LM317AHVT and Heatsink

[mike_mike]

Should I test again the modified schematic with the oscilloscope ?
The only difference between the tested and the attached (current) schematic are the 3x8.2K parallel resistors (which was initially only one 3.3K resistor) and the 100R resistor (which was initially 122R resistor).

And I am facing another problem. If I set the output voltage to 26.4V when I connect a 9Amps load, the voltage drops by 0.5V, reaching 25.9V. I did not find this issue last year when I tested this power supply. Why this voltage drop appears ? What should I do to correct this voltage drop ?


If I set the output voltage to 19.53V and I connect a 4.74A load, then the voltage drops to 19.27V. I did not found this behavior of the power supply at previous versions.

I re-tested using an older PCB and I obatined 25.9V with 7A load, and 26.0V without load. So, probably the PCB is the problem.

[David Hess]

Should I test again the modified schematic with the oscilloscope ?
The only difference between the tested and the attached (current) schematic are the 3x8.2K parallel resistors (which was initially only one 3.3K resistor) and the 100R resistor (which was initially 122R resistor).

That change should not matter.

And I am facing another problem. If I set the output voltage to 26.4V when I connect a 9Amps load, the voltage drops by 0.5V, reaching 25.9V. I did not find this issue last year when I tested this power supply. Why this voltage drop appears ? What should I do to correct this voltage drop ?

If I set the output voltage to 19.53V and I connect a 4.74A load, then the voltage drops to 19.27V. I did not found this behavior of the power supply at previous versions.

It looks like you have a high enough minimum load on the output.  Make sure that you are not falling below the dropout voltage of the combined regulator and transistor and make sure there is not a parasitic oscillation.

C3 should be located at J2 and not near U1.  R3 should not be required if R6 is low enough to provide the entire minimum load.

I re-tested using an older PCB and I obatined 25.9V with 7A load, and 26.0V without load. So, probably the PCB is the problem.

It could also be a layout problem.  The lower part of the divider should terminate at a single point near J2 but the top of the divider should terminate at a single point close to U1 for best load regulation.

[mike_mike]

What is the difference between A and B from the attached schematic, besides the direction I need to rotate in order to increase or decrease voltage ? Is there any difference between A and B ?

[David Hess]

There is no difference.

[mike_mike]

Thank you for the reply David Hess.
I tested the power supply and the results are (the schematic is the one from reply #27, with C4=2200uF/63V instead of 1000uF/63V)
without load V=19.80V, with 6.17A load V=19.79V
without load V=26.0V, with 7.24A load V=26.0V

I don't know if the wire that come from the pin 1 or another pin of LM317 to the PCB was making good connection with the screw terminal located on the PCB.
If there was a bad connection, could the power supply work good and if I repair the bad connection by tightening the screw could the power supply to work bad ?

[MarkF]

Dear Forum,

I made a layout for a LM317 power supply, using the attached schematic.
I don't know if I correctly made the layout. Please have a look at the layout and tell me if it is good.
I want to know only if the layout is correct. The schematic that I draw in KiCAD is correct because I checked it. I also added a LED and a current limiting resistor (R4 and J3).
The refdes of the components does not correspond from the schematic to the layout.
I was a little bit confused by the copper pours, because I don't know if I correctly made the copper pours.

I did a similar PCB which will accept either AC or DC input.

   

[Zero999]

Thank you for the reply David Hess.
I tested the power supply and the results are (the schematic is the one from reply #27, with C4=2200uF/63V instead of 1000uF/63V)
without load V=19.80V, with 6.17A load V=19.79V
without load V=26.0V, with 7.24A load V=26.0V

I don't know if the wire that come from the pin 1 or another pin of LM317 to the PCB was making good connection with the screw terminal located on the PCB.
If there was a bad connection, could the power supply work good and if I repair the bad connection by tightening the screw could the power supply to work bad ?

