer. Some transformers have a low-voltage (often 50VAC) tap for this purpose. Other transformers do not. When you have a common-cathode rectifier and CT winding for your positive plate supply, you could add two low-current diodes to make a nevative supply. But it would come out 300V-500V, and you really want 20V-50V. Yes you could get the negative 400V DC on a 400V cap, and then divide-down with resistors. But 400V caps cost a lot. The cheap trick is to divide the AC, before the capacitor.

> no residual ripple

Plate-filter caps will be a major expense. You always want to use Full Wave rectification above the one Watt level.

The bias supply is typically 50V 1mA, 0.05 Watts. For this amount of power, in context of 25W-100W plate supply, the amount and cost to filter the bias supply is small, even when using half-wave.

For the AA964, 340VAC, 100K, 25uFd, 27K, the output ripple is 0.55V peak-peak or say 0.2V rms. This is further attenuated by the 220K 0.1uFd grid coupling network, and then cancelled by push-pull operation. It is generally not the major buzz-source of the complete amplifier. If it was, a 50uFd 50V cap costs very little more.

Note that the RMS current in the 100K resistor is much higher than you might think, 2.2mA, which makes about 0.4 Watts heat. Half-Watt resistors may be OK for 90-day warranties, but tend to fail. Most resistor failure "won't play" but little harm is done. Loss of output tube bias forces output tubes to MAXimum current, which usually kills these costly tubes and very often smokes the transformer. And this same trick is used on bigger amplifiers. So it would be wise to find a TWO Watt resistor for this chore.

 

[seavote]

didnt want you to think your input was not appreciated and/or i gave up on this project.
i have resigned myself to bringing the amp back to stock.(not very difficult now that i made up my mind to do it and started) salvaged a power transformer from a dynaco st34 that has close(slightly higher)specs.
should have it hooked up in a few days,a week? depends on how busy i am. the power tansformer that was used had a normal B+ current but the low V off the transformer measured 34v. as far as i can make out the rectifiers were used to create the -34V used by the 6v6 and was sent to the 5 heaters wired in series.  now that i have the correct power transformer i can wrap my head around the stock circuit.
thanks again

 

[seavote]

thanks living note, PRR, the amp is up and running as a stock princeton
now the only problem is how my brother is going to get it back.i'll need to test it out for a while