rectifying line volts

[shooter]

I was looking at the 12ax7's datasheet and they show a 90 plate example, is there a down-side to using rectified line vs R'in down the HT line?
thx

[silverfox]

If you are asking what I think you are asking? Which is: Do you have to use a power transformer in a design? No you don't. It was common in designs of the past- 50's and prior on many consumer electronics, to use a line voltage without a power transformer.

Very unsafe however due to the fact there is no isolation between the line source. A potential lethal shock hazard.

Not sure if this is what you're referring to.

Silverfox.

[darryl]

Using line voltage ( or "mains" voltage as it is called here in Australia ) without transformer isolation is potentially lethal, and I'm certain it is now illegal to do so in most countries.

[PRR]

> line voltage ...without transformer isolation is potentially lethal, and I'm certain it is now illegal to do so in most countries.

Not illegal, strictly.

However you must meet the Leakage Current requirements for anything the user can touch.

And under modern leakage codes, the transformer is the ONLY practical way. (Either 50/60Hz transformer, or now-a-days a switcher with 20KHz transformer.)

(Under older code, a very-low current stage "could" function yet not leak "excessively". This was always a bad idea, and went out of style in the 1960s.)

> using rectified line

Plastic-case radio "can" run transformerless. Input is through radio-waves, no external antenna contact (or isolated with a teeny high-voltage cap). Output is air-waves from a speaker behind a grille. Controls are all plastic knobs. Ideally you can NOT couch any conductor (metal). Such radios were common in the US and Canada 1930s into the 1970s. Similar models existed in Europe (adapted to ~~200V B+.)

The 1st-audio stage in a radio is a 12AX7-like tube and yes it may run only 90V supply. The power stage runs on maybe 105V. Power tube is selected to pass big current (40mA) at 105V, and need only about 15V grid drive which a 90V-power driver can just about do.

There's also battery-power radios. Most used a different set of tubes with filamentary cathodes which use less battery-power to heat. 90V B+ was a fairly common value for battery loudspeaker radios.

If you HAVE 300V handy, there's about no reason to be thinking about 90V power.

[HotBluePlates]

I was looking at the 12ax7's datasheet and they show a 90 plate example, is there a down-side to using rectified line ...

Most of the time, something in the chassis needs more than 120vac * 1.414 = 169vdc.

[silverfox]

Coincidentally I found this over at RickTone.com as I was surfing tonight. How to make safe is in the diagram. The schematic isn't showing up in the preview so not sure if it will in the post; Here goes!!

Silverfox.

[jazbo8]

This says it all - death cap, from the schematic:

[shooter]

Thanks, I will use an isolation transformer, my primary PT will be like 800vdc and I didn't wanna  put in a "heater" circuit to get the volts down for the driver-pre circuit.

[HotBluePlates]

Thanks, I will use an isolation transformer, my primary PT will be like 800vdc and I didn't wanna  put in a "heater" circuit to get the volts down for the driver-pre circuit.

Whatcha building that needs 800vdc?

[shooter]

Whatcha building that needs 800vdc?

it'll be a 5E'ish build running a ktt88 close to the melting point - 800vdc plate max.  I was gonna ask this in a new thread, how do builders base their amps power rating? being an electronics guy I tell the guitar guys what I measure for PA plate volts X the PA current.  They all says, "It seems like more power than that" anyway, thanks

[HotBluePlates]

Whatcha building that needs 800vdc?

it'll be a 5E'ish build running a ktt88 close to the melting point - 800vdc plate max. 

... my primary PT will be like 800vdc and I didn't wanna  put in a "heater" circuit to get the volts down for the driver-pre circuit.

The Kt88's are gonna need ~300vdc or less for the screens. Absolute maximum screen voltage is 600vdc, but this high screen voltage causes the plate current to be way too high with that high of plate voltage.

You will find that most amps either supply the plates 600v and the screens ~300v, or they run the output stage ultalinear at closer to 500vdc (plates and screens are nearly same voltage in UL mode).

[PRR]

> PA plate volts X the PA current

Idle current or full-roar current?

IAC.... there's NO reason to run KT88 at 800V in audio. 800V on G2 may kill them. It can certainly unleash MORE than enough current to melt the tubes.

Remember voltage is NOT the only way to power. High-current makes power too.

Paralleling is another path.

http://www.mif.pg.gda.pl/homepages/frank/vs.html

Take the M-O Valve sheet. Page 3 shows P-P fix bias tetrode. 560V supply in 4.5K load will give 100 Watts output. Vg2 is just 300V. (This is easily gotten with a 440V CT AC winding and a bridge rectifier. This also has 300V for your little bottles.)

