uA723 voltage regulator viable in 21st century?

[Richard Crowley]

I can see why LM555 continues to be used, but why would antiques like uA723 and uA709 continue to be used in new designs?
Over in the GeekSlutz forum I was helping a guy from Sweden fix his audio mixer apparently designed in the mid 2000s.
The power supply is dead because the custom transformer (made by respected audio transformer maker Lundhal) has an open primary winding.

But the regulator design just seems completely antique.  It was obsolete when the thing was designed.  If you need + and - 15V @ 200mA, then why wouldn't you just use a positive and negative 3-terminal regulator (LM7815 and LM7915 or a modern descendent).



Service Manual PDF: https://data2.manualslib.com/pdf5/101/10098/1009764-satt_electronics/sam_82.pdf?373d3cacdef1450e71f59cd434f92921

[Conrad Hoffman]

If you need it, you can customize fold-back limiting with the 723. It's still a good part!

[pelule]

But the regulator design just seems completely antique.  It was obsolete when the thing was designed.

Obsolete?
At least at TI the uA/LM723 is still active:
http://www.ti.com/sitesearch/docs/universalsearch.tsp?searchTerm=LM723#linkId=1&src=top
http://www.ti.com/sitesearch/docs/universalsearch.tsp?searchTerm=uA723#linkId=1&src=top
Call me old fashioned, but I am still use the uA/LM723 in my designs, if I need a bit more sophisticates linear regulation (i.e current protection or precisely adjustable voltage. I also still use the 723 as a very cheap temp-stabelized reference voltage (an old Elektor circuit to temp-regulate the on-chip reference using the on-chip pwr transitor as heater and the on-chip current sense transistor as temp-sensore).
/PeLuLe

[Bassman59]

I can see why LM555 continues to be used, but why would antiques like uA723 and uA709 continue to be used in new designs?
Over in the GeekSlutz forum I was helping a guy from Sweden fix his audio mixer apparently designed in the mid 2000s.
The power supply is dead because the custom transformer (made by respected audio transformer maker Lundhal) has an open primary winding.

But the regulator design just seems completely antique.  It was obsolete when the thing was designed.  If you need + and - 15V @ 200mA, then why wouldn't you just use a positive and negative 3-terminal regulator (LM7815 and LM7915 or a modern descendent).

Read more of the service manual! It uses a 709 in the input stage and the quad 4741 elsewhere. An RC4558 is used in the oscillator.

And zoom in on the schematic. It's dated 1976. So, no it's not modern, and it certainly wasn't designed in the 2000s.

(This kinda silliness is why I don't bother with GearSlutz ... it's all idiots talking in a circle jerk about audio gear they don't understand and have never used.)

[schmitt trigger]

My very first project employing an IC that actually worked, was a variable regulated power supply using a uA723. Which I used for my personal experiments for about 10 years.
So, I feel a lot of nostalgia for the uA723, and know that it is a very flexible building block which allows certain configurations like foldback current limiting.

But I would not use it today, for a couple of reasons:

The series pass transistor is not thermally protected.
I prefer spending my limited time, board real state and components in something more meaningful than a power supply.

[Richard Crowley]

Why do you say 2000s?  It looks like the schematics are dated 1976. 

Doh! You are right.   

From that perspective, it seems more reasonable to see uA723.
I don't know where I got the notion that it was from the mid-2000s

[chris_leyson]

As far as I know they're still available. Built a super low noise low power regulator a long long time ago, it performed much better that the low noise linear regs from TI, National etc. Can't remember the circuit but I think there was a big tantalum capacitor on the reference pin. They're very versatile regulator chips and like Conrad mentioned they're great for foldback current limiters. Even seen them used as switching regulators in the HP9825 power supply.

[bd139]

Indeed. It’s a power supply construction kit not a voltage regulator. And each chunk of it has pretty good performance.

[rdl]

... it’s an incredibly useful part because it does not present you with a regulator as such, instead it’s a kit of all the parts required to make a regulator of almost any description.

https://hackaday.com/2016/12/02/get-to-know-voltage-regulators-with-a-723/


I bought a handful of them years ago, and one of these days I'm actually going to use one for something.

[DimitriP]

If you need + and - 15V @ 200mA, then why wouldn't you just use a positive and negative 3-terminal regulator

Stellar noise performance (with very few- maybe 6 - expections) and you don't need  magnifying glasses to see it.