The is no issue with those measurements. A 10mV change in output between no load and 6.17A of load is equivalent to a resistance of 10m/6.17 = 1.62mOIhms, which is very good. If there's a problem with a mechanical connection then it isn't affecting the circuit operation but it should still be resolved to improve the reliability.

[mike_mike]

1. Can someone explain how to calculate the voltage on the potentiometer ?
It is the output voltage minus the voltage on the resistor (1.25V) ?
2. If I remove the 3 8.2K resistors and the 2 potentiometers and I use instead of them only a single 2K/2W multiturn potentiometer, what can change in the power supply operation ? excluding the output voltage ?

[MarkF]

1. Can someone explain how to calculate the voltage on the potentiometer ?
It is the output voltage minus the voltage on the resistor (1.25V) ?
2. If I remove the 3 8.2K resistors and the 2 potentiometers and I use instead of them only a single 2K/2W multiturn potentiometer, what can change in the power supply operation ? excluding the output voltage ?

First, get rid of all that resistor nonsense you have.
Keep the two potentiometers if you want a coarse and fine adjustment (with the fine adj. being about 10% the value of the coarse).
On the side, your capacitor values are way out of wack.
Then watch this video:

   

   

[AngraMelo]

Im sorry to go out of the main topic here but can someone please explain to me how Vin goes to the base of the transistor but Vout comes out of the collector?

[mike_mike]

On the side, your capacitor values are way out of wack.

You mean that the capacitors have a too high or too low value ?

[MarkF]

Im sorry to go out of the main topic here but can someone please explain to me how Vin goes to the base of the transistor but Vout comes out of the collector?

The two main current paths are through the transistor and through the LM317 (green and blue).  The resistor R1 (yellow) determines how much current bypasses the LM317 by setting the emitter/base voltage as a function of current.  The  more current required the more is bypassed.

   

On the side, your capacitor values are way out of wack.

You mean that the capacitors have a too high or too low value ?

Far bigger than necessary for a DC input.  Maybe a 100uF and 1uF on the input.  And a 10uF and 0.1uF on the output.  For me, I would just have a 100uF on the input and a 10uF on the output and let whatever is being power have it's local decoupling.

[mike_mike]

Thank you for your help.
I watched your video, and I understoon the following: If I need to calculate the power dissipated on the potentiometer, then I need to calculate the output voltage of the power supply, then decrease the value by 1.25V, then calculate the current through the potentiometer using Ohms Law and the using the formula P[W]=I[A]*V[V] to calculate the power ?
For example:
The output voltage is 26.25V, then the voltage on the potentiometer is 26.25-1.25V=25V, the current through pot is 25V/2000R=0.0125A, and the power dissipated on the pot is 25V*0.0125A=0.3125W.
Is that correct ?
Please consider the attached schmatic, the DC input is unfiltered.

[Zero999]

Yes, you've calculated the power dissipation in the potentiometer correctly.

Note for future reference the maximum power rating of the potentiometer is only for the whole track, when current is passed through half of the track, the power dissipation will be halved. Thsi isn't a problem for this circuit since the current through the potentiometer is constant, but it's good to know.

[mike_mike]

If I switch from the first schematic (with dual pots) to the second one (with 2k pot), what would change in the functionality of the power supply ?

[MarkF]

It's NOT a matter of one or two pots.

It's the 3 resistors you have in parallel with the pots.  Get rid of them!


Edit- Removed bad info

[MarkF]

I just noticed that you have the resistors on the emitters. 
They should be on the collectors for PNP transistors.

Remember the outputs are being summed together and
you want to compensate for small variances of each transistor.
You always want the resistors on the output side.

And REMOVE R3.  If you can't set R6 to meet the minimum load, then add a constant current source (not a resistor)

If I switch from the first schematic (with dual pots) to the second one (with 2k pot), what would change in the functionality of the power supply ?

