Mutual-conductance Characteristic Curves of Tubes - how to ? - schematics here

[kagliostro]

 Hi

anyone has ever built something like this ??

I'll like to know your opinion and to have news about (also about reels construction)

have someone a layout for that ??

MANY THANKS IN ADVANCE

kAGLIOSTRO

[FYL]

anyone has ever built something like this ??

No need to. You can measure Gm by using a stable signal generator, suitable power supplies (plate and screen plus bias) and an anode load - for instance a 10R resistor. Inject say a 5 KHZ square wave @ 100 mV, vary the bias voltage, get the current variation across the load, correct for plate resistance and you're done.

[kagliostro]

suitable power supplies

I find out this schematics for power supply, can this be ok for that use ??

(and also can I use this to have a laboratory PS for test, is possibol to increase the Voltage and more the current ??)

THANKS and excuse for many requests

Kagliostro

[FYL]

It's a pretty crude variable PS but it should be OK. Use imperatively a 500V-rated pot (RV4 Mil style or equivalent), heatsink the Mosfet and take the usual safety precautions.

You can also consider adding a voltage range limit resistor - say 100K or so - between the pot and ground as well as using uprated parts : the Mosfet is close to it's max voltage, a 600V / 8A model would be happier.

R2 sets the current limit thru T2, you may play with it if you want higher max current.

[kagliostro]

MANY THANKS FYL

If I have a higher rated Mosfet can I think also about to increase voltage ??

Kagliostro

[FYL]

If I have a higher rated Mosfet can I think also about to increase voltage ??

You may, provided that you use the right parts: very few components are qualified above 500V - voltage rating of caps is one thing, of pots, resistors, wires, etc. another.

[LooseChange]

That power supply IS Kevin O'Connors basic power scaling circuit.

[FYL]

That power supply IS Kevin O'Connors basic power scaling circuit.

KOC's basic PS is slightly different.

[kagliostro]

Thanks for info

here is Duncan's Munro schematics

Can this be a better regoulated Lab PS than the others ??

which it is the resistor to replace with a pot to obtain a regoulation in the output ??

http://www.duncanamps.com/technical/mosfet.html

MANY THANKS

Kagliostro

[FYL]

which it is the resistor to replace with a pot to obtain a regoulation in the output ??

Zeners are used for voltage reference, it's basically a fixed regulated PS.


May I suggest some reading re. tube testers?

http://www.triodeel.com/tester.htm

[PRR]

> which it is the resistor to replace with a pot

How does it work?

If you can't even guess, you won't be able to modify unless someone else does ALL your thinking.

There are thousands of voltage regulator circuits and hundreds of discussions on the web.

This one is similar to many. D10 is fed a semi-constant current and has a constant voltage. This is fed to Q2 base. Q2 emitter connects to Q3 emitter, a standard differential-pair. So Q3 base voltage "should" be similar to Q2 base voltage and thus to D10 voltage.

But what is feeding Q3 base? R9 R10 divider from output. If the R9-R10 voltage is not equal to D10 voltage, Q3 collector yanks on M1 gate and M1 controls output voltage. Some study will show that the Q3 M1 loop will tend to try to keep Q3 base voltage equal to D10 voltage.

So at a glance, you can make small trims of output voltage with small trims of either R9 or R10.

But large change of voltage are not so simple. M1 gate and thus OUT voltage can't go as high as Q4 emitter voltage. Which must be less than the voltages of D9 and D10. Which must be less than the input voltage. This seems to have been designed for a very specific situation. At other voltages it will need deep changes.

Why do you need a regulated supply?

Why do you need a Mutual-conductance Characteristic Curve?

Good vacuum-tube gear uses simple power supplies and almost any tube the right size and with reasonable gain. And for most audio purposes, the best test is to use the tube in an audio amplifier.

(The Mutual-conductance tester in your first post is specifically for something audio people almost never do: mis-bias a tube for very low gain. Radio men do this routinely: a radio needs large gain for weak signals and very much lower gain for strong local signals. The way the gain goes-away with excess bias can be important in radio design. But 99.9% of audio amplifiers just want fairly high voltage/current gain; anything less is not interesting. When we want "less" we use a volume potentiometer.

[jhadhar65]

>Radio men do this routinely...

AVC stages?

>But large change of voltage are not so simple.

What about a B250k for R5?

[kagliostro]

MANY THANKS TO ALL FOR ANSWERS

Thank you also for your patience

I'm not very expert at all and the most that I know I learned reading here in the forum,

so excuse me if sometime my questions are stupid or, at last, naive

for PRR

> Why do you need a Mutual-conductance Characteristic Curve?

some days ago, in an italian forum, there was another user asking about how to test tubes without a tube tester, in my mind I remembered a movi of a french man building his own tubes, he had a plotter and plot with it the tube curves, so I start to read here and there and I find the schematic I post with my first question

Why do you need a regulated supply?

FYL responding to my question told about an adeguate PS

well, I'm looking to build my own Lab PS and I was dragged asking about the way to have a good schematic for that project

My Lab isn't very well equipped and I'm always interested to project to increase his endowment as I can't afford large outgoings but I'm sure that with the right councils it will be possibe

Kagliostro

[PRR]

> What about a B250k for R5?

