the 1.25v of lm317

[Albrerta]

I built a LM317 based ccs for tube heater PS, in order to get 600ma I used a 2R resistor, it supposes to have 1.25v voltage drop on it, but I only measured about 1.11v.
Then I added another 20R resistor in parallel to lower the total value to about 1.8R, surprisingly the voltage drop also lowered to about 1.03v
Finally I settled it with 1.5R and 0.9V which is perfect.
The LM317 is new from newark.com, ST brand. So I guess the 1.25v rule is not always the case.
Am I correct?

 

[AndrewT]

does not sound right.
Have you got the pin connections right.

 

[woody]

How many volts are being droped across the regulator?

 

[Omega_Void]

Did you include the bypass cap on the adj pin parallel to the voltage-setting resistor?

I built an LM317 circuit, the output voltage fell under medium loading, and voltage from out to adj also fell as loading increased. The cap turned out to be the problem. Clipped it off and the reg behaved itself properly. Stuck a new cap on and it was fine.

 

[tomchr]

The 1.25 V is set by a bandgap and should be fairly solid (+/- 4 % AFAIR) over process, voltage, temperature as long as the regulator is stable and has enough voltage across it (Vin-Vout) to allow regulation.

Vin-Vout should be greater than 2.5 V and the regulator should be bypassed to ground both on input and output with 100 nF film cap.

~Tom

 

[Albrerta]

that's exactly I did
12.6AC ->rectifier ->18000uf ->0.75R(2x1.5R) ->1.8uf film -> Vin ->Vout ->1.5R ->connect to adj no bypass cap ->47uf

DC output is about 12.3 with load

Can you tell anything wrong?

 

[Omega_Void]

Did you include the bypass cap on the adj pin parallel to the voltage-setting resistor?

No, of course you didn't 'cos you're using it as a CCS. Sorry, I managed to miss that, even though you stated it plainly.

 

[infinia]

These same questions keep happening to the tube guys and their %#@$ filaments.
Its mostly comes down to using a filaments cold resistance and making sure the three terminal Vreg is not
1) Thermal/Current Limiting
2) Has enough overhead to allow regulation ie actual measured Vin -Vout

 

[benb]

To be more pertinent on the questions of others, is the power supply at least four volts or so (for Vin-Vout>=2.5V, plus Vout-Vref=1.25V for a minimum of 3.75 volts) higher than the heater voltage rating? If not, the current regulator circuit will drop so much voltage that the heater won't get full voltage and likely won't pull full current.

What is the voltage across the filament?

 

[Albrerta]

with load ACin is 12.6 and DCout is about 12.3, totally fine with me.

Since I used it as CCS, I don't set the output voltage, the 317 will take whatever it needs, (2.5v or so), so that's not the point

BTW, thank you all

 

[infinia]

If the band gap is not stable what is the point. > It's not working right
Why dont you put up an actual schematic and measured voltages, hopefully taken with a scope.

 

[infinia]

If you push too much current into a fixed load resistance Vo rises. When Vo raises higher it takes away from the min voltage drop the regulator needs to see = (Vi-Vo)+Vref ~ 3.5 -4.0 V

Note If you lower the current or raise Vin it should work properly. Try using 500mA and eliminate Rdrop 0.75R