Plate 800v / G2 300v for GU50 PP 120W - Which PI and how arranged ?

[kagliostro]

I and a friend are planning a 120W GU50 all purpose amp

the datasheet followed say

Plate 800v
G2 300v
Plate at idle 2 x 50mA
Plate at max 2 x 130mA
G2 polarisation 2 x 0.9mA
G2 at max 2 x 13mA
Bias 2 x -51v
Load 8k
Input signal for max output 2 x 36v AC

I'm in trouble on identifying an adeguate PI for this amp

LTPI or Concertina ??

and which values for components ?

the GU50 is a deaf tube, a 6L6 has a sensibility of 6mA/V, the GU50 only 3-5.5mA/V (yes unfortunately not all are at 5.5mA/V)

Can we plan to plagiarize the PI of an existing amp, which one ?

Thanks

K

[jazbo8]

That's not that difficult a load to drive, any LPT used for the larger tubes like KT88, 6550, etc. can do the job.

[kagliostro]

Thanks Jazbo8

..... any LPT used for the larger tubes like KT88, 6550, etc. can do the job.

Also if KT88 and 6550 has a sensibility of ~11.5mA/V and the GU50 only 3mA/V till 5.5mA/V ?

Franco

[shooter]

Input signal for max output 2 x 36v AC

I just finished a build that was driving 2 EL34 at about 32Vrms, I'm currently using KT88's instead without any affect on the grid signal.  I'm using a USA-1 tranny as my PI.  I also did a build using a cathodyne? PI and was able to get about 28Vrms to the grids of 6L6s, both builds were cathode biased about 20vdc.  Have yet to build a LTPI.

has a sensibility of ~11.5mA/V

 This spec is new to me, is it plate current vs grid drive?

[PRR]

When in doubt, copy Bassman 5F6a.

[DummyLoad]

k, have you sourced the 8K 120W OT? just curios?      


--pete

[kagliostro]

@ Shooter

Thanks for the info

This spec is new to me, is it plate current vs grid drive?

11.5mA/V = 11500µMho the transconductance of the tube

@ PRR

as always Thanks

@ DummyLoad

My friend wraps by himself, entirely by hand, the transformers
(he also get beautiful calluses on hands  )

Here an example of PT and OT for a 4 x 6L6 amp he did





Franco

[Lindsay]

Looking forward to the result here Franco. Are those specifications calculated or from experience. It doesn't seem like much bias volts considering my SE build at low voltage was -46V. In an experiment to see two valves in PP perform  in makeshift chassis with plate at 700V we needed much more bias.
 
Lindsay

[DummyLoad]

k, thanks! that's some nice iron... post progress, please.


--pete

[kagliostro]

At the moment I can post only a preview of the piece of metal that will become the chassis



Ciao

Franco

[PRR]

> It doesn't seem like much bias volts considering my SE build at low voltage was -46V. In an experiment to see two valves in PP perform  in makeshift chassis with plate at 700V we needed much more bias.

G1 voltages are more relative to G2 voltages, not plate voltages.

When you go to crazy-high plate voltage you do not need (or want) high G2 voltage. G2 voltage sets maximum current. With way-high plate voltage you do not need much current to make a lot of power. Generally, for many tubes, as you go over 400V-450V you should *reduce* G2 voltage. Look at the 100W conditions for 6550 and 8417: Vp 600V, Vg2 300V. 6146 can be worked at Vp=600V or 750V with Vg2 at 200V or 150V!

GU-50 test condition is Vp=800V, Vg2=250V, G1 bias of -25V to -55V to get 50mA plate current. K's -51V is reasonable, though at Vg2=300V he may need a few more volts to hold the most excitable GU-50s down to 50mA.

[kagliostro]

I Apologize Lindsay, I forgot to give you an answer 

---

The data are borrowed from the EL152 datasheet, you know that is the civil version of the LS50

and the spec of this datasheet are used by Wavebourn and many other without problems

seems that russian military tubes on the datasheet often are indicated to have a lower G2 voltage than it is

I think this is done for very long performance use

Franco

[Lindsay]

Thanks PBR and Franco for the explanation. I might post my build details in another thread.

