5654 tube as power tube - which is the correct OT primary impedance in SE & PP ?

[kagliostro]

Some time ago at Amp Garage there was a guy asking about the use of a 5654 (6ak5) tube as power tube
as I've many of those tube I had interest on that thing and I started a search, but the only schematics
I found where about the use of the tube as preamp tube
this days I've found a schematic which confirms that the 5654 can be used as power tube

but the use of a PT as OT isn't comforting
I do a bit of math and with the 230v/6v PT used, the results are
8ohm speaker = 11755ohm primary
4ohm speaker = 5877ohm primary
but I'm not sure that this values are the right values
Please, someone knows which will be the right value for a 5654 in SE & in PP ?


Here the datasheet of the tube
http://www.tubebooks.org/tubedata/HB-3/Miscellaneous_Types/5654.PDF
http://www.r-type.org/pdfs/cv4010.pdf
http://tdsl.duncanamps.com/pdf/5654.pdf
http://www.mif.pg.gda.pl/homepages/frank/sheets/009/5/5654.pdf


Many Thanks


K

[tubenit]

Thanks for sharing that!

Some yrs ago, Sluckey gave me some to try out in the Tweed Overdrive Special in the V1 position.  They sounded GREAT! & were pretty comparable to the 5879 in tone. 

I also later tried one out in a tremolo position and that worked quite well also.

With respect, Tubenit

[jojokeo]

Why not use the 6AK6 instead? I would highly suggest these instead and save your 6AK5's for where they're better suited...in the preamp.

[kagliostro]

May be I've 25-30 5654 and no one 6ak6 


K

[HotBluePlates]

I do a bit of math and with the 230v/6v PT used, the results are
8ohm speaker = 11755ohm primary
4ohm speaker = 5877ohm primary
but I'm not sure that this values are the right values
Please, someone knows which will be the right value for a 5654 in SE & in PP ?

Your math is correct for the impedance presented at the primary, based on the voltage ratio shown in the schematic.


I'd guess the intended speaker load is 8Ω, to reflect ~12kΩ. But it could also be 4Ω for ~6kΩ (see below).


I looked at the Sylvania 6AK5 data sheet hosted at Frank's (they have a lot of 6AK5/5654/EF95 data sheets). The usual guidelines to speed output stage design go out the window when you use a tube like this, intended for (high frequency) voltage amplification (and into a reactive load instead of a plate load resistor). I noticed many sheets gave curves only up to -1v on the grid, and took that as a clue to avoid driving it to 0v.


We should note that your schematic has a B+ of 120v, applied to both plate and screen. This is handy, as there are data and curves for 120v on G2.


Looking at one of these curves (page 2 of the data sheet), for single-ended class A operation, the idle point must sit on the loadline and at 120v (because that is the schematic voltage). The 390Ω cathode resistor is a clue, and after some hunting, I found that an idle point of ~6mA yields a bias of -2.34v at 120v on the plate.


You'd like the tube to swing a similar amount above and below the idle point and pass through a point above the knee of the curve. I picked an arbitary point above the knee at 40v plate and 12.5mA of plate current, and drew a line passing through the idle point I'd previously found. It intersects the 0mA axis at 195v.


Load resistance = (195v - 40v)/12.5mA = 12,400Ω. This is shown by the red line on the graph below.


Because we probably could drive the 6AK5 up to 0v on the grid (though no line is shown on the graph), I also plotted a 6kΩ load line in blue. The G.E. data sheet from Duncan Amps that you posted does have a 0v grid line, and the 6AK5 doesn't quite hit the 19mA that the 6kΩ load line would imply. So either load could probably work.


Power output for the 12.4kΩ load is (195v-40v) * (0.0125A-0A)/8 = ~0.24w
Power output for the 6kΩ load is (157v-40v) * (0.019A-0A)/8 = ~0.28w (though the tube probably won't swing that high a peak current)


If you want push-pull operation at this same plate and screen voltage of 120v, then you'd use the same loads of 6kΩ or 12kΩ plate-to-plate.


It would be possible to figure class AB operation with smaller load impedances for more power output, but with a tube this small, that's pretty silly to sped time on...

[kagliostro]

Many Thanks HotBluePlates

for the indication about the impedance

and


the very good explanations


---



For those who can have interest on this small tube, here some schematics I've find about the use of 5654 as preamp tube






K

[kagliostro]

Hi Drgonzonm  Thanks for your contribute

in the schematic I posted the 5654 is used as power tube in SE

a PT rated 230v - 6v is used as OT and no impedance indication is given about the speaker that is used

squaring the voltage labeled on the transformer is possible to know the reflected impedance seen by the tube

assuming the used speaker is an 8ohm speaker the result is an impedance of 11755ohm

assuming the used speaker is a 4ohm speaker the result is an impedance of 5877ohm

I supposed that the impedance of the speaker was one of those (8ohm or 4ohm) as those are the more common

impedances used in speakers

So I was asking if the math was correct and which will be the impedance that better suited the tube

(my knowledge is limited and I'm not able to read datasheets for this purpose by myself)

HotBluePlates has a large knowledge in this matter and I suppose that he has seen that the 12k load

is a good load for the tube, but he investigated also the 6k load, also if less suitable

---

If the datasheet of the tube give a direct indication of the r_{a - a}, that is a good thing that we can follow, but not always that indication is given

---

5694, as a power tube

there is a little typo, the tube is 5654, not 5694, can you edit your post to correct it ?

