LM723 based PSU with min voltage of 0,00V

earckens said:

spec, I

spec I am puzzled by your comment: 1N4148 Irev=50uA, 1N60 Irev=0,5uA. 1N4148 is worse, or not?? https://www.vishay.com/docs/81857/1n4148.pdf

Click to expand...

Hi earckens,

Yes, the leakage current data on the 1N4148 data sheet it is confusing.

One thing to know is that the reverse leakage of a diode is proportional to the reverse voltage.

Also, when you power supply is operating in the normal voltage stabilization mode (not in the current limit mode) the voltage at the LM723 pin 4 is constant at 6V nominal. That means that the reverse voltage across the diode is always 6V, so you need to look at the leakage current for 10V reverse voltage on the data sheet, rather than the reverse leakage current at the maximum safe reverse voltage of 75V for the 1N4148.

Another point is that a particular version of an 1N4148 is liable to have a reverse leakage current close to average rather than maximum, while a Schottky diode, being leaky by nature, is more liable to have a leakage current higher than average (I know that mathematically that is garbage, but that is what I have found and I have messed a lot with both types of diode).

spec

PS: I could not find any data on the 1N60; do you have a link to the 1N60 datasheet?

spec said:

Hi earckens,

Yes, the leakage current data on the 1N4148 data sheet it is confusing.

One thing to know is that the reverse leakage of a diode is proportional to the reverse voltage.

Also, when you power supply is operating in the normal voltage stabilization mode (not in the current limit mode) the voltage at the LM723 pin 4 is constant at 6V nominal. That means that the reverse voltage across the diode is always 6V, so you need to look at the leakage current for 10V reverse voltage on the data sheet, rather than the reverse leakage current at the maximum safe reverse voltage of 75V for the 1N4148.

Another point is that a particular version of an 1N4148 is liable to have a reverse leakage current close to average rather than maximum, while a Schottky diode, being leaky by nature, is more liable to have a leakage current higher than average (I know that mathematically that is garbage, but that is what I have found and I have messed a lot with both types of diode).

spec

PS: I could not find any data on the 1N60; do you have a link to the 1N60 datasheet?

Click to expand...

Hi spec, very interesting; I did not know this. Here is a link to a 1N60 pdf: **broken link removed**
Indeed, at 20V Irev for the 1N4148 is 0,025uA, for the 1N60 at 15V the Irev is 0,1uA.

earckens said:

Hi spec, very interesting; I did not know this. Here is a link to a 1N60 pdf: **broken link removed**
Indeed, at 20V Irev for the 1N4148 is 0,025uA, for the 1N60 at 15V the Irev is 0,1uA.

Click to expand...

Hi earckens,

Thanks for the data sheet link- not heard of the 1N60 before.

Just a small point- the worst case leakage current for the 1N60 is actually 0.5uA at 15V reverse.

By the way, the BAT41 is my go-to small signal Schottky diode for low leakage: 0.1uA worst case at 50V reverse.

Another snipit of information about silicon semiconductor leakage current is that it increases exponentially with temperature.

There is a very approximate rule of thumb that says for every 10 Deg C rise in junction temperature the reverse leakage current will double.

spec

Ordinary small silicon junction diodes with low reverse leakage are:
BAS21, BAV20, BAV21

And in the super group of low leakage diodes, the Fairchild FJH100 is king at 10pA reverse leakage.

Diodes are lovely components- so well behaved and forgiving. There is more to them than you might at first think too.

spec

earckens said:

View attachment 104556

Click to expand...

Hi earckens,

Pretty awesome what you did with the Lab PSU LM723....I was wondering if you would OK with sending me the eagle sch. file...I would like to upgrade my lm723 psu with current limit indication and would like to create a PCB from your file...

Thanks and great work

Roger

Yes of course I can send you the Eagle files, just pm me your email.
I plan on adding the controller section too in this file but haven't done this yet. The controller measures voltage, current and 2 critical temperatures (the Darlington 18V regulator and the heatsink temperatures).
These get all displayed, as well as the necessary alarms get blinked and sounded when needed. I use a Arduino Pro Mini for that. Pretty big program because I use statistical outlier exclusions and some other programming tricks.
If you want the source file, feel free to ask if you want it, however the diagram I am due to draft is necessary too.

And thanks for the compliments but they should be directed to everyone here in this thread and in the other related thread! Without all the fantastic help I received on this forum I would not have been able to accomplish this.

Rgds,
Erik

Hi Erik,

Here is my email, email address removed by moderator Thanks for your reply and congrates for this work and for what is coming...I’m not yet very knowledgeable with Arduino, but if I get into it, I may try the same and let you know...

I have seen that you are from Belgium... I’m from the Montréal area (Québec, Canada) Was wondering if you spoke French? If so don’t hesitate to write in French!!!

Thanks a lot!

Regards,

Roger

Email addresses in public posts are an invitation to spammers.

Roger Albert said:

Hi Erik,

Here is my email, (xxxx) Thanks for your reply and congrates for this work and for what is coming...I’m not yet very knowledgeable with Arduino, but if I get into it, I may try the same and let you know...

I have seen that you are from Belgium... I’m from the Montréal area (Québec, Canada) Was wondering if you spoke French? If so don’t hesitate to write in French!!!

Thanks a lot!

Regards,

Roger

Click to expand...

Hi Roger, yes I speak french altough my native language is flemish. We (spouse and me) have a holiday home in the Provence (southern France), so that helps.
Arduino: not that hard if you have some knowledge of programming (my experience dates from the early 80's and then with Fortran and Pascal), and loads of internet info and helpforums, so you should be able to get going.
I will pm you,
Erik

Hi everyone, finally I did get all paperwork updated, and can present you with a full set of drawings ready for use including the controller program code (hex code for Atmega 168 or 328) and the digital section of the psu.

Little recap: it all started when in October last year I blew up the transformer of a 30 year old lab psu I had made in my student years and I wanted to rebuild the thing. Because it had the nice feature to supply 0,000V lowest output voltage up to 30V at 3A. But this time I wanted it do go from 0,000V to about 46Vdc, still at 3A continuous rating.
I also wanted to include a controller to display on a 16x2 LCD the output voltage, output current, and the temperatures of the darlington (Q6) inside (which goes at a steady 60 to 80°C), together with appropriate audible warnings as well as display warning messages, as well as a LED to indicate the current limitation is active, as well as another LED to show extra warning signals for overtemperatures.

Again a big thank you to all who did help me with this project since October last year.

I will of course gladly help anyone wanting to fabricate this unit.

Grts,
Erik

Edit 16/06: changed Q7 fan driver from npn to n-channel MOSFET 2N7000
**broken link removed** **broken link removed**

And here is the brd file.

Edit: and a png picture

Update: C5 to be changed from 470pF to 100pF.

Just a few days ago I received a new oscilloscope and I checked the output of the PSU. The output vo ltage is fed to a microcontroller for display, to 10mV resolution. Until now this display jumped all over the place within 800mV range; I had been looking into software issues, or ADC problems but with the scope I established in effect up to 400mV 5kHz to 8kHz sawtooth ripple. A number of hours were spent trying to solve it, and finally the culprit was found: C5.

The .brd file needs no update, just the schematic: