DIY 0-30V 0-3A ...Again

[Jwillis]

I know this darn project has been done to death and I apologize for bringing it up yet again.
I've poured over all the posts here and across the Net  concerning this project and have found nothing related to my problem.
I understand that it's unlikely to achieve the advertised output  of 3 amps without over heating and possibly destroying the circuit.That would be more impressive than what I getting right now.
With no load it does indeed supply 0-31VDC. The current control seems to work as it effects the voltage and the LED does change off and on.
My problem is despite all attempts,checks ,rechecks and more rechecks I can't get more than 30ma.On load test at full voltage and full amps all it does is 30mA and drops the voltage right down to 0.8V.Current can be adjusted between 0 and 30mA with delicate movement of the potentiometer.
The transformer is a Hammond   120VAC to 24VAC 350VA .Now this should be plenty to power this PSU.
I purchased new TL081 op amps from Digikey and  tried 3 different sets. No change on load
I replaced the filter Cap from 3300uFwith 4700uF.No change on load
Replaced all diodes including zener .No change.
All diodes have checked OK
Removed Q1 suspecting it may be faulty.No change on load.
Changed Q2 from 2SD882 to a BD139.No change on load.
Checked the  Q4 2SD1047 and it checks OK.
Replaced Q1 (9014) and Q3 (9015) .No change
All resistors check within 1%.
All diodes have checked OK
All potentiometers work as they should.
I've checked all voltages across pin4 and pin7 supplying the op amps .All around 32 volts.
I checked the zener diodes and get the expected  5.1Vdc across D7 and -5.1Vdc  across D8.
I load tested pin6 on U2 and get 600mA at 31V which should be enough to drive Q2 with the (R15)1K limiting resistor.Which should give 31mA at 31Vdc.
Checked all traces and compared to working PCBs and found no difference aside from pattern.
Checked and rechecked orientation of components.
Checked continuity of traces and all solder joints.All functioning.

The only thing that seems odd is that RV1(100K Trim pot) doesn't do any thing .
Despite all this I can't get more than 30mA.
Frustrated and out of ideas.

[Andy Watson]

Double check the value of R7.
If it is correct, load the output to 30mA (or whatever the maximum attainable is) and report the voltage across R7, and at pins 2 and 3 of U3 (relative to 0V side of R7).

[Kleinstein]

The circuit is known from Chinese kits with several known issues.
The most important is that with a 24 V transformer the supply for the OPs U2 and U3 will be too high (e.g. nominally some 38 V, worst case maybe more than 42V)  So expect these to blow.
Even if the OP survive (lucky or using other OPs like 741 , TLE2081) one would not get 30 V out under a significant load. With a 24 V transformer its more like 20-24 V. The current rating is more like short time peaks and should be reduced to something like 2 A.
To keep the original OPs in spec, the transformer should be no more than 18-20 V AC and thus some 16 V maximum output voltage under load.


For the broken circuit, having 31 V at the supply to the OPs suggests the negativ (-5 V) supply is not working. This could be a wrong part (e.g. zener D7 the wrong way around) or a broken OP, drawing to much supply current.
600 mA out of U2 suggest it's broken. Without the negative supply Q1 blocks any significant current - a little odd to still get 30 mA. 30 mA would be more like the normal current limit out of U2.

[IanB]

This may seem obvious, but if we take the transformer at face value, then with a supply of 24 V AC, the maximum output current at 33 V DC using a perfect rectifier will be zero amps. If we subtract some diode drops in the rectifier and the necessary voltage drops across the regulation circuitry, then the limiting voltage for zero amps out will drop down to 30 V or so. If you are actually managing to get 80 mA at 31 V DC I would say you are doing remarkably well. It is way better than zero which is all you have any right to expect.

[Jwillis]

Oh wow! Thanks for the reply s

Double check the value of R7.
If it is correct, load the output to 30mA (or whatever the maximum attainable is) and report the voltage across R7, and at pins 2 and 3 of U3 (relative to 0V side of R7).

