L200 Bench Power supply.

[hackableFM]

I built a bench supply using a typical L200 diagram which is commonly available across the net, but I've used THREE L200's now and I am pulling my hair out now trying to get this working.

I'll try to describe this circuit as accurately as possible...

I'm using a 24volt (4amp I believe) torroidal transformer fed through a full wave diode bridge consisting of 4 very chunky diodes which I think can handle the 4 amps. This is then fed to a 6800:mu: f 50wv capacitor. The current from this is then fed to an L200C regulator directly,the regulator is mounted on a very large heatsink (Which at no point has even been warm!). The external control components used for the L200 are 1khm: resitor, 10khm: pot, 0.33hm: Ilimit resistor and a 10:mu: f cap and a '104' cap on the output of the regulator. I know this is probably not relevant but the output is fed through an ammeter and on to a voltmeter, Both are analogue 'needle' type meters.

On first switch on I thought I had struck gold as it appeared to work immediately with no problems, The voltage from the rectifier to the regulator is around 36.4 volts DC and I was getting in the region of 3 - 30 Volts from the PSU. I only noticed there was a problem when I connected any load to the output, I tried pulling approx 12volt @ 200mA from it and the voltage dropped off to approx 5 volts, It appeared the regulator had reached the set current limit??? After a few checks I decided that the only thing causing this could be the L200 regulator so I changed it. It worked fine again and this time I immediately connected the same load (Approx 200 mA) and it worked well, I then spent 10 minutes using the power supply for the 'other project' I built this unit for, (Testing & setting up a couple of lead acid battery monitors based around the LM3914 bar chip.) Anyway, I was happy as larry at this point, I thought I was finished for the night but I then started to 'play' with the PSU I had just built. I stupidly connected a large solenoid to the supply which gave off a few sparks which promptly blew hell out of the L200 regulator, My output voltage shot up the the full 36volts which was being fed to the regulator. Obviously the regulator had gone short circuit. I again changed teh L200 and again all worked fine.

Now, I connected a load (Approx 200 - 300mA) to the PSU and decided to leave it running for a while as a soak test if you like. All went well for approx 25 - 35 minutes after which the LED's on the load began to 'blink' dull then immediately go back to bright, this would happen every second or two... I watched of the next 20 minutes and the problem got worse, The LED's eventually got duller and duller until they were hardly light at all while still blinking even duller occasionally. Now obviously the output voltage from the PSU was dropping gradually and again I feel the regulator was going into it's current limit routine.

I'm totally scratching my head as to why this thing is doing this, Obviously the second L200 was my fault out of stupidity but why is it behaving in this was especially after it worked well for a short period, Surely this says that the actual build quality is ok since it worked for a while?

Can anyone shead any light on this for me please as I am stumped by it now.

Thanks for taking the time to read this...

hackableFM.

 

[ericgibbs]

hi,

>> Now, I connected a load (Approx 200 - 300mA) to the PSU and decided to leave it running for a while as a soak test if you like. All went well for approx 25 - 35 minutes after which the LED's on the load began to 'blink' dull then immediately go back to bright, this would happen every second or two... I watched of the next 20 minutes and the problem got worse, The LED's eventually got duller and duller until they were hardly light at all while still blinking even duller occasionally. Now obviously the output voltage from the PSU was dropping gradually and again I feel the regulator was going into it's current limit routine.

What was the ouput voltage when you had the LED [200/300mA] load connected.???

 

[mvs sarma]

Hi,
A picture speaks 100 words

try posting a circuit diagram. it would be easy to discuss.

 

[mvs sarma]

I built a bench supply using a typical L200 diagram which is commonly available across the net, but I've used THREE L200's now and I am pulling my hair out now trying to get this working.

I'll try to describe this circuit as accurately as possible...



Can anyone shed any light on this for me please as I am stumped by it now.

Thanks for taking the time to read this...

hackableFM.

what actually did you mean by -- used three L200 now.

did you replace the chip twice or your design has thee circuits paralleled?

2. Did you ensure that the wires (both on +ve and return)used are of proper gauge, including those at rectifier to the regulator input - as your current sense resistor is at 0.33 ohms?

 

[Hero999]

Please post a schematic in case there's anything you've missed in your description.

The only thing I can think of is the safe operating area protection is kicking in. When Vin-Vout > 16V the output current limit starts to fall to protect the regulator. In your case the voltage on the capacitor is 36.4V and the output voltage is 12V, therefore Vin-Vout = 24.4V which would limit the output current to 1A if the regulator is cold or 800mA if it's hot. I know you were only drawing 300mA but please note that these characteristics are typical, some regulators might cut the current limit at a lower Vin-Vout than others and that it's temperature dependant. This would explain why the first regulator you tried didn't work and the second one did but started to fail after half an hour, as it warmed up.