[schmitt trigger]

A SMPS using a 723!
Now I remember a fascinating Unitrode app note describing a hysteretic regulator utilizing a 723.

[chris_leyson]

Still available in TO100-10 package but £8 or £9 each !

[richard.cs]

I used one at work a few months ago. I needed low noise and there are very few options for that in the voltage range I required. It's versatile enough to have a lot of uses, if nothing else keep a few in the junk box for when you need to build a slightly unusual regulator.

[pelule]

Still available in TO100-10 package but £8 or £9 each !

Conrad, a German (not the cheapest) electronic seller offers:
Texas Instruments UA723CDR, SOIC-14 for 1,12 EUR/pcs
Texas Instruments UA723CN, PDIP-14  for 0,89 EUR/pcs
Texas Instruments LM723H/NOPB, TO-100-10 for 11,70 EUR/pcs
So plenty options still active & available
/PeLuLe

[glarsson]

I don't know where I got the notion that it was from the mid-2000s

The enclosure that other manufacturers might have designed in 2006?

[rsjsouza]

I too agree the uA/LM723 is still a useful part.

Funny how TI's product page for the LM723 shows an unwarranted comparison with the LM317L-N in the front page. (perhaps trying to steer people towards the other part).

http://www.ti.com/product/LM723

[Synthtech]

I run a vintage synthesiser repair workshop, the uA723 is used in a lot of instruments and it is a part that rarely if ever fails. They are a workhorse that runs reliably for decades and it does it’s job very well. Three terminal regulators are also incredibly reliable of course but the 723 has advantages over them.

[glarsson]

A SMPS using a 723!

In the early eighties I used a 723 as a heater element and heater controller for a crystal oven. The output driver transistor was the heater, a PN junction of the "free" transistor (normally used for current limiting) was the temperature sensor. Glue the crystal to the top of the DIP and pack with some insulating material.

[KE5FX]

Only very recently have linear regulators been introduced that rival the ua723's noise performance.  The ability to bypass the reference is a big deal in some applications, and for a long time it was either use a 723 or roll your own.

[dardosordi]

Just recently stumbled upon this guy, he likes the lm723 so much that it had made me appreciate it a lot: https://electronicprojectsforfun.wordpress.com/power-supplies/a-collection-of-proper-design-practices-using-the-lm723-ic-regulator/

Also check out lm723 self heated reference: https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/

[edpalmer42]

FYI, the elektor version of the self-heated reference was in the 1978 July/Aug. issue on page 7-89 which is page 94 of this pdf.

https://archive.org/download/ElektorMagazine/Elektor%5Bnonlinear.ir%5D%201978-07_08_text.pdf

[texaspyro]

Only very recently have linear regulators been introduced that rival the ua723's noise performance.

Yep, and that new fangled stuff  can cost 10x-100x what you can get 723's for...  I recently bought 100 of them for $5.00.

[BravoV]

FYI, the elektor version of the self-heated reference was in the 1978 July/Aug. issue on page 7-89 which is page 94 of this pdf.

https://archive.org/download/ElektorMagazine/Elektor%5Bnonlinear.ir%5D%201978-07_08_text.pdf

Thanks for the link, also extracted that page and attached below.

About the uA723 used as TC voltage reference, any idea what is the realistic performance ?

[Wolfgang]

Hi,

I think they chose this old chip because it has much less noise than a LM317, LM78XX and many others, even much more modern chips.
For low level audio, measurement and other sensitive app like low phase noise oscillators this can be important.

I learned a lot from a german paper about regulator noise (just looking at the graphs tells you everything you need):

https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern

Another aspect is a datasheet value for long time drift and tempco. The cheap 3-legged regulators can be off by 5%, have up to 10mV of noise, a bad PSRR at higher frequencies, and no long term drift specs, plus a bad tempco in comparison. I agree that a 723 is more effort, but it can make sense.

The question for me is not if a chip is old or new, but if it does the job I need. Electronic design should not be based on what is fashionable, but on specs. 
There must be a reason why TI, ST and others still sell this methusalem ...

For prototyping, the availability of a DIP or TO-99 is also a plus.

Regards
   Wolfgang

PS: If you want to see some 732 designs for all kings of projects:

https://electronicprojectsforfun.wordpress.com/power-supplies/

and something a bit off-center:

https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/

[blackdog]

Hi,

And do not forget my uA723 Oven designs 

https://www.eevblog.com/forum/metrology/small-oven-controler-for-voltage-reference/

Kind regards,
Bram