Thanks Andy.I checked R7 and pis 2 and 3 on U3 as you said. R7 is 47.3 Ohms and under max load (a whole 30mA) it reads 1.667 Vdc
Pin 2 is 1.672Vdc and pin 3 is 1.723Vdc

The circuit is known from Chinese kits with several known issues.
The most important is that with a 24 V transformer the supply for the OPs U2 and U3 will be too high (e.g. nominally some 38 V, worst case maybe more than 42V)  So expect these to blow.
Even if the OP survive (lucky or using other OPs like 741 , TLE2081) one would not get 30 V out under a significant load. With a 24 V transformer its more like 20-24 V. The current rating is more like short time peaks and should be reduced to something like 2 A.
To keep the original OPs in spec, the transformer should be no more than 18-20 V AC and thus some 16 V maximum output voltage under load.


For the broken circuit, having 31 V at the supply to the OPs suggests the negativ (-5 V) supply is not working. This could be a wrong part (e.g. zener D7 the wrong way around) or a broken OP, drawing to much supply current.
600 mA out of U2 suggest it's broken. Without the negative supply Q1 blocks any significant current - a little odd to still get 30 mA. 30 mA would be more like the normal current limit out of U2.

Thanks Kleinstein. I don't expect much from this kit without modifications. But was hoping to get at least 1 Amp.So far I've retested those Zeners and they both read 5.1V and the orientation is correct.Double checked all the diodes for orientation.
The Ops are all fine .I've switched them out several times with the same results.

What value is R15? Is D10 fitted the correct way?

Thanks xavier. I recheck D10 and its the correct way. R15 is 1K ohms But I found another schematic where R15 is 100 Ohms and this person claims a much higher current output.
http://electronics-lab.com/community/uploads/monthly_08_2014/post-107142-14279144694286.jpg given the extra power transistor I don,t doubt it.
I was kind of leaning toward the idea that R15 may be to high but the circuit is identical to others that function fine.

[IanB]

The key thing is can you get a reasonable current at lower voltages. For instance, what can you get at 24 V, or at 12 V ?

[Zero999]

audioguru made some modifications to that project many years ago, but the webmaster on electronics lab (that site has plenty of other broken/bad projects/schematics) didn't bother updating it. I strongly recommend implementing them.

[Kleinstein]

.....
Thanks Andy.I checked R7 and pis 2 and 3 on U3 as you said. R7 is 47.3 Ohms and under max load (a whole 30mA) it reads 1.667 Vdc
Pin 2 is 1.672Vdc and pin 3 is 1.723Vdc

At 47 Ohms R7 would be way too large - it more like 0.47 Ohms to be right. So the 30 mA could be about right given the resistor is high by a factor of 100.

[soldar]

R7 is 47.3 Ohms and under max load (a whole 30mA) it reads 1.667 Vdc

R7 should be 0.47 Ohm or 0.5 Ohm. A value too high will limit the current output sooner.

[Jwillis]

.....
Thanks Andy.I checked R7 and pis 2 and 3 on U3 as you said. R7 is 47.3 Ohms and under max load (a whole 30mA) it reads 1.667 Vdc
Pin 2 is 1.672Vdc and pin 3 is 1.723Vdc

At 47 Ohms R7 would be way too large - it more like 0.47 Ohms to be right. So the 30 mA could be about right given the resistor is high by a factor of 100.

R7 is 47.3 Ohms and under max load (a whole 30mA) it reads 1.667 Vdc

R7 should be 0.47 Ohm or 0.5 Ohm. A value too high will limit the current output sooner.

You know you guys are right I neglected to notice that .The schematic even requires a  0.47 Ohm  not the 47 ohm that was supplied.Damn my poor vision.Now that explains whats going on.
A couple 0.25 in series should be alright to test.

Edit: I have a variac so Would it be ok to set a lower voltage at the primary with the Variac just for testing?

[soldar]

I have a variac so Would it be ok to set a lower voltage at the primary with the Variac just for testing?

I can't see how that helps any. You wer using the stated 24 Vac and it was working well except for the current being limited. Why do you want to change something that is working as it should?

[Audioguru]

You bought a cheap Chinese kit that has all the errors of the original Greek kit plus the value of R7 is 100 times wrong!

[Jwillis]

I didn't bother with the Variac. Anyway I changed the 47 Ohm for (2)  0.25 in series (0.5 Ohm ). No change to current output of 30mA and voltage is now 3mV instead of zero to 31V.Darn!
Still have full range of voltage without load but it still drops to 200mV at load full available current (30mA).
Voltages across pin 4 and pin 7 for U2 remains 31 volts with or without load. U3 and U1  those same pins drop to 28 volts with load.
Going to attempt a lower value at R15.

