Estimatings max cathode current?

[kiira]

I'm trying to design a line stage with a tube for which there isn't much data and the max cathode current is not listed... max plate current is 25ma, max grid current 8ma and plate voltage max is 250. Any advice how I could go about determining if the operating conditions are not being exceeded for the cathode current?

thanks,

Kiira

 

[NewYorkDave]

Well, you gotta give US more data. Is it a triode? If so, then the cathode current is the same as the plate current. Is it a tetrode or pentode? Which grid do you refer to when you say "grid current?" I'm guessing you mean the screen grid. In that case, the cathode current is equal to the plate current plus the screen current. The cathode is, after all, the source of all the electrons.

Come to think of it, I'm not sure if I've ever seen "maximum cathode current" listed on a tube datasheet. It's just implicit from the max plate and screen currents. Plate and screen dissipation set the limit on when the tube melts down, anyway.

Posting the tube type doesn't hurt, either.

 

[Mark Burnley]

...and even the weirdest tubes can usually be found here:

http://frank.pocnet.net/

:thumb:

Mark

 

[kiira]

oops sorry I forgot to post the link to the data sheet
http://www.2tough.com/~kiira/rec/1626.pdf

It's a triode so the cathode max will be 25ma like the plate then. agh I spent a day looking for this info in "Tube Amps for Maths Geniuses" by Morgan Jones.

thankyouthankyouthankyou!

Kiira

 

[CJ]

Try 100 ohms.

Cathode resistors are self regulating. The more current, the more neg volts the grid feels, the less plate current. Less current means less neg volts = more plate current. The tube will settle by itself. Hench the word "self bias".
You can't blow up a tube that uses a cathode resistor. Unless its zero ohms.
And even then it will be tough. Unless it's a 6L6 at 500 v.

 

[kiira]

This is what I came up with for the simplest thing. Gain is about 4 and 12 db. I think. I hope.
I think maybe I'll just build this first and see how it sound then maybe get fancier. Orcad LIte won't let me edit symbols so Ihad to use a pentode just because it had the right pinouts.

 

[pstamler]

[quote author="kiira"]This is what I came up with for the simplest thing. Gain is about 4 and 12 db. I think. I hope.
I think maybe I'll just build this first and see how it sound then maybe get fancier. [/quote]

It looks good on paper. Probably a good idea to make that plate resistor at least a 5-watt and the cathode resistor at least a 1-watt.

Peace,
Paul

 

[NewYorkDave]

Orcad Lite does include a triode symbol. It's listed as TRIODE/DISCRETE.

 

[kiira]

[quote author="NewYorkDave"]Orcad Lite does include a triode symbol. It's listed as TRIODE/DISCRETE.[/quote]

Yah but I couldn't edit the properties to change the pin numbers. Orcad says I'm not allowed to edit any properties or save a new symbol to a library bigger than 15. Mybe I'm doing something wrong

Kiira

 

[NewYorkDave]

Drop the symbol into your schematic, then right-click on it and "edit part." When the part editing window opens, double-click on any pin and that opens a dialogue that permits you to change pin names/numbers/length, etc.

 

[CJ]

Lower that cathode resistor and just plug it in.
Be brave.

:razz:

 

[kiira]

Oh I see I can edit it in the schematic if I don't choose global save. Yay. Thanks.

I changed the cathode resister to 100 ohms and raised the bypass cap to 100uF and that puts more volts across the tube. I'll play with losing the bypass cap.

KIira

 

[Guest]

[quote author="kiira"]http://www.2tough.com/~kiira/rec/1626.pdf
[/quote]
Oooh, grid current for 5 W triode, somewhat unusual...
For audio grid current is used for 1 kW + power, because
efficiency is higher.

But really. This is tube intended (only hypothesis ????) for solo - oscilator transmitter,
then grid current was used (with grid bias resistor) to stabilize
amplitude of the oscillator (still working in linear region).

But you use triode as resistor coupled amp? why. If you use
pentode with the some transconductancy, you will have gain 5 X
higher.
Triode is unpractical in resistance coupling, mainly because inner resistance dominates and decrease gain.
In all causes, you will be still in the grid-cathode negative voltage
region without grid current. Icat = I plat.

By the way. It is relative big triode, and 1 M grid - leak resistance
can be too much. See similar circuits, or measure R1 voltage drop.
May be under 100 mV. If not, use resistor 100k or lower.

xvlk

 

[CJ]

What?
Arafat, are you on crack?

Try 1k cathode R first, and work from there. minus 25 vdc seems like a good starting bias.

 

[kiira]

[quote author="xvlk"][quote author="kiira"]http://www.2tough.com/~kiira/rec/1626.pdf
[/quote]

But you use triode as resistor coupled amp? why. If you use
pentode with the some transconductancy, you will have gain 5 X
higher.
Triode is unpractical in resistance coupling, mainly because inner resistance dominates and decrease gain.
In all causes, you will be still in the grid-cathode negative voltage
region without grid current. Icat = I plat.
[/quote]

Ok, yer foolin' with my right? It's a resistor cathode biased triode the most common circuit prolly like forever. I don't care about the gain so much anyway though because 12 db is ok for a line stage I think. And besides this is just a simple circuit using this particular tube and not meant as any kinda ground breaking design... others are handling that fine here. I'll build it and see how it works, if I plotted the load line and chose the operating point ok and did the maths and if it sounds nice I'll take it apart and build something else like a plate follower or a mu follower or something.

thank you for the feedback.

Kiira

 

[PRR]

> you use triode as resistor coupled amp? why.

Triodes make fine resistor-coupled amps.

Triode gain only goes to 70 or so, while pentodes can do 100-400 IF the load resistance is very high. But sometimes loads are low, and triodes are easier.

I believe that tube is safe with a 1Meg grid resistor. It is a big "small tube", not a big tube.

The tube is rated for grid current (unusual in a tube this small). It was mostly intended for a very stable RF oscillator. In that work, the grid current would be small (~0.2mA) once the amplitude stabilized. But as long as you had that tube in your design, you might as well use a second one for a buffer. If you beat the grid with current you cam make several watts of RF. For some small transmitters, that might be the final output stage.

In audio, of course, we "never" want grid current. Grid voltage less than cathode, grid impedance is 1 Meg. Grid higher than cathode, grid impedance is 1K. 1K is a very low value for any tube to drive. Grid current makes sense when we have super-high plate voltages, tuned-circuits to transform impedance and average-out current pulses, and when we have a big need for power in a limited number of bottles.

On a simulator we have to be careful: some simulator models don't know about grid current.

> grid on pin5, plate on pin 3...

The pin numbers in the simulation do NOT have to be the same as the pin numbers on the bottle. Simulator models always number the triode 1 2 3 even if the real tube has the pins scattered all over the tube. Just be sure the order is right: some models order the pins G K P and some K G P.

This is a kinda-big tube, and in a totally logical universe you would not use it for small chores. With 300V supply and 13K plate resistor, the maximum possible plate power is 1.6 watts, much less than it is able to handle. I like the condition 300V (or 290V) supply, 150Vp, 25mA (plate and cathode). That gives grid bias about -12V, so cathode resistor is 12V/0.025= 480 ohms. Plate resistor is about 140V/0.025= 5K6. Since in this real world you sometimes use a tube that is "too big", I agree with using 10K-13K plate resistor, which leads to the 1K cathode resistor that CJ suggests.

With a 100 ohm cathode resistor you will be sitting barely-below the zero-bias line. Gain may be good, but you can only get a few volts of plate swing without going into grid current. And we never do that, because the great thing about tubes is no-input-current (as long as the grid stays negative at all times). This point may actually be ample for a HiFi line-amp (we only need abot 3V peak) but not for studio work.

 

[vibrolax]

The 1626 is a power triode designed as a transmitting tube. It has gained some fame among the low power hi-fi set a few years back after Bob Danielak's published his DIY "Darling" design on the web. http://www.geocities.com/bobdanielak/darling.html

By googling or following the links, you can see where other people have taken the design.

-Jon

 

[kiira]

hI

[quote author="PRR"]

This is a kinda-big tube, and in a totally logical universe you would not use it for small chores. With 300V supply and 13K plate resistor, the maximum possible plate power is 1.6 watts, much less than it is able to handle. I like the condition 300V (or 290V) supply, 150Vp, 25mA (plate and cathode). That gives grid bias about -12V, so cathode resistor is 12V/0.025= 480 ohms. Plate resistor is about 140V/0.025= 5K6. Since in this real world you sometimes use a tube that is "too big", I agree with using 10K-13K plate resistor, which leads to the 1K cathode resistor that CJ suggests. [/quote]

Could you explain how the 13k resistor is more appropriate for this tube? I chose it based on the plate resistance and the operating point I wanted and 10ma max I.

With a 100 ohm cathode resistor you will be sitting barely-below the zero-bias line. Gain may be good, but you can only get a few volts of plate swing without going into grid current. And we never do that, because the great thing about tubes is no-input-current (as long as the grid stays negative at all times). This point may actually be ample for a HiFi line-amp (we only need abot 3V peak) but not for studio work.

I must be doing my math wrong because I got about -29V bias with the 100 ohm resistor and -27 with the 1K.

I'd like to eventually make this a line stage that could drive 600 ohm inputs or phones without needing a xformer. Not as a grounded cathode of course.

Thanks for the feedback.

Kiira

 

[CJ]

Your plate resistor will define your max plate current for a shorted tube. In other words, lets imagine 300 volts of B+ on a 15,000 ohms resistor. Since a tube is just a variable resistor inside a glass cage, your theoritical max Ip will be an Ohm's Law equation, right?

So even if 100 ohms is too low, and puts you on a distorted part of the load line, your tube won't fry. It will just be operating at a bad bias point.

 

[PRR]

> I must be doing my math wrong because I got about -29V bias with the 100 ohm resistor and -27 with the 1K.

29V drop across a 100 ohm cathode resistor is almost 300 milliAmps. Yes, you must be doing math wrong. Show your work and someone will spot the slip-up.