[HotBluePlates]

This spec is new to me, is it plate current vs grid drive?

11.5mA/V = 11500µMho the transconductance of the tube

Side-topic:

Ohm's Law arranged to calculate resistance: Ohms = Volts/Current

"Transconductance" on American data sheets is defined (in the front-matter of the tube manual) as "change in plate current for a change of grid voltage." That's Current/Volts, to take the words and put them in math form. That's also the form of Ohm's Law quoted above but flipped upside down. So they also wrote "Ohm" backwards in the old days, and called the unit "Mho."

1 Mho = 1 Ampere/1 Volt but we might get a tube plate to pass 1mA more for a 1v change of grid voltage (in a preamp tube), so transconductance is 1 milli-mho. Some tubes had even less current change for 1v of grid voltage change, so micro-mho seemed to be the norm. Our 1 milli-mho tube above then can be called a 1000 micro-mho tube.

But the European system conveys more useful information if you didn't already know all of the above. How do you use it?

Look at Franco's original post: data says 130mA at max output with a 8kΩ load, which implies 130mA² * 8kΩ = 135w (assuming everything is RMS; overage above 120w goal is probably distortion and some other factors). The sheet says 36v RMS for that 130mA, so you need/have a tube with 130mA/36v = ~3.6mA/v or 3600 micromhos in the American naming system. Assuming we didn't know the driving voltage from the data sheet, we could note what Franco says that the GU-50 runs from 3-5.5 mA/v and know that for our needed 130mA plate current the drive voltage will need to be between 130mA/5.5mA = 23.6v RMS up to 130mA/3mA = 43.3v RMS. Since the tube's grid shouldn't run positive unless a special driver is designed to deliver grid driving power, the output tube bias and peak drive to the output tubes should be up to 43.3v RMS * 1.414 = 61v.

The normal bias will probably less than the worst-case noted above, and the driver/phase inverter probably won't have to output 61v peak, but a good design should allow adjustment to both that bias level and a phase inverter capable of throwing 61v peak to allow for the range of output tube Gm tolerance.

[kagliostro]

Ciao HotBluePlates

Many Thanks, very good and detailed explanation 

---

@ DummyLoad

the PT is ultimated, now my friend has to perform some test on it   





a bit of preblems on fitting the sprocket on the laminations (gently solved  )





Franco

[tubeswell]

Nice iron Franco

[kagliostro]

Oh, is my friend that has the "green thumb" 


100% hand winding


Franco

[shooter]

FIne hands indeed!  Wonderful work

HPB thanks

European system conveys more useful information

[kagliostro]

The normal bias will probably less than the worst-case noted above, and the driver/phase inverter probably won't have to output 61v peak, but a good design should allow adjustment to both that bias level and a phase inverter capable of throwing 61v peak to allow for the range of output tube Gm tolerance.

I was thinking to this raccomandations and if I find easy to plan the way to have a large bias voltage accomodation, I realized I don't know how to increase the "swing" of the PI, which are the components to be modified as to obtain a larger output signal from a PI ?


EDIT:

Is the way to increase the "swing" of the LTPI the same of a single stage triode ?

Does a higher plate value increase the output voltage in relation to the input voltage ?


Thanks

Franco

[PRR]

> how to increase the "swing" of the PI

Use the 5F6a driver. Power it from same/similar voltage as G2. It will have ample swing.

[kagliostro]

I apologize PRR

you told me it before

When in doubt, copy Bassman 5F6a.

Franco

[kagliostro]

Update

The PT, due to poor wire, had problems and the project stopped for a few time


now it is newly on the bench but it will be a 4 x GU50 amp, not a 120W but a 240W amp 

My friend, after to have wind some other transformers, has wind a new PT

(this time only near by hand, not fully hand, a wooden dowel was used as an aid  )

on the left you can see a PT for Fender 25W amp, on the right the new PT for the revised project (column 50mm)



the spec of the new transformer are:

Primary

220v (292 twist able for 2.8A)

Secondary

3 windings 221v + 108v + 255v (total 799 twist - 580v AC @ 1A)

The old chassis for two tubes is abandoned and a new one will be used (and may be build)

K