You bought a cheap Chinese kit that has all the errors of the original Greek kit plus the value of R7 is 100 times wrong!

I checked the board for any errors in the traces and they match all working products.Most of the components have already been changed to fully tested ones from reputable dealers.Aside from the error with R7 that is partly my fault for not paying attention.Besides, changing R7 did little . 

[IanB]

Still have full range of voltage without load but it still drops to 200mV at load full available current (30mA).

I do not think you have not clarified whether this restriction in current applies at all selected voltages, or just at the maximum voltage?

[Audioguru]

You were using the stated 24 Vac and it was working well except for the current being limited. Why do you want to change something that is working as it should?

That circuit cannot survive since many parts have too much voltage and current. The TL081 opamps have a maximum allowed supply of 36V but the 24V transformer produces a peak of 34V plus the 5.1V negative supply plus maybe 2V more with low load current= 41.1V. Then POOF go the opamps.
The circuit has no voltage and current calibration trimpots.
The main filter capacitor is WAY too small.

My improved circuit uses a 28V or 30V 4.2A transformer, 44V opamps that use a 1.3V negative supply instead of the higher negative voltage needed for the old TL081 opamps, two output transistors to share the heat, some larger resistors to survive the heat, and calibration trimpots. It produces a regulated 30VDC at 3A as it should.

[Jwillis]

Still have full range of voltage without load but it still drops to 200mV at load full available current (30mA).

I do not think you have not clarified whether this restriction in current applies at all selected voltages, or just at the maximum voltage?

Oh Sorry Ian .The restriction is is the same at all positions of the voltage pot.At maximum load with the current 30mA of the voltage can be adjusted between 0 and and around 200mV
It's acting like Q2 and Q4 aren't  turning on at all and all the current is being drawn from U2 which just doesn't make sense .
I have some 3055 s to replace the  2SD1047 if need be.
I'll take a fresh look at it again tomorrow .It's probably something stupid .

[IanB]

Oh Sorry Ian .The restriction is is the same at all positions of the voltage pot.At maximum load with the current 30mA of the voltage can be adjusted between 0 and and around 200mV
It's acting like Q2 and Q4 aren't  turning on at all and all the current is being drawn from U2 which just doesn't make sense .
I have some 3055 s to replace the  2SD1047 if need be.
I'll take a fresh look at it again tomorrow .It's probably something stupid .

OK, thanks for the clarification. But I'm still having a little trouble following what you are doing.

The normal way to test such a power supply would be to set the voltage with no load, set the current limit to maximum, and then increase the load gradually and watch what happens to the voltage. When the voltage starts to dip or lose regulation you have reached the limit on how much current can be supplied.

A common way to adjust the current limit would be to short the output terminals and then turn the current adjustment until it is at the current you want. When doing this it would be normal for the output voltage to be close to zero (since the output is shorted).

[Audioguru]

The driver and output transistors are simple emitter-followers. The output transistor has an hFE of at least 25 and the driver transistor's hFE is at least 80. Then the total current gain is 2000 and only 3A/2000= 1.5V will appear across R15 if Q1 is properly turned off. U2 will have a max output current of only 1.5mA.

Maybe the voltage at pin 3 of U3 is not about 1.5V when the current-setting pot P2 is set to 3A to match the 3A x 0.5 ohms= 1.5V of your R7. Because any current more than 3A will cause the output of U3 to go low enough to reduce the output voltage and current.   

[Cliff Matthews]

Note: the garbage pots from Banggood could be a source of a lot trouble (at least they were for me).
I cut-and-pasted a side-by-side schematic to help future modders.