What about it?

That sets tail-current. This has nearly no effect on output voltage, except that as R5 goes below 100K it pulls all of Q1's current, the Zeners starve, the output collapses. (Unless the input is way over 300V: then Q1 may fail from over-voltage, probably fail short, and things go very wrong.)

D10 is 200V, D9 is 100V. We need over 301V input to get D10 happy and constant-voltage. The nominal output is 250V; this can be tweaked with R9 down to maybe 197V and up to 290V. Load current is said to be 25mA, so R1 drops 25V, and we need over 326V input at 25mA, over 401V input at 100mA load. Over 601V input violates Q1 rating, and 700V is begging for major smoke.

It is a driver power supply, will feed a very generous driver and lots of preamps. It would drive screen grids for many gitar amps. It could feed the big plates in a Champ, maybe a DeLuxe, if you like 200V-300V on the tubes with 420V raw power.

Duncan himself asks "Is it over the top?" IMHO, it is a fancy way to refine hi-fi audio, not really suited for stage-amps or for all-purpose bench work. KKT,MNY!

[kagliostro]

Here again

May be that an usual dimmer for inductive load

(used to drive the halogen lamps transformer - someting like Lutron products - http://www.lutron.com/applicationnotes/362219b.pdf)

can be used to drive a TA as to obtain a variable HV output ??

the idea is to drive a 220v primary / ~450v secondary PT to obtain the variable HV output

is that a feasible thing ????

Thanks

Kagliostro

[FYL]

is that a feasible thing ????

I wouldn't even think about it.

Build a real PS, even a basic version with a single pass Mosfet and a control pot will be fine.

[PRR]

> usual dimmer for inductive load

NO!

I did it once. It took weeks for the smoke-stink to fade away.

> variable HV output

I don't think I own a variable supply; certainly not a high voltage variable. When I need a little high voltage, I tap out of another amplifier. When I need a LOT of high voltage, unless I have something very similar to tap from, I don't want an "exact" voltage, I want a voltage which will sag like the real parts I will use. So I buy the real parts.

I don't own a regulated supply. The wall-voltage in my shop varies from 116V to 108V, which does cause a 7% change in voltage and maybe 15% change of power output. Other places around here run up to 124V. So when I was working on the "20 Watt", wall-voltage was 107V (it was a bad month), barely reading 17 Watts, I computed about what it would be with more normal 120V-124V power (like 22W) and was happy. Yes if I did this more often I should get "normal" wall-voltage. In fact I have a booster which will give me 3% 6% 9% step-ups... and I don't bother to use it.

> how to test tubes without a tube tester

Audio amplifier tubes, put them in an audio amplifier.

I have never measured transconductance.

Do you own a light-meter to test light-bulbs and see if a 100W lamp makes a full 1,000 Lumens? They may really give 900 or 1,100 Lumens, and you don't care. They almost never give much more light than rated. What usually happens is they give NO light, and you do not need a meter for that.

OK, a vacuum tube can get "soft", not dead, just weak. This will be obvious if you wire it up with standard bias and load resistors. For 12AX7 with 100K and 1.5K, the plate voltage will be 70% of the supply voltage. You don't need an exact supply voltage, you just observe if it is 65%-75%. You may use different resistors for different tube types, and you may not know what result is "normal" for that type, but if you have several different tubes of equivalent type then you will probably find most in a small zone and "bad" ones obviously "different".

> My Lab isn't very well equipped

Many geeks wind up with more toys than they know what to do with.

More toys does not give more insight.

I use one or maybe two voltmeters for almost everything. When I know the DC is happy and some sound comes out, I may add a signal generator and an oscilloscope. While I have some very fine sig gens, I rarely use more than a couple volts or more than a few frequencies. And my favorite silly-scope is completely un-calibrated and barely gets to 20KHz. I have a fancy-brand decayed-sweep 40MHz 'scope, but it has too many knobs and so many parts that it is often set-wrong or has a bad connection inside.

That's gitar amps and general hi-fi. If an amp is flat at 80Hz and at 160Hz, it is probably fine everywhere in between. Yes, when building narrow-band filters I may need a fine sweep of 119Hz, 120Hz, 121Hz.... but that's unusual. When building a recording limiter I may need 3 or 4 voltmeters (I own about ten) watching different places, to see why the control voltage is not getting through, but that is unusual. Working on transistor amps, I like to have a 20MHz 'scope because they tend to take-off at high radio frequencies; tubes are less likely to do this and when they do, it can be so far past 20MHz that no affordable 'scope could reveal it. 

[kagliostro]

Many thanks for explanations and councils

about PS now I can see the thing another way

about tube test I remember the tread on wich you (PRR) explain about to build a little "champ amp" to test tubes

the question about tube testing I read in an italian forum was in a forum were they often mean HiFi tube amp when the speak about tube amp and they have a "particoular" vision about how an amp must be and how must be build whit selected parts (I really dislake some "speaking about the sex of angels") and I can well understand your argoument about "selecting" a tube


AS ALWAYS MANY MANY THANKS

Kagliostro