5Y3GT data sheet?

[bluefoxicy]

I'm having trouble understanding the 5Y3GT. I am looking into using a recto tube in the next amp I build, as a matter of behavior; aiming for a classic rock or blues sound, so I'm mimicking a Fender amp. It's likely I'll try to use a DPDT to switch to some Zener diodes with an appropriate voltage drop to emulate the tube, but have solid-state behavior (i.e. fast response, less sag).

What kind of capacitor bank should I use for this? The data sheet I'm seeing looks as follows:

http://www.nj7p.org/Tube4.php?tube=5Y3GT

I want to drive a handful of 12AX7 preamps (I'm thinking 5, maybe just 3 though) and possibly up to 4 tubes in push-pull configuration -- two EL84 and two 6V6GT. The power tubes will probably suck down 100mA per pair, which fits within spec but only for one pair (if I decide to run the EL84 and 6V6 at the same time I'll blow the 5Y3), so I'm thinking I might need to run multiple rectifiers in parallel to handle possible 200mA draws?

Probably want to use a choke after the first cap, seems to be more reliable at higher current draws....

[Enzo]

Can you re-explain that? Are you planning to switch between the actual tube recto and a zener/solid state rectifier stage any time you like? Or are you trying to use a tube but get solid state rectification performance?

Eliminating the sag pretty much makes it pointless to use a tube rectifier, I would think.

If you want to toggle between the two, you could switch voltage reduction in and out at the same time. But if you do that, you might try without and see how the higher voltage sounds. If you want sagless, then maybe higher voltage works for your sound.

Probably want to use a choke after the first cap, seems to be more reliable at higher current draws....

If you are getting that from the curves and specs, look at it again. The higher current ratings are for choke INPUT filters, not chokes after the first filter cap. Note that from the same AC into the recto, choke input filters have substantially lower output voltage.

[Tubenit]

5V4 rectifer

I don't think with the current draw you are considering that a 5Y3 would be a good choice.

You might consider a 5V4 rectifier? I have three amps which use a 5Y3, 5V4, and GZ34 respectively.

5Y3 = 1.1x VAC(divided by 2)
125ma 60v (5Y3GT/20v)

5V4 = 1.2x
175ma 25v

5U4 = 1.2x
225ma 44v

GZ34 = 1.3x
250ma 30v

SS = 1.4x

Guesstimated.

With respect, Tubenit

[bluefoxicy]

I don't think with the current draw you are considering that a 5Y3 would be a good choice.

You might consider a 5V4 rectifier? I have three amps which use a 5Y3, 5V4, and GZ34 respectively.

The 5Y3 responds to sudden voltage demands in 0.44 seconds, so produces significant signal compression; the 5U4 responds faster, and the GZ34 even faster. Some recto tubes respond in 0.02 seconds, with solid state responding in about 0.01.

http://www.guitaramplifierblueprinti...ectifiers.html

It would not be a crime to fall around 0.3-0.4.

The Fender reference design I'm looking at uses a 5Y3, single stage pre-amp, and a 6V6 power amp. (I don't understand the input stage; why the arrows, and why not have the 68k resistors as ONE resistor just before the tube?)

http://www.ampwares.com/ffg/schem/pr..._5f2_schem.gif

Amazing blues and rock tone:

http://www.youtube.com/watch?v=sNV_I...eature=related

I believe I want some softer cathode power caps than I use in my current amp (which uses 1000uF for the cathode cap on the 6V6, nice and stiff), and lighter filter caps to make the amp more responsive (sag). I want a tube that will act like the 5Y3.

5Y3GT:


5U4G:


These seem acceptably close, although the 5Y3 will have a 15% slower response than the 5U4.

GZ34:


Silicon:


By contrast, a GZ34 will respond like solid state diode rectifiers.

A 5U4GB gives -50 volts @ 275 mA; a set of zenner diodes with 50V drop would allow me to add a switch to change to solid state rectification. This would sound similar to a GZ34; why use the GZ34 at all then?