[Mate_Well]

•Actually, i was very interested in this project some time ago, after digging into it, i found there is a great document ( in French and it is not idealy translated but it is understandable into english), it has very nice measurements, facts, and additions to it https://drive.google.com/open?id=1qaKc0H9951NUQiRNrA4dNyz8iehMwrSn (it has the most stable part list/diagram from all that are floating arround, i hope this will help out (also  i did not tryed to build it yet)


•I will also try to make it much more powerfull with more powerfull transformer, which leads to more powerfull opamps, and other parts, air coling and such, i will do it long time from now on (i need to save up for the parts sadly) 

[Richard Crowley]

•Actually, i was very interested in this project some time ago, after digging into it, i found there is a great document ( in French and it is not idealy translated but it is understandable into english), it has very nice measurements, facts, and additions to ithttp://electronics-lab.com/community/applications/core/interface/file/attachment.php?id=42970 (it has the most stable part list/diagram from all that are floating arround, i hope this will help out (also  i did not tryed to build it yet)


•I will also try to make it much more powerfull with more powerfull transformer, which leads to more powerfull opamps, and other parts, air coling and such, i will do it long time from now on (i need to save up for the parts sadly) 

This attachment is not available. It may have been removed or the person who shared it may not have permission to share it to this location.

[Microcheap]

I am well aware of the limitation of this circuit but I've assembled this project over a decade ago when I started learning electronics, and the psu still works, of course I never pushed it to its limits, I think the only time was when I left it overnight recharging an old lead acid battery at around 13V@2A, and it survived. The only thing I remember I had to replace 1 or 2 times were transistors Q2 and/or Q4 but I was using Chinese knock-off transistors with improper heat-sink.

Back to the original question, have you checked Q1, R13 and R14 already? As Audioguru pointed, Q1 must be off, test the Vbe of Q1, it should be lower than 0.7V (I measured 0.2V in my board, with R13=10k and R14=1k5).

[Mate_Well]

•Any chance the u3 is damaged, or has wrong pins conection (could be caused by putting the chip in 180°rotated also) ?

link fixed

[IanB]

This attachment is not available. It may have been removed or the person who shared it may not have permission to share it to this location.

I am able to download it from Google docs, however it is 5 MB so it is too big to attach here.

[Jwillis]

Hello folks.sorry for not getting back sooner.
All the information you guys provided is a great help.

•Actually, i was very interested in this project some time ago, after digging into it, i found there is a great document ( in French and it is not idealy translated but it is understandable into english), it has very nice measurements, facts, and additions to it https://drive.google.com/open?id=1qaKc0H9951NUQiRNrA4dNyz8iehMwrSn (it has the most stable part list/diagram from all that are floating arround, i hope this will help out (also  i did not tryed to build it yet)


•I will also try to make it much more powerfull with more powerfull transformer, which leads to more powerfull opamps, and other parts, air coling and such, i will do it long time from now on (i need to save up for the parts sadly) 

Thank you  Mate_Well ,I'll call a friend to see if she can translate that PDF file for me.
I'm currently testing  the ops to see which one failed .The voltage adjust is fine but lost current control.I suspect U3 as you mentioned .

Note: the garbage pots from Banggood could be a source of a lot trouble (at least they were for me).
I cut-and-pasted a side-by-side schematic to help future modders.

Cliff Matthews I had the same problem at the start and changed those to 2W 10K for testing.


So I pulled all the Ops to measure the voltages at pin 4 and pin 7 in relation to output ground.For U1 I can't identify a true zero point to compare the values of pin 4 which should be negative and pin 7 which should be positive for any of the Ops .So from what I understand the TL081 requires a supply of 7 V ~ 36 V, ±3.5 V ~ 18 V  (single or dual) So I just took a reading pin 4 to pin 7 and got 33.8Vdc.Since pin 4 looks like it gets a variable voltage off of U3 pin 2.That's of course a guess on my part. Taking a reading at pin 4 on U1 from ground gives 0Vdc.Pin 7 I measure 33.5Vdc.
For U2 pin 4 reads from output negative for pin 4  -5.1Vdc and pin 7 33.8Vdc . Across from pin 4 to pin 7 is 38.9Vdc.Which is to high.
For U3 pin 4 reads from output negative for pin 4   -5.1Vdc and pin 7 33.8Vdc. Across from pin 4 to pin 7 is 38.9Vdc.Which is also to high.
Since the additive values of pin 4 and pin 7 can't exceed 36 V,I wondering if theirs any reason why I can't  incorporate  zener regulators for U1 ,U2 and U3 to bring  pin7  voltage down so the additive voltage of pin 4 and pin 7 won't exceed 36 V ? Could even bring D7 down as well since it really only needs to be >3.5V