Ciao

Franco

[HotBluePlates]

I couldn’t find 5694 tube data for a push pull application, even reviewing the variants of the tube ...

You are correct, as it doesn't exist. The designers and manufacturers of the tube planned for it to be a voltage amplifier in a preamp (where it will be most often single-ended), not as an output tube. Strictly, you can operate any voltage amplifier tube in push-pull in a preamp, but for the most part the style of circuit you'd use allows calculation as though you were running the tube single-ended.

Am I incorrect in using the r_a-a  data found in push pull data as the first indicator to select the primary impedance of the O/T?  For example the EL84, shows an r_a-a of 8Kohms, so I would select a transformer with primary impedance of 8K? 

And so, as you had suspected in the earlier post, you won't find R_a-a data from any manufacturer, as the tube wasn't intended to be run that way. If you'd held a 6AK5 in your hand, you'd understand why the thought of it being a "power tube" is laughable... (see pic at bottom) Same basic size as a 6AL5 if you've ever encountered one of those.

[By the way: I like how a 6AL5 was used as a rectifier in the original schematic, given the small power demands of the radio depicted. Gives me an idea for all the 6AL5's I've got that had been lumped in with other tube-lot purchases. Normally, these tubes are beyond useless unless you have something that aleady uses them and want to stay all-original.]

With the square of the voltage, then HBP info makes sense.

You know now, but for everyone else:
For a transformer (assuming no losses), Power In = Power Out. There are turns on the primary and turns on the secondary, and therefore a ratio of the number of turns on each. Voltages transform from primary to secondary in a proportion equal to the turns ratio just mentioned (so that means Voltage Ratio = Turns Ratio). Kagliostro's OT was specified as "230v:6v" which is the same as 230v/6v = 38.33:1 Turns Ratio.

If Power In = Power Out, and there is a set Voltage Ratio between windings, then how does impedance transform from one winding to the other. You could write the equation for power in the form Power = Voltage²/Resistance, then convert to Resistance = Voltage2/Power.

At this point, you'd need to know that the equations for a transformer invert this relation for power from one winding to the other, so with the same power present on each winding and using the form which solves for resistance, you get:

Power/Voltage2= Voltage2/Power, and the term for Power cancels (because they are equal), leaving Voltage²/Voltage2 as our ratio for Resistance (really "Impedance"). Therefore, the Impedance Ratio of the transformer is the square of the Voltage Ratio, which is the same as the square of the Turns Ratio.For Kagliostro's 230v/6v transformer, the 38.33:1 turns ratio yields an Impedance ratio of 1469.44:1, and an 8Ω speaker on the secondary reflects 1469.44*8Ω = 11,755Ω to the primary. Similarly, a 4Ω speaker reflects 5878Ω to the primary (and I just rounded those up to 12kΩ and 6kΩ, as they're close enough).

[HotBluePlates]

An interesting thing I saw when looking for a picture of a 6AK5 was that there is a 6Ж1П (6Zh1P or sometimes 6Z1P) made by Russian tube companies which is essentially the same tube.

The 6AK5 is now being used in some hi-fi products by Little Dot, and so where you can buy them for ~$4/ea or less from tube dealers, folks on Ebay are asking $20-30 or more for a pair. Of course, if you look carefully you can find the 6AK5 cheaply on Ebay in large lots, but at least one seller has the 6Ж1П at $0.80/ea.

Worth keeping in your back pocket if 6AK5's ever become scarce.

The Cyrillic under the Russian tubes says "Novosibirsk" and "Svetlana" and is the manufacturer of those particular examples.

[HotBluePlates]

I can see HBP, going into business of making very low amperage push pull amps.  6AL5's as rectifiers and 6AK5's as the power tubes.

Sure! Two 6AK5's in push-pull making 0.7w output. We'll call it the "AK-7" (Trademark, All Rights Reseved).

 

[kagliostro]

We'll call it the "AK-7"

Be careful about the use of russian tubes


someone can misunderstand and think about an AK-(4)7 supply 

---

If I can, I suggest those as basis for the development of your new guitar SE & PP tube amp line


http://jjs.at/electronic/

K

[HotBluePlates]

We'll call it the "AK-7"

Be careful about the use of russian tubes ... someone can misunderstand and think about an AK-(4)7 supply 

Well, that's what I get for coming up with a name... "AK-47" is what I meant!