I suppose it would not be a crime to go with a 5U4 based design and try to get 275mA out, which should be more than enough to drive the power and preamp tubes. I'll use 1Nxxxx diodes for the filaments if at all (more likely, twist the wires together and just use AC).

[Tubenit]

rectifier tubes

You know one option you have would be to try different tubes to see what you like best? Rather than go by the math, ...... try an audio test. Better yet, have a friend change the recto tubes out not telling you what he/she put in ....... and then decide what captures your tone.

I do think the 5Y3 will not handle the current you're referring to. So try one of the others.

I play blues and I tried different rectifiers in my amps and ended up with a different tube preference in each amp.


The purpose with the two 68k resistors is that one actually give you 33k paralleled with the other one.

Hope you find the tone you want. Rectifier tubes are lower on my list of what I think contributes to the tone I'm looking for.

With respect, Tubenit

[bluefoxicy]

You know one option you have would be to try different tubes to see what you like best?

Changes the voltage output etc.

I do think the 5Y3 will not handle the current you're referring to. So try one of the others.

And running two such tubes in parallel won't help? I was under the assumption that putting two resistors (motors, light bulbs, whatever) in parallel creates a current divider....

The purpose with the two 68k resistors is that one actually give you 33k paralleled with the other one.

Except the two inputs should be isolated, besides that I don't understand the arrows... are those diodes or what? Is that 1M thing supposed to be a pot?

[Tubenit]

rectifiers

http://www.schematicheaven.com/fende..._5d8_schem.pdf

This has two 5Y3's with 6L6's. You can try that.

On the input jacks.
One arrow is a ground the other represents a shorting jack.
http://www.el34world.com/charts/fenderservice2.htm

Regarding the "blues tone". You have individuals like BB King and Albert King who got an amazing blues tone with solid state amps (Gibson and Ampeg). Blues runs such a gamet of tones from Ronnie Earl, to Johnny Winter, to Tab Benoit, to Larry Carlton, to Robben Ford to Sonny Landreth ............... etc that I don't tend to associate a particular amp or rectifier tube with blues.

Since your amp doesn't seem like a copy of any particular amp given you're thinking of running EL84's and 6V6's .... you may need to just experiment with different rectifiers to capture "your" tone. Have fun with it.

With respect, Tubenit

[bluefoxicy]

Nice. Damn that schematic is weird. Is there a book or a site where they keep diagrams and explain wtf parts of them do? Multiple preamps, that stuff after the second 12AY7 has to be a tone stack... how does that standby switch work?

[Enzo]

Those arrows on the input jacks represent the cutout contacts on those jacks. For example the arrow on Jack 1 is grounded, so that is the contact that grounds the tip input when nothing is plugged in. The arrow contact on jack 2 puts the two resistors in parallel so ther is a net 34k in series with your pickup., but only when using jack 1.

But look at what happens in jack 2. Plug in there and the arrow contact opens. ANd that means there is now a 68k resistor in series with your input. But the jack 1 is still grounded by its arrow contact. SO its 68k resistor is grounded at the jack end. Together the two 68k resistors form a voltage divider which knocks the signal voltage in half. That is a 6db reduction in level. And that is how the two jacks get their different sensitivity.

You COULD just wire in one 34k resistor, but then ther would be no point in having two jacks. Look at the two jack system as a 6db pad. Nowdays they call it active and passive inputs.

Why are they arrows? Because those contacts are really a switch. Look at the power switch. Note its symbol also uses an arrow for the switch element.


Speaking of switch, the standby switch works by breaking the current path through the transformer. There are still several hundred volts between the recto plates, but there is no path to ground, so no current flows. You could break the path between the transformer and the plates but that would require a two pole switch.

[bluefoxicy]

Speaking of switch, the standby switch works by breaking the current path through the transformer. There are still several hundred volts between the recto plates, but there is no path to ground, so no current flows.