Yes it is possible to squeeze a bit more, but is it safe? Wise? Economical? Already a quad of 6L6GC or EL34 will do 100W with similar or lower tube costs. 150W suggests a quad of KT88/6550 *or* a sextet of 6L6GC/EL34. Availability is better with the more-common tubes. Reliability may be better with more tubes (one of a pair fails, the gig is over; one of a sextet fails, nobody notices).

[shooter]

 but is it safe? Wise? Economical?

  that's why i'm askin questions.  My thought process is to develop the largest ac I can without distortion in the PA stage.  I was thinking the "bigger" the plate volts, the "bigger" signal I can produce without clipping.  I don't have the datasheet here but I think the max peak g1 signal is 16v peak, guessing that is 32ish p-p.  The last kt I built was at the datasheet typ class a,  about 400 plate, 84ma idle but I could only drive the g1 to about 16vp-p before the tube started distorting the output.  That was loud enough for me but....... 

[HotBluePlates]

... I think the max peak g1 signal is ... 

Maximum G1 input before distortion is solely about how you have the tube biased. So if it distorted with inputs greater than 16v, you had maybe 17v of bias or less.

...  The last kt I built was at the datasheet typ class a,  about 400 plate, 84ma idle but I could only drive the g1 to about 16vp-p before the tube started distorting the output.  ...

A KT88 is rated for 42w, and class A amps are typically biased to 100% dissipation at idle (you could idle cooler, but you just get less power output). 42w / 400v = 105mA

Had you idled the tube hotter, you would have used an even smaller bias voltage than you did, so distortion would have began with a smaller signal.

That sounds like a problem, but you'd also be getting more power output with less driving voltage so really your amp would have had better performance. I dunno why 16v peak is a magic number, because if you're getting distortion you simply turn down the volume and apply less signal.

My thought process is to develop the largest ac I can without distortion in the PA stage.  I was thinking the "bigger" the plate volts, the "bigger" signal I can produce without clipping.

Remember voltage is NOT the only way to power. High-current makes power too.

You listen to a speaker. The speaker is a low-impedance coil of wire in a magnetic field; current is what moves it.

If you have 40w applied to an 8Ω speaker, then √(40w*8Ω) = 17.9v is across the speaker impedance. If you have 500-600v of signal swing at your tube plate, the output transformer still has to step it down to ~18v to deliver that 40w. The bigger the voltage step-down ratio, the more likely you will run into limitations in your transformer.

17.9v / 8Ω = 2.24A

Power = voltage * current
~40w = 17.9v * 2.24A  (some rounding error)

If you have an 800v supply in your amp to allow the KT88 to have a ~510v RMS signal swing, then 40w / 510v = 78mA of current swing. High voltage times small current or small voltage times big current. Either way delivers the same power.

And distortion isn't predicted by deciding high or low supply voltage (loading changes everything).

[shooter]

Thanks hotbluesplate, that helps a lot, I don't know why I keep ignoring ohm's law when I "look"" at the ac signal.   

suggests to me that I have a dual set of electric chairs, and the state I live in outlawed capital punishment.
 I have the HV supply for an x-ray tube, 140,000vdc at .1A  I just haven't figured out how to get the x-ray tube to operate in the audio range!

[PRR]

> the "bigger" the plate volts, the "bigger" signal I can produce

You ever do any arc-welding?

The arc voltage is always 40 Volts. When you want "MORE!!", a BIG arc, you dial-up the *current*. 10 Amps will tack beer-cans. 70 Amps for a Mustang inner fender. Tractor axle, dial-up 250 Amps. Steel-ship welders run in the thousands of amps. All at 40 Volts (pretty near).

In tube amps we can take a wide range of voltage and current, then transform it to speaker level.

[shooter]

The only thing I know about welding is why they call it a "stick welder" every time I tried to draw a bead the rod stuck!!!  I will stick to soldering n let my master welding brother weld!

I think i'll keep the volts to the 400-500 range, tweak the bias for best un-distorted signal.  but 1st I need my trannies to complete the el84 build.  Once that is up n workin then I have gain no#'s and power no#'s I can use to set up the kt88 since it'll be the "right" channel of a pair of mono blocks.  I'll let the guitar guys decide what side they like best!

[shooter]

It is indeed!  with the anode spinnin at like 10,000rpm so she won't melt!