Hello,
I am planning to build a few LM/UA723 based power supplies and I need to know how long this IC will still be manufactured by the producer ?
I already seen that some manufacturers (ST for example) has been discontinued this IC and I also seen that LM723CN has been discontinued at TI, while UA723 is still in production at TI.

This is the project: http://www.electronics-lab.com/project/3-30-v2-5-a-stabilized-power-supply/

 

It is impossible to predict. It is a very old design, originally from Fairchild (µA723), but still seems to be moderately popular. It could be around for many more years or announcement of end of life could come tomorrow.

Since TI acquired National Semiconductor there would be duplication of lots of common parts that both companies made, so it isn't surprising that some variants would be dropped.

 

Buy all the parts you think you'll ever need or use a newer part.

I think the only thing going for the LM723 is foldback current limiting. Unless that feature is being used, why use a part that's obsolete or quickly becoming so?

 

That is one of the lessons you have to quickly learn if you want to be in the electronics manufacturing business.

Every component you design into your product is obsolete or near end of production by the time you bring your product to market.

Make your designs adaptable so that you can substitute newer components into your design when required to do so.

 

You can talk to the manufacturer. TI usually gives warning if they are going to cease production of a part. You'll see the "Not for new designs" watermark on the data sheets of near end production parts. They may also suggest newer parts that are compatible.

How many is a "few"? Would it be worth it to buy extra to have as repair parts?

Lastly, what is the intended use for the power supplies? Can you use a different design, perhaps a switching supply?

 

What is the acceptable voltage drop under load for a LM723 based power supply ?
For example for this power supply: http://www.electronics-lab.com/project/3-30-v2-5-a-stabilized-power-supply/ ?

 

What is the acceptable voltage drop under load for a LM723 based power supply ?

That depends on the conditions under which the supply will be used.

 

What is the acceptable voltage drop under load for a LM723 based power supply ?
For example for this power supply: http://www.electronics-lab.com/project/3-30-v2-5-a-stabilized-power-supply/ ?

0.6% of Vout

 

@dl324 This applies even if there are more than 1 (for example 4) power NPN transistors ?

 

Hello,

Did you ever had a look at Tony van Roon's design for a 723 powersupply?


The description can be found in the attached PDF.

Bertus

 

@dl324 This applies even if there are more than 1 (for example 4) power NPN transistors ?

As long as the power transistors are in the feedback loop of the regulator.

 

As long as the power transistors are in the feedback loop of the regulator.

If the transistors are like in the attached schematic (in the feedback loop) ? The feedback loop is the loop that starts from pin 4 and goes through the transistors to the pin 10 ?

@bertus I found that schematic a few years ago, but I did not built it yet. And I did not read the documentation. I hope I will have time to read the documentation. Thank you.

 

If the transistors are like in the attached schematic (in the feedback loop) ?

Yes.

 

I calculated 0.6% × 19.97 = 0.119V. Is that correct ? The voltage could drop by maximum 0.119V ?

 

One of the things that limits regulation performance at DC is the gain of the error amplifier.

In an ideal operational amplifier, the DC gain would be infinite, but in practice it never is. This means that in order for the amplifier to produce a voltage at its output, the voltage difference between the inverting input (ii) and non-inverting input (nii) will not be zero but some fraction of the voltage at the output. For example, if the output was to be 10 volts and the gain of the amplifier was 100, there would be a voltage difference of 0.1 V between the inputs. If the nii were connected to a voltage reference of 1 V and feedback from the output of the overall circuit was taken directly from the final output via a voltage divider, the ii would be maintained at 0.9 V. This would mean that in equilibrium, the output of the overall circuit, in our case a power supply, would be 9/10 of what it would be if the error amplifier's gain was infinite. We can improve our performance by using an error amplifier with higher gain at DC. There isn't a specification for the error amplifier "open loop gain" as such in the 723 datasheet, but it is probably a few thousand.

Another limit on regulation, both accuracy and stability, is the voltage reference. Good voltage references are quite complex circuits and can be expensive - as much as hundreds of dollars. The reference in the 723 has quite poor initial accuracy: 6.8 V minimum, 7.15 V typically and 7.5 V maximum at 25 °C. For many purposes, this would be regarded as extremely bad and unusable. For a power supply made with fixed resistors it would just be bad. For a power supply that is adjustable and set using an accurate meter, it is tolerable. The other factor to consider is the stability of the reference with time and temperature. All references will change voltage over time. Generally the change is unpredictable and can be moderately large in the first few tens to hundreds of hours of operation, then they tend to become more stable. Variation with temperature is usually more predictable. There is no separate specification for the reference stability in the 723 - it is just included in the overall stability values in the datasheet.

There are other factors in power supply performance. One that is very important but often overlooked by inexperienced designers is dynamic performance - how well the supply maintains accurate voltage when the load current is changed quickly. It is quite possible to design a power supply that will regulate at (say) 10.000 volts with a steady load, but overshoot to 11 volts if the load current is suddenly reduced or drop temporarily to 9 V if the load current is suddenly increased.

 

In my experience of designing circuits fro different applications I quickly learned that it makes a lot of sense to use parts available from a number of manufacturers.. So while the designs never used "bleeding edge" brand new components, all of the equipment can still be repaired 20 years later, which is what my industrial customers wanted. Parts not being available can certainly stop production and cause a lot of problems. Multiple sources is the simple solution.

 

Hello, I have the attached schematic. I tested the schematic and it seems to work.
I want to add a red led to light up when a short circuit occurs.

Explanation of the schematic is:
1. I use an 2x15V/150VA toroidal transformer connected at X1-1 and X1-2.
2. One 15V secondary is connected at X2-2 and X2-1.
3. The Darlington power transistor is connected at X3 (X3-1 = E, X3-2 = B, X3-3 = C).
4. The shunt resistor is connected at X4-1 and X4-2. The value of this resistor will be calculated when I will decide the maximum output current.
5. The X5-1 and X5-2 is the output of the power supply.
6. T1 is BD139.
7. C2 and C3 are 4700uF/63V.
8. C5, C7 and C12 are 100nF/100V.
9. C6 and C13 are 220uF/63V.

I added the LED1, R15, T3 and R14. R15 will limit the current through the LED1, but I am not sure about the value for R14. T2 and T3 will be the same part number, BC547.

Please have a look at the schematic and let me know if the overload led is correctly connected into the original schematic ?

 

You can use a regular transistor symbol for the darlington transistor and it will make the circuit easier to follow. And a standard convention is to use the letter "Q" to describe transistor devices, reserving the letter "T"T for Transformers. And showing the shunt resistor rather than a couple of points will also make the circuit easier to understand.
The transformer connections, as stated and drawn, do not seem to make sense . The winding across X1-1 and X1-2 puts an AC voltage across the C2, C3, and C5 capacitors, which makes no sense. The winding connected across X2-1 and X2-2 will charge capacitor C1, but there is no connection from the negative side of C1 for any current to flow anywhere. So I do not see that making any sense either. So as printed and described I do not see that the supply will work at all.

 

Mouser has tens of thousands of them in stock.

 

Ok, then I will build the attached schematic.
Could you please recommend some equivalents for TR2 and TR3-TR6 ?
They do not need to be pin to pin compatible.
I searched on the internet and I did not found the datasheet for TR3-TR6. I think that their part number is 2SA823.
For TR2 I think it can be replaced by BD140.