I've been googling this a lot since last night and having no luck; I'd like to take some notes and throw them in a Wiki. Do you think you could explain the single 5Y3 rectifier (I'll assume that using two just does the same, but with two tubes), including how and why that standby switch works (especially with the ground path through the tubes which I assume exists*); and what the top coil on the transformer there does?

All kinds of other shit is grounded. I'm not sure on this, if I tape off the center tap and fuse the two high voltage leads together does it pass no current? Also that tap at the top is 5V@2A for filament/cathode, interesting but how does this make any sense electrically? Does the 32uF of filtering effectively smooth out the paltry 5V of ripple caused?


Do tubes act as a variable resistor controlled by grid voltage? I'm told tubes work by having a grid block flow of electrons from negative cathode to positive anode when the grid becomes negative, and allow flow when the grid becomes positive. This indicates to me that the tubes act as a resistor, with higher resistance with a more negative voltage on the grid and lower resistance with a more positive voltage on the grid (we call voltage variations "signal").

This means that the recto output connects to ground... how does this not matter? For that matter, how does the 5VAC filament coil not present a good DC flow path for the DC coming off the recto, at least half of the time?

[Enzo]

Do you think you could explain the single 5Y3 rectifier (I'll assume that using two just does the same, but with two tubes), including how and why that standby switch works (especially with the ground path through the tubes which I assume exists*); and what the top coil on the transformer there does?

I will be happy to try, but may I suggest you get a copy of the RCA tube manual? Reprints of the RC-30 (the last edition they made) are available from many sources. I am sure it is scanned online somewhere too, but really a book in hand is a lot easier to browse. Not only does it have the tube specs, but also tutorial chapters on all manner of tube technology. It explains how tubes work, and how tube circuits work. I don't mean this unkindly at all, but I think it is far more important to know the fundamentals before concerning yourself with the esoterica and minutia of tube circuits. FOr example trying to work with tube curves without knowing how the tubes work.

And there are other fine books, but that RCA is one I use every day, and I have worn out numerous copies in my lifetime of soldering. That is worth the $30 or whatever they are asking.


Yes, you can ignore the second tube for discussion, it is just in parallel with the other for more current capability. And for now let's skip HOW the tube rectifies, it is just a diode after all. You can replace the tube with a pair of diodes, as I believe you are aware. I know the voltage would rise, but that is irrelevant to how it works. The tube might as well be two diodes with their cathodes tied together.

The transformer is center tapped, and the center tap is normally grounded.

Let me digress: we use the term ground to mean the earth ground in the power wiring in the wall, but what we have inside the amp is really not a ground so much as a COMMON. I just happens we usually connect the circuit common to ground, so we call it that. I mention this because the importance of the center tap going to ground is not so it is "grounded," but more because it completes a circuit path to the common for the current in the circuit. Without a complete circuit, no curent flows. We could make an amp that ran on batteries, and nothing in it would be connected to earth ground, but it still would require a common in the circuit.

ANyway, grounding the center tap really results in two AC windings that happen to be out of phase with each other. The two ends of the winding have AC on them in equal amounts, but when one goes positive, the other is going negative. Since we want positive from this supply, essentially the two halves of the winding take turns spitting positive pulses at us.

The rectifier allows current through one way but not the other, so when the AC cycle goes positive, then a positive pulse appears on the cathode of the diode - or tube, both have cathodes. When the AC goes negative, the diode blocks it.

(Yes, I know which way electrons flow. I am a conventional current guy - the arrows point the right way for me.)

So if you think about it, what we have here are a pair of single rectifiers - half wave rectifiers - operating into the same supply, taking turns. SO let us look at one side as if it were indeed a half wave circuit - just like the bias circuit in a Fender.

Current from the transformer flows through the diode, into the load (resistor, light bulb, motor, amplifier circuit, whatever), through the load to common, and back to the other side of the transformer. That is a complete circuit. If I replaced the diode with a wire, I would have AC in my load instead of DC. The tube heaters adn pilot light are straight AC. Imagine if I disconnected one side of the transformer winding to the heaters - they would go out. Having the diode in there changes it to DC, but doesn't change that disconnecting one side of the transformer would turn the circuit off.

