pre regulator doesnt work when entering current limiting

[xReM1x]

hi, I'm using the LT1074 as a pre regulator and when I'm entering current limiting mod the LT3080 (post regulator) just regulate to 3V doesnt matter what voltage I set for current or voltage.

as you can see I set the Iadj to 5V but the voltage is at 1V so it doesnt enter the current limiting mod and it works great :

but when I enter current limiting well thats what is happeing: (Voltage set to 25V, current limiting is set to 1V(1A). the voltage at the output supposd to be 1V because Rload is 1R)

and the regulator output voltage is always 3V at current limiting no matter what.

when not using the pre regulator the circuit works good.

[Kleinstein]

The preregulator looks Ok. With this circuit its normal to have about 2,5-3 V headroom from the preregulator.

However the circuit around the LT3080 does not look ggod at all. Voltage regulation is likely prone to oscillation, at least with some more problematic loads (e.g. constant current source). Using an external transistor to boost the current is messing up the chip internal regulation loop. Usually the chip is too fast for most external PNPs.

The second problem is current regulation. Here the loop is external and likely not stable with many loads. Also C2 and R6 make current limiting very slow - so it may not provide short circuit protection.
C2 and R6 may also make U4 unstable.

[xReM1x]

The preregulator looks Ok. With this circuit its normal to have about 2,5-3 V headroom from the preregulator.

However the circuit around the LT3080 does not look ggod at all. Voltage regulation is likely prone to oscillation, at least with some more problematic loads (e.g. constant current source). Using an external transistor to boost the current is messing up the chip internal regulation loop. Usually the chip is too fast for most external PNPs.

The second problem is current regulation. Here the loop is external and likely not stable with many loads. Also C2 and R6 make current limiting very slow - so it may not provide short circuit protection.
C2 and R6 may also make U4 unstable.

removed pnp boost transistor, C2 redueced to 22uF and R6 the same 1K.
I will deal with all of those later, I just need to figure out the problem with th pre regulator. it is stuck on 1.5V now instand of 3V.

[Kleinstein]

I am afraid the current limiting needs a total rework. Just scaling down C2 a little won't save the circuit. U4 would likely need a lokal feedback if C2 is larger than about 100 pF.
Compensation of current regulation is better done around U3 instead of the C2 R6.

With current regulation there is also trouble if the preregulator goes below the set voltage IADJ. So a lower range like 0-2 V should be choosen here. The exaxt limit also depends on the current sense amplifier. Here one might need a negative auxillary supply or a higher setting for the preregulator.

Normally the shown circuit should give a dopout of something like 0.6 Volts plus 2 times Vref of the LT1074 - so something like 3 volts over the output voltage.

[xReM1x]

I am afraid the current limiting needs a total rework. Just scaling down C2 a little won't save the circuit. U4 would likely need a lokal feedback if C2 is larger than about 100 pF.
Compensation of current regulation is better done around U3 instead of the C2 R6.

With current regulation there is also trouble if the preregulator goes below the set voltage IADJ. So a lower range like 0-2 V should be choosen here. The exaxt limit also depends on the current sense amplifier. Here one might need a negative auxillary supply or a higher setting for the preregulator.

Normally the shown circuit should give a dopout of something like 0.6 Volts plus 2 times Vref of the LT1074 - so something like 3 volts over the output voltage.

shit goes real.
EEVBLOG Dave also used pre regulator and the same current limiting technique.
here is the schematic :
http://www.eevblog.com/files/uSupplyBenchRevC.pdf
you can see the dc-dc convertor . how did he do it?

[Kleinstein]

Daves DCDC Converter seems to set the preregulator with a separate EPOT. So the voltage is following the set voltage - which is not a good idea: if there is a short, the current limiting is reducing the voltage and thus the poor LT3080 sees a high current and high voltage. So it will soon go in SOA limiting or thermal shutdown.

So at this stage dave's circuit is not a good solution. Not ate the preregulator and also not at the current limiting part. As far as I understood Daves power supply project did not work out well because of this problem: voltage regulation is good (unless U3A is oscillation, which is well possible) and rather fast but current regulation is slow, possibly unstable. SO better not use this circuit as a basis for a supply.