**broken link removed**

I assume the regulator broke when you connected the solenoid because the back EMF caused the output voltage to become higher than the input voltage for a short period of time. You can protect against this be connecting a diode from pin 2 to pin 1 (anode going to pin 2 and cathode to pin 1), this would provide a current path round the regulator so the transient would be absorbed by the input capacitor.

 

[hackableFM]

Wow.. excellent responce, sorry for my late reply but I've only just got in from work.

I'm gonna try to answer all of the questions in one reply!

What was the ouput voltage when you had the LED [200/300mA] load connected.???

The output voltage was 12Volts as it should have been set by the pot. The voltage did drop off to approx 5 - 7volts as the fault appeared.

Hi,
A picture speaks 100 words

try posting a circuit diagram. it would be easy to discuss.

Welll a full circuit diagram isn't actually available as such but I can point you **broken link removed** for the diagram I used in setting up the regulator part of the PSU, even then I missed out C1 as I thought it wouldn't be necessary considering I have used a nice 6800:mu: f capacitor for smoothing. Now basically I built a basic rectifier (4 diodes) and a nice big smoothing capacitor as mentioned and fed this into the diagram as posted.

what actually did you mean by -- used three L200 now.

did you replace the chip twice or your design has thee circuits paralleled?

2. Did you ensure that the wires (both on +ve and return)used are of proper gauge, including those at rectifier to the regulator input - as your current sense resistor is at 0.33 ohms?

Well I blew the regulators up and replaced them twice.

The wires I used are ample for the supply of 2 amps @ 30V, I dont actually have any idea of the guage but they are from a PC PSU and are plenty to handle the current. (I hope!)

Please post a schematic in case there's anything you've missed in your description.

The only thing I can think of is the safe operating area protection is kicking in. When Vin-Vout > 16V the output current limit starts to fall to protect the regulator. In your case the voltage on the capacitor is 36.4V and the output voltage is 12V, therefore Vin-Vout = 24.4V which would limit the output current to 1A if the regulator is cold or 800mA if it's hot. I know you were only drawing 300mA but please note that these characteristics are typical, some regulators might cut the current limit at a lower Vin-Vout than others and that it's temperature dependant. This would explain why the first regulator you tried didn't work and the second one did but started to fail after half an hour, as it warmed up.

**broken link removed**

I assume the regulator broke when you connected the solenoid because the back EMF caused the output voltage to become higher than the input voltage for a short period of time. You can protect against this be connecting a diode from pin 2 to pin 1 (anode going to pin 2 and cathode to pin 1), this would provide a current path round the regulator so the transient would be absorbed by the input capacitor.

Now, I think I get what you're saying about the safe operating area thingy, BUT I would have thought it would supply 12V and only a couple of hundred milliamps??? It's bloody useless if it doesn't and I've wasted my time building the thing lol.

You assume correctly about the back EMF blowing the poo out of the regulator, Now I never thought about the diode across pins 2 & 1 but rest assured I will be installing it if I can sort out the main problem and actually make this PSU useful!

Thanks for the replies so far!

hackableFM...

 

[Hero999]

Now, I think I get what you're saying about the safe operating area thingy, BUT I would have thought it would supply 12V and only a couple of hundred milliamps??? It's bloody useless if it doesn't and I've wasted my time building the thing lol.

You assume correctly about the back EMF blowing the poo out of the regulator, Now I never thought about the diode across pins 2 & 1 but rest assured I will be installing it if I can sort out the main problem and actually make this PSU useful!

Does the transformer have a centre tap?

Are you getting the output by connecting the two coils in series?

You can lower the input voltage when you set the output voltage to a lower setting.

You could add a resistor in series with the input of the regulator to lower the voltage as the current increases for the lower voltage settings. If you choose this option, you'll have to be careful with the value you choose and make sure it's rated for the current you plan to draw.

 

[hackableFM]

Does the transformer have a centre tap?

Are you getting the output by connecting the two coils in series?

You can lower the input voltage when you set the output voltage to a lower setting.

You could add a resistor in series with the input of the regulator to lower the voltage as the current increases for the lower voltage settings. If you choose this option, you'll have to be careful with the value you choose and make sure it's rated for the current you plan to draw.

No, the transformer doesn't have a centre tap, I have already thought about this as an option for reducing the input voltage to the regulator for lower output voltages but that option is not available.... This transformer was ripped out of a fixed 4A Regulated PSU so at least it's not like I went out and bought it.

I like the idea of the resistor on the regulator input but I truly can't bothered with what seem like workarounds, I just want a variable PSU that actually works for general test purposes ideally 3 - 30 volts at an amp or two....

Has anyone got any recommendations for this?? Maybe I can rip out the L200 jobby and use the existing rectified & smoothed DC to run some other adjustable psu?

 

[hackableFM]

Here you go chaps, this is pretty much the Schematic for the PSU I have built!

 

[hackableFM]

Right.....Here's a question for you's........