And that is what we have here. It doesn;t matter which side of the transformer you disconnect, any more than it matters which terminal of a battery you disconnect, either way it turns the circuit off. SOme power supplies put the switch on the top, between rectifier and transformer. That disconnects the top wire, but disconnecting the bottom wire works just as well.

But here we have the two halves of the winding - one half for each rectifier - the center tap is shared by both circuits, so disconnecting the center tap interupts the current and turns the circuit off to both halves with a single switch. We could do it Marshall style and put the switch on the ends of the transformer winding - between transformer and diode - that is OK too, but now you need a dual switch.

Personally I'd prefer the Marshall approach since it removes the AC from the rectifier tube socket, but electrically it doesn;t matter.

All kinds of other shit is grounded. I'm not sure on this, if I tape off the center tap and fuse the two high voltage leads together does it pass no current?

NO no, remember those two places are at exactly opposite phase - opposite polarity. The 120VAC in the mains has about a 170v peak. So when one end of the winding is at +170v, the other end is at -170v. That is 340V across the winding 120 times a second. COnnecting the ends together would be a dead short across the transformer. COnsider the tube heaters again. You can ground their center tap or not. But if you connected the two ends of the winding together, you'd be blowing fuses. Having a center tap doesn't change that.

SO shorting the ends together will make MAXIMUM current flow through the transformer. But what it WON'T do is send any current through the rest of the amp. ANy electrons - any current - that wanted to leave the transformer would need a return path, and ther is not one without that center tap grounded.

You still would have high voltage between the plates of the rectifier tube, but there is bo longer a complate circuit to common.

Also that tap at the top is 5V@2A for filament/cathode, interesting but how does this make any sense electrically? Does the 32uF of filtering effectively smooth out the paltry 5V of ripple caused?

For the rectifier tube to work it must have its heater powered. That 5VAC is used for that. That 5VAC is isolated from ground - well really all windings are isolated from ground until you connect them to it.

Imagine we have 300VDC in a circuit with respect to ground. Now I take a 9 volt battery and connect its negative terminal to the +300v. If I measure the voltage at the positive terminal of the battery, with the meter grounded, I will see +309VDC there. Yet if I measure the voltage across the battery itself, I still get 9v. In fact I could connect the battery to a 9v light bulb or even a littel portable radio and it would run just fine off that 9v battery even though it was still connected to the +300v. At least as long as the radio never touched any other part of the circuit. The 9v circuit is floating on the +300v. Just like birds can sit on high voltage power lines harmlessly. In both examples, there is no path to ground for the high voltage. at least not through the bird, the transistor radio, or the 5Y3.

The cathode is the heater in some rectifier tubes, and in others the cathode is separate but connected to the heater. When the circuit is on, assuming 300v again, there will be +300v on any point of the heater and the heater winding WITH RESPECT TO GROUND, but there will only be 5VAC across the heater itself.

The 5VAC does not cause 5v of ripple. To add that ripple, the 5VAC would have to add and subtract its voltage to our +300v. But that would require a current path in series with the 300v. We don't have that. Imagine a shelf 300mm off the floor. Ther is an object on the shelf, so it is also 300mm off the floor. Now grasp the thing and shake it side to side 5mm. It is moving 5mm, but it stays at a constant 300mm off the floor. This is how the 5VAC thing works. Within itself it has 5VAC across it, but it stays at whateve the DC level is in the circuit. The 32uf needs only smooth out the pulsing DC from the rectifier

Do tubes act as a variable resistor controlled by grid voltage? I'm told tubes work by having a grid block flow of electrons from negative cathode to positive anode when the grid becomes negative, and allow flow when the grid becomes positive. This indicates to me that the tubes act as a resistor, with higher resistance with a more negative voltage on the grid and lower resistance with a more positive voltage on the grid (we call voltage variations "signal").

