Bench power supply problem

[chrish]

Hello Guys,

I normally hang out in the tube amp forum, but have a specific power supply issue. My 16 year old son has started to develop an interest in electronics/computers/mechatronics. For his birthday, I bought him a bunch of components, a multimeter, second hand scope and the parts to make a bench power supply. I used this schematic from the Eliot sound products site Variable Dual Lab Power Supply



We built this part of the power supply with the same values/components as the schematic and it works fine. We then decided that variable current limit would be a useful feature, so we found tis schematic Adjustable Current Limit For Dual Power Supply Circuit Diagram

Values:
R1 R2 1.2R
R3 R5 120R
R4 22R
P1 100R

We built the current limit circuit (the first part of the schematic) to insert prior to the above voltage reg, but it does not work. To troubleshoot, it would help if I understood the circuit, but I must admit I am confused as to what exactly is going on with the voltages/currents and what the regulator it trying to do with the 1.25 volt reference.

For info, I could get the IC1 and IC2 LM317/337 working OK with just the 1.2R sense resistor to limit the rails to 1 Amp without the IC3 circuit connected. I could get the IC3 circuit working independently to limit current from 10mA to 60mA when connected across a 10 volt supply, but (from memory as I am away from home with work at the moment) could not get it to work when connected across the +/- rails. It is worth noting that the +/- rails are about +/- 34 volts unloaded. I note that the LM317 data sheet states 40 volt max across the chip, might this be the problem? (If it is, the variable current limit article states that the supply voltages should be at "least +/- 22 volts").

Any help understanding the circuit or obvious problems would be greatly appreciated.

Regards,

Chris

[thaumaturge]

First thing I see is that you lifted the ground refference pins on the 317's. See where VR1 a/b connect to ground? Your circuit attemps to refference each from the other rail. Don't know why you followed the 15V regs with the five volters, each will tollerate roughly the same input. So if you want regulated +/- 5 as well as +/- 15 then just paralell the inputs to each. That said, don't know how the +/- regs will like having their input voltage reduced by the current reg section. But it might work.
Doc

[chrish]

Thanks Doc,

Perhaps I should clarify. I built the supply in the first picture with the variable voltage. It works fine. I then tried to insert the variable current section of the second picture, that is IC2, 2 and 3 and their associated resistors to the variable voltage supply. So, I have taken the output of the top picture at the 4,700uF filter caps to "+Uin" and "-Uin" of the bottom picture and then taken the output of the current reg at the junction of R1/R3 and R2/R5 and taken them to the inputs of the LM317 and LM337 of the top picture. I have not implemented the fixed 15 and 5 volt supplies in the bottom picture.

Cheers,

Chris

[magnoman]

Chris,

My take is that IC3 is configured as a current source established by the 1.25V reference divided by the sum of R4 and P1. This current puts a starting bias across both R3 and R5. When this bias along with the load current x R1 (R2) reaches the 1.25V reference for IC1 (IC2) then that series regulator will start to drop the output voltage in order to maintain this 1.25V limit.
That is prior to getting to this 1.25 reference IC1 should be running "full on" such that its output equals the input less it compliance voltage of a few volts.
By changing P1 you change the initial bias or the amount the R1 x I load as to develope before the output starts to drop (current limit).

Hope this helps
-Antonio

[chrish]

Thanks Antonio.

Individually, I think I understand how the LM317/337 limits current by maintaining 1.25 volts across a current sense resistor. I can see how varying the 100R resistor from 0 to 100R on IC3 gives a combined resistance of 22R to 122R varying current from roughly 60mA to 10mA. I am having difficulty understanding how this biases IC1 and 2. Lets say that we set P1 to give a current of 10mA. This should give a voltage drop of 1.2 volts across R3 and R5 (as only a few uA of current flows from the Adj pin of the 317/337)? If we increase this to 60mA, the drop across R3 and R5 would be 7.2 volts? As I understand it, IC1 tries to maintain 1.25 volts across R1+R3. If the voltage is <1.25 volts, no limiting. If the voltage is > 1.25 volts, it will limit its output to reduce the voltage back to 1.25 volts.

I am probably looking at this wrong or making a wrong assumption, but I would have thought you would want to bias from 0 (for all current sensing coming from the 1.2R resistor R1) for a 1 Amp output to a bias of 1.25 volts that would cut the current off completely, not the 1.2v-7.2v bias applied?

[magnoman]

Chris

Well in looking at the actual values I can see your point. Seems that the pot should be 100 ohms minimum, even 100 ohms should cause the series parts to limit with no additional load current.
Larger values on upto 1K (roughly 1ma current, and still much larger than the adjustment pin leakage current) would seem to be near the built in LM317 protection. Decreasing R3 should have the same effect but unnecessarily consume more power.

Hope this helps
-Antonio

[chrish]

Yes, it helps!

Thanks Antonio. I was trying to work out where I was going wrong with my thinking. Looks like there might be an error in the schematic values.

Since it looks like you confirmed I was thinking on the right track re bias, I was thinking that something like this might work:



R1 and R2 are 1.2R to limit output to 1A. IC2 sets a current of about 15mA with R3 at 82R. R5 and R6 each drop about 15 volts to keep the voltage across IC2 at about 34 volts (+/- rails are +/- 34 volts) (or perhaps 15v zeners might be more appropriate here?). LEDs drop about 1.7 volts, so voltage divider dual potentiometer (R4) can set a bias of 0-1.7 volts to the adjust pins of the LM317/337 IC1/3.

Am I on the right track?

Thanks for the help so far.

Regards,

Chris

[AndrewT]

What about replacing Vr1 & Vr1 with switched resistors for a range of fixed current limits.
A 2pole 6 position rotary switch would parallel a fixed value of say, 100r. With the switches open circuit, position 0, the current limit would be ~12mA,
then add progressively lower resistances in parallel to that 100r to reach your maximum limit of ~1A.

This allows the Vr4a and Vr4b to be omitted.

Now go back to post1, top figure and use that pair of voltage regulators.

Take care a range of 1v2 to 25V will require a 25Vac transformer and that could give 38Vdc as the input to the current regulator. The Vdrop of the voltage regulator could approach 35V and this combined with a 1A output would result in each voltage regulator dissipating 35W. That is very hot !!!!

[chrish]

Thanks for the suggestion Andrew. I had thought of that, but it is not quite 'the elegant solution' I had hoped for. Someone else on the EEVblog forum suggested something like this:



Yes, big power dissipation on voltage regs, I have a big heat sink! Will breadboard above in a few days when I get the LM385 1.24 volt reference devices and report back.

Thanks for the help and suggestions!

Chris

[magnoman]

Chris

Except with the approach recommended by Andrew, another issue you have to watch out for is that the bias current which sets the lower current limits does not decrease when the outputs are in full current limit. That of course would allow the current to increase.
You have to protect against both outputs being shorted together, while at the same time ensuring you have enough bias left across what ever is setting the current limit. This will depend on what regulators are used downstream (how much voltage / current they will produce under shorted conditions).
There are alternatives (one can always reference to the input side of IC1 and 2) but again depending on the regulators characteristics you may not need to, but please have a look at this case.

Hope this helps
-Antonio

Edit: rushing out the door, hope this is still somewhat descriptive.