THIS datasheet as kindly posted by Hero999 **broken link removed** On page 8 there's a diagram for a high current PSU with a 'Pass' transistor, Now, If I modify my circuit to include this 'pass' transistor will it sort out my problems? If so what is the value or Rsc???

Thanks for your help!

Fatty. ops I mean hackableFM...

 

[mvs sarma]

what is the spec of your psu?

 

[eblc1388]

The current from this is then fed to an L200C regulator directly,the regulator is mounted on a very large heatsink (Which at no point has even been warm!).

Cold heatsink is not a good sign with 5W heat dissipation given out by the L200 under 12V 0.2A output loading.

Make sure thermal conduction compound has been used for the mounting interface between L200 and heatsink.

 

[hackableFM]

Spec of the PSU? Well if you mean what I want from it then basically I want upto 30volts (Ish) and an amp or two.

Heatsink compound hasn't been used BUT The regulator doesn't even get warm at all never mind the heatsink... Maybe I have been generous with the current I claimed to draw from this PSU at test, I was running a previous project as a load, The project was a home made PIC driven Digital Clock which uses 3 inch LED 7 Segmants to display the time. I ESTIMATED the 200 - 300 mA, Maybe it's much lower.

hackableFM...

 

[ericgibbs]

Spec of the PSU? Well if you mean what I want from it then basically I want upto 30volts (Ish) and an amp or two.

Heatsink compound hasn't been used BUT The regulator doesn't even get warm at all never mind the heatsink... Maybe I have been generous with the current I claimed to draw from this PSU at test, I was running a previous project as a load, The project was a home made PIC driven Digital Clock which uses 3 inch LED 7 Segmants to display the time. I ESTIMATED the 200 - 300 mA, Maybe it's much lower.

hackableFM...

If you want about 1 to 2 amps, use the circuit in the application notes that has a bypass power transistor.

 

[Hero999]

Right.....Here's a question for you's........

THIS datasheet as kindly posted by Hero999 **broken link removed** On page 8 there's a diagram for a high current PSU with a 'Pass' transistor, Now, If I modify my circuit to include this 'pass' transistor will it sort out my problems? If so what is the value or Rsc???

Thanks for your help!

Fatty. ops I mean hackableFM...

Rsc is the sense resistor that deturmines the short circuit current.

Any circuit that reduces the power dissipation of the regulator will help solve your problem.

Another thing I've noticed is that the maximum drop out voltage is 32V. This means that when you set the output voltage to less than 4V it could damage the regulator.

 

[hackableFM]

OK, thanks all, I will now revise my PSU and include the 'Pass' transistor, I may need to wait til after the holidays to get one but I will sort it and let you all know how I get on!

Actually, I have here a 2N3055 transistor mounted on a large heatsing already.... It came from the same fixed PSU I salvaged other parts from. Can this be used as the Bypass tranny?? If so what modifications will be needed to make this work?

Many thanks for your replies on this over the past couple of days, If you hadn't replied then I would probably have launched this PSU out the window by now!

hackableFM..

 

[Hero999]

Just use the circuit on the datasheet, any high power transistor should work.

 

[hackableFM]

Right now....

There's been a change of plan. I decided to build one of THESE instead.

I'm not doing too bad so far, I have been out and bought all the bits & pieces (Including all the necessities for photo etching PCB's) I have now printed the board and mounted most of the components onto it.

My problem is that I have not got a transistor (Q2) which is listed as a "2N2219". At the shop I bought the bits from the salesman forgot to check out an equivalent device for me since he didn't have one in stock....

Now I have built the majority of the PSU I am kinda stuck until I get this tranny sorted. I've searched about a bit in my junk box and found a few old transistors which I am hoping might do the job.

I have some 'BC108', 'BC109', 'ZTX300', 'BC182', 'BC212', 'BC557' and '2N4401' all old but unused. Can anyone answer if any of these will substitute the 2N2219 I need?

Thanks for any help on this!

 

[audioguru]

That 0V-30V/3A power supply doesn't make 30V at 3A. Its power transformer's voltage is too low. The main filter capacitor is too small. It makes about 25V at 3A with lots of ripple.

With a 3A load, its transformer, rectifier diodes and some resistors get overheated.
With a 3A load and a low output voltage or a shorted output then the little 2N2219 transistor gets way too hot and might melt. I recommended a TIP31A power transistor but a member simulated that a BD139 transistor will have a better transient response. They need a small finned heatsink.
The single 2N3055 output transistor frequently also overheats.
Without a load, its opamps operate at a supply voltage that is higher than their absolute max allowed voltage.
The offset voltage adjustment circuit is wrong.

Read in the forum about my fixes to it.

 

[hackableFM]

Ohhhh BUGGER!!

where are these fixes?? Can they be implimented on the already printed PCB??

Just for info, I am using 4amp Diodes for the rectifier and a 6800:mu: f main filter capacitor already.

It sounds like I have made another bad choice in building this pig!