Well sorta. Without a grid, a tube is a diode, like the rectifier, and it will conduct as much current as it possible. Either the transformer will run out of current, or the rectifier will melt down if you push things. Add a grid, and the more negative it gets, the harder it is for the electrons to pass it. It is not as simple as negative/off and positive on. It takes a lot of voltage for the grid to "cutoff" the tube current. Zero volts on the grid and it essentially disappears, then the mor negative you make it, the less current flows until cutoff is reached. The grid does not have to go positive to allow current to flow. In fact if the grid goes positive it starts gathering electrons itself, acting like a little plate. Ther is no current flowing through a grid normally, but make it positive and it starts conducting current. We don't allow that in our amps much. SO I make the distiction between the grid going positive and the grid going "more positive" than some negative voltage it was at a moment earlier.

I always imagined a triode like a rubber hose with water flowing. Left alone water flows full force. If I then start squeezing the hose with my fingers - the grid - I can control the flow of water. Ideally for largest signal headroom, I would start with it half squeezed - well, squeezed for half flow.

This means that the recto output connects to ground... how does this not matter? For that matter, how does the 5VAC filament coil not present a good DC flow path for the DC coming off the recto, at least half of the time?

How did we make that assumption? There is no connection to ground in the recto tube. What you see on the schematic is the whole story. The output of the rectifier - its cathode - connect to the hot side of the load, NOT ground. It indeed matters a lot. There is a current path, a circuit to ground, but that is through the load, or through the transformer. The transformer being the source of the current.

And the 5VAC winding? How would it offer a path for current? WHere would that current go? Yes, it sits at +300 or +390 or whatever the circuit uses WITH RESPECT TO GROUND, but there is no path to ground THROUGH it.

I hope this helps. If I have made anything unclear, please ask.

[Steve Conner]

The 5Y3 responds to sudden voltage demands in 0.44 seconds, so produces significant signal compression; the 5U4 responds faster, and the GZ34 even faster. Some recto tubes respond in 0.02 seconds, with solid state responding in about 0.01.

Where on earth did you get these numbers from? The concept of a rectifier tube taking 0.44 seconds to respond is absurd, when you consider that they happily chop up current at 120 cycles per second. Imagine Pete Townshend doing a big windmill arm chord and nothing comes out of the speakers till half a second later. Rock and roll could never have happened.

You probably mean that 0.44 seconds is the RC time constant of the power supply filter, which is a function of all the components in it: transformer winding resistance, rectifier tube effective "resistance", and filter capacitance.

If you like to look at pretty glowing things, fit a tube rectifier. If you're not bothered, fit silicon diodes, and use series resistors and appropriately sized filter caps to get the "response speed" you want.

Turning your computer off and cracking some books wouldn't hurt either, as others have suggested. (For our younger readers, a book is like a wiki except nobody can edit it, and you can read it in the tub without people IMing you. And these are both advantages.)

[bluefoxicy]

Where on earth did you get these numbers from? The concept of a rectifier tube taking 0.44 seconds to respond is absurd

The 5Y3 responds to sudden voltage demands in 0.44 seconds, so produces significant signal compression[...]

http://www.guitaramplifierblueprinti...ectifiers.html

[...]

5Y3GT:


[...]
Silicon:

QED

[bluefoxicy]

Thanks Enzo! I think I get it now....

• The current WILL flow from one tap on the tranny to the other, but because of the diode (rectifier) in the way, the direction of flow is inevitably stopped at one side or another; therefor it needs to reach the center tap (which is hooked to the common/ground).
• Current coming from the 5VAC filament leads only wants to get back to the opposite lead of the 5VAC coil, not to ground or wherever else. It won't flow through the circuit.
• In either case, current doesn't magically run to ground; both ends of the power source have to effectively reach ground (and thus, each other) for that to work.

One of the guys I work with says I should study electronics instead of jacking around inside hot circuits. Probably right. But you've been way helpful. I'll have to add this to my notes some time this weekend :3

[Wakculloch]

I should like to suggest that these curves show that an amp without smoothing caps would sound boggin.