If there is trouble with the preregulation, one should test it separately.
One point calling for trouble is that the LTC6101 needs at least 4 volts, and can only drive 1 V at the lower supply limit. As a quick fix changing R12 to about 5 K might work - however with higher voltage at the preregulator.
It's likely better to get rid of the LTC6101.

Adding current regulation the lt3080 to make a real lab supply is really difficult. So don't expect good and fast current regulation.

[xReM1x]

Daves DCDC Converter seems to set the preregulator with a separate EPOT. So the voltage is following the set voltage - which is not a good idea: if there is a short, the current limiting is reducing the voltage and thus the poor LT3080 sees a high current and high voltage. So it will soon go in SOA limiting or thermal shutdown.

So at this stage dave's circuit is not a good solution. Not ate the preregulator and also not at the current limiting part. As far as I understood Daves power supply project did not work out well because of this problem: voltage regulation is good (unless U3A is oscillation, which is well possible) and rather fast but current regulation is slow, possibly unstable. SO better not use this circuit as a basis for a supply.

If there is trouble with the preregulation, one should test it separately.
One point calling for trouble is that the LTC6101 needs at least 4 volts, and can only drive 1 V at the lower supply limit. As a quick fix changing R12 to about 5 K might work - however with higher voltage at the preregulator.
It's likely better to get rid of the LTC6101.

Adding current regulation the lt3080 to make a real lab supply is really difficult. So don't expect good and fast current regulation.

so should I just abonden the LT3080 and use other linear regulator or something like that?

[Kleinstein]

Using a voltage regulator chip and adding variable current limiting later is usullay a poor choice. Current regulation tends to be unstable with complex loads or very sluggish.

At high power it's usually better to have a regulator with two OP-Amps and discrete BJTs (maybe MOSFETs) as a power stage. So think about LM358 and transistors instead of LM317 or LT3080. There are mainly two types of circuit to choose from, that are found in commercial lab supplies:
1) Using a positive side BJT (e.g. NPN darlington) as a follower and low Side shunt. This type of circuit gets difficult at more than about 30-35 V. Many LM723 based circuits are of this type.
2) Using a floating auxilary supply for the regulator. E.g. using a NPN Darlington in Emitter circuit. This type is used in most HP supplies. The supply for the regulator itself is often conntected to the positive Output.

There are plenty of threads about lab power supplies.

[xReM1x]

Using a voltage regulator chip and adding variable current limiting later is usullay a poor choice. Current regulation tends to be unstable with complex loads or very sluggish.

At high power it's usually better to have a regulator with two OP-Amps and discrete BJTs (maybe MOSFETs) as a power stage. So think about LM358 and transistors instead of LM317 or LT3080. There are mainly two types of circuit to choose from, that are found in commercial lab supplies:
1) Using a positive side BJT (e.g. NPN darlington) as a follower and low Side shunt. This type of circuit gets difficult at more than about 30-35 V. Many LM723 based circuits are of this type.
2) Using a floating auxilary supply for the regulator. E.g. using a NPN Darlington in Emitter circuit. This type is used in most HP supplies. The supply for the regulator itself is often conntected to the positive Output.

There are plenty of threads about lab power supplies.

I saw the general purpose power supply thread here and I wonder how could I change it to make the current limit up to 5A?

[C]

The LM3080 is not a voltage controlled voltage regulator.

The LM3080 is a resistance controlled voltage regulator.

The resistor connected to the set pin becomes a high quality voltage reference that has low noise inside the LM3080. Note the +-0.3 Volt limit on the set pin. With the source and sense inside the LM3080 this is not a problem as long as what is connected to set pin is just resistance.

Current Limit:
You want the Pre-regulator to be X volts above LM3080's output. When you start current limiting reduce the X so that you are source current limiting. The LM3080's data sheet has an example of two LM3080 in series doing this.

Voltage Set:
The LM3080 controls the current to the set resistor. If the LM3080 turns off the current it expects the set pin to be a resistance to ground not some voltage.

C

[Kleinstein]

With the LM3080 one can use a voltage or a resistor to controll the output. Its at the users choice, both ways and combinations work. The controll input delivers a constant current, so a resistor to ground can be used to set the voltage. However it also word by setting the voltage directly from a (low impedance) voltage source.

Using an external voltage ignores the internal reference current.