None with corrected two pot circuit.
 - You would design for the single pot to get your desired voltage range.
 - Then you could just use a single 3/4 turn pot
 - Or you could add two 3/4 turn pots with the smaller value being around 10% of the bigger depending on how fine of a control you want
 - Or you could use a single 10-turn pot

[macboy]

I just noticed that you have the resistors on the emitters. 
They should be on the collectors for PNP transistors.

Remember the outputs are being summed together and
you want to compensate for small variances of each transistor.
You always want the resistors on the output side.

No, it is definitely correct to have the resistors on the emitter side!

They act as "emitter degeneration resistors" in order to effect the current sharing. As more current flows through one of these resistors, the emitter (and therefore, base) voltage is affected in the way which reduces base current, creating a negative feedback. This enforces tight current sharing between the transistors, as changing Vbe by just a few millivolts can dramatically affect collector current. This scheme is not as effective for MOSFETs because of the much looser relationship between gate voltage and drain current.

[mike_mike]

Thank you for the replies.
1. I have a 2K/2W potentiometer with the part number WXD3540. In the datatsheet it specifies "Independent Linearity". This means that the potentiometer have linear characteristics ?
I don't see any other information about the linearity in datasheet.

2. I made an accidentally short circuit between C and E at one of the transistors, with the multimeter probe, while I was taking some measurements and while was connected a 6-7A load at the output. After the short circuit I tested the power supply and it worked good. If the power supply works good, is there any recommendation for replacing the transistors or other components ?

[MarkF]

If it works?  -  Don't fix it.

[mike_mike]

Thank you for your help.
I observed that sometimes, at random moments of time the voltage is decreasing without any reason from 26.1V to about 24.3V. This happens randomly, and it happens with load and without load at the output of the power supply.
So I tested with the scope - again - and I attached some of the screenshots.
I found that the voltage is sometimes coming back to the initial value if I move the PCB. I checked the connections on the PCB and they are good.
The screenshots were took when the power supply had 26.1V at the output. I did not managed to check with the scope when the voltage was 24.3V.
Why the voltage drops suddenly ?
Are the screenshots good ?

And sorry for bothering you with my noob questions
Please help me. It is about 2 months since I started this project and I still have problems with it and I know that it is a simple project.

[mike_mike]

I found that sometimes the voltage at the output of the power supply drops suddenly from about 19V to about 17V without any reason. This happens is both situation - with load and without load.
Sometimes, if I move repetitive the PCB, then the voltage comes back to 19V.
What should I do to solve this problem ?
Please have a look at the above message.

[Zero999]

There's no reason why that would happen, other than overheating/load. Are you sure about the connections? Try moving and shaking the PCB with the power on and no load connected.

[mike_mike]

I checked the circuit and the LM317 and the power transistors are not overheating I can hold relaxed the hand on the case of those components.
The behavior is the same with and without load if I shake the PCB, sometimes the voltage drops and sometimes it comes back to the initial value.
I checked the connections and they seems to be good.
I used a 1.5A load at 19.87V for the last tests.
I checked the voltages while the output was 19.87V and while the output dropped to 17.87V and I found that in both cases, the input voltage in LM317 and the voltage on the filter capacitors was the same. The problem is on the output of the power supply, because there the voltage dropped. I also tried with different LM317AHV and the results were the same.

[floobydust]

It's a bit puzzling because it is an intermittent problem.

I would usually suspect the parts on the ADJ pin like a noisy potentiometer or leaky C5 (rated for >50V?).

The LM317 might have gotten damaged if you shorted C-E (input-output) earlier.
Careful because a damaged LM317 likes to smoke and burn the potentiometer- the ADJ pin can stay at high voltage even though the pot is towards zero ohms and then it smokes. Try jumper across the pot and see if you get 1.25V out.

A pass-transistor TIP36C could also get damaged with a C-E short. I would test them out of circuit.

If this isn't it this stuff, then worst-case the circuit is oscillating.