[PRR]

Sure looks to me like a small 6AU6.

FWIW: 6AK5 is the first hi-voltage tube in the Miniature bottle and was meant for VHF gain.

At a glance: for Vg2=120V the knee is around 10mA 20V. Vg2 limit is only 140V so up-volting won't change things much.

In SE you would idle at half that, 5mA. With plate idled at 120V, you could swing to 20V 10mA, or a 100V 5mA swing. Impedance must be in the area of 100V/5mA or 20K. Peak output is 100V*5mA or 500mW; Sine output would be 250mW. Idle dissipation is 0.6 Watts, far short of the 1.7W limit.

This IS a small-signal tube. In VHF (radar) IF amps we don't aim for the center as in class A power designs, we aim high for more gain with less snow. Note the 7.7mA condition. You can't take a large swing from there, but the gain is high.

Going to 140V on both elements might give 0.8W Pdiss and 340mW out.
 
Going all the way to 160V on plate makes the already-high (for audio) load impedance even higher and the very slight added power-out is probably not worth the OT compromise.
 
Other tradeoffs (impedances, voltages) are possible. Doesn't make a large difference.
 
> I've 25-30 5654
 
OK, parallel ten of them.
 
Now you have about 2K load and 2 or 3 Watts out. You also have more heater current than a KT-88. And a LOT of pins to wire. The maximum grid resistance is not listed on 6AK5 sheet (because VHF circuits are naturally low impedance); but say it is 500K. Ten grids "should" have no more than 50K grid resistor. That may be a hefty load if the driver is a pentode. (OTOH you only need 2V of drive...)

Design of the 6AK5

[kagliostro]

Hi PRR

Many thanks for the explanations

---

Here is Fig 4-60 that is on the pdf file you linked


And this is my last source of supply for 5654


I got it at the hamfest of Saturday last

Franco

[jojokeo]

Now here using 6AK6 instead gives you more bang for your buck at same voltages a bit more current - 1.1 watts, two would give much greater benefit with same OT as you'd use w/ 6AK5. Kag, I know you don't have any now but you could find them easily and cheaply. It would be much more worth it and rewarding in the end. Have you played something with this little of power before? Better for driving a pair of headphones not a speaker cone. You can hear your guitar strings more than the speaker, lol. Anything less that a watt is virtually useless and like drinking watered down tea at best. Not a great example but if you're putting in the time and money you will want something better out of your efforts. I have actually built with this and know...but it's for you to decide in the end.

[kagliostro]

To be honest, at all hamfest where I go, I try to look at all kinds of tubes that may have some use, but until now I've found only one 6ak6 and it was broken 

Franco

[PRR]

> my last source of supply for 5654
 
There's a 6CL6 on there. 6CL6 *is* a power tube.
 
The cathode is a bit larger for more gain; the grid-plate system a bit smaller for less stray capacitance, better treble (out around 4MHz). So not designed for big slow audio, but not bad.
 
The data shows a happy 2.8 Watt audio amplifier with Champ-like OT. It only needs 3V of drive, so any small pentode (maybe you can find a 6AK5/5654) should bring guitar up to a volume control and into the 6CL6. Use the early (no 12AX7) Champ plan. Call it a half-Champ.

[PRR]

Bucket of 6AK5....  Eight bottles....  180 Vp, 140Vg2, 6.6K p-p load, 6 Watts out.

[kagliostro]

There's a 6CL6 on there

absolutely nothing escapes to you huh 

---

maybe you can find a 6AK5/5654

Oh, yes, may be 

---

Many thanks for the advice 

------

Eight bottles....  180 Vp, 140Vg2, 6.6K p-p load, 6 Watts out

это фантастика ! 

----

Many Thanks PRR

Franco

p.s.: BTW in the small chassis there is also a 5670 (1W output with 27k load) 

http://www.tubebooks.org/tubedata/HB-3/Receiving-Type_Industrial_Tubes/5670.PDF

[Bergera]

Bucket of 6AK5....  Eight bottles....  180 Vp, 140Vg2, 6.6K p-p load, 6 Watts out.

Very tempting I'm also looking at a quad 6ak6

[PRR]

> there is also a 5670

Pretty much a re-rated 6BQ7, a very popular TV tuner tube.

Long-long ago I built a 6BQ7 headphone amp. Didn't take measurements, didn't get long life, was probably over-abusing them. (But they were VERY cheap when tube TVs went out of style.)

6AK5 6BQ7 5670 are all Mu= 35 with 1.1W-1.3W heater. You can think of them as the "same" tube, except the Pentode allows more pull-down in large-signal work (but really less stray capacitance in VHF work). Not surprising they make similar power, the pentode a bit better especially at low voltages.

[kagliostro]

Thanks


Franco