Can someone please check the pcb i made for my bench supply, my hava fake chip

[multime]

I have made and populated a pcb for my new bench power supply but its not working as expected it is designed around an obsolete chip MC1466l/1566L. just wodering can some check the pcb against the circuit schematic, its not a complicated circuit. there were some errors which i have corrected but any errors you spot let me know. Thic pcb is not done with any circuit design software but hand drawn.

The three images show the pcb design and a copy of the acetate images back and front.

I am hoping some one can find errors i have not spotted otherwise maybe these chips are fakes.

 

[indeterminate]

Try drawing a circuit diagram based on your pcb art work
it should be fairly easy to spot the problem then

and measure sum voltages at various points and enplane what the problem is

[SteveThackery]

You will need to post the schematic as well.

[multime]

Sorry i thought i had uploaded the schematic, I've attached it now.

The problem i am having is the output voltage will not rise above 7.25 v when turning p1 the voltage output control pot as soon as 7.25v turning the pot any further has no effect. Adjusting rv1 has very limited effect it only increases output voltage by 0.10 - 0.15v. 7.25v also happens to be the  internal voltage regulator value.

When looking at the schematic the LM393 (now changed to LM339) has now been isolated to aid in fault finding so no longer has connections to pins  8 and 9 of MC1466L.

Transistor Q1 2N5401 is now installed correctly i originally had it inserted with collector and emitter reversed.

I have checked the pcb, and component values and as said previously have corrected any faults found (wish i had made a note of them) but i just cannot get this chip to go above 7.25v. So just hopping a fresh pair of eyes may spot something i have missed.

 

[multime]

I have now after checking the pcb again today for about the tenth time and finding no faults  have resigned myself to the fact the chips i have are fakes. I will just put the board to one side in the hope i may find some genuine MC1466L / MC1566L at some point and move on to another design.

[SteveThackery]

Can you confirm that the circuit as drawn is definitely good? Is it a known design that definitely works?

[multime]

The circuit was found on here and the poster had help from people on here with aspects of the design, protection diodes etc. The circuit minus those and the op-amp mode detection is similar to what Motorola
depicts in its data sheet (applications) Datasheet found here

https://www.alldatasheet.com/datasheet-pdf/pdf/87107/MOTOROLA/MC1466L.html

[Doctorandus_P]

I do not understand why you want to build a power supply around an MC1466. It's been obsolete for maybe 30 years or so. There are a gazillion power supply schematics in all kinds of configurations and power classes. My own preference is something build around generic opamps. They never go obsolete, can always be serviced or replaced and they are much easier to design around then discrete transistors (although that can be a fun exercise too, so as always, it depends on your goals).

Also, have you ever thought about some EDA? I've been using KiCad for quite some years now, and even before I was using KiCad, I put about a week of my life into finding a better EDA suite as I had previously used. Doing schematic and PCB design on a PC is such a huge advantage that I would never could go back to the old flattened dead tree carcasses.

Even old guys can learn new tricks. Some claim it can delay dementia / alzheimer (Not altimeter you stupid *&^%$#@! spelling thing) for a few years, and a few more good years is worth fighting for.

[indeterminate]

 similar to what Motorola depicts 
That is guaranteed to get you in trouble , have you taken any voltage measurements from the pins of the ic.
if you purchased the MC1466 on ali its 100% face
there are sum NOS traders listing them for ~$10 each

and +1 on using KiCad 

[xavier60]

The IC isn't simple at all. Pin 7 should always be about -7V WRT the PSU's + output. Pin 7 should never make contact with GND. This will likely cause damage.

[coromonadalix]

does your schematics come from this site ?   is it a rip off out of it ?

http://chemelec.com/Projects/MC1466/MC1466.htm


and another project on this one

https://www.eeweb.com/mc1466l-driven-power-supply-with-30-volts-at-2-amperes/


i have 4 of supposed genuine mc1466  that i never used   loll

mc1466  use a floating supply topology ... 

[xavier60]

Yes, it works in the same way as regulators that need one voltage setting resistor like the LT3081. The Reference pin outputs a small precise current.

[multime]

Doctorandus_P

This power supply started life about 3 years ago ( long time i know but when your the head cook and bottle washer there isnt much free time). and was going to be based around 2 Lm723,s which was an old Velleman design sold by maplin. I had built one before and it was a very good supply apart from it could go down to 0V despite being sold as 0 to 30 volts. This really irritated me being a bit of a purist and i eventually cured that but i still cant recall how i done it, so while had had amassed all the parts to build it that stuck in my. Searching around i came across MC1466L and liked the tight regulation of voltage and current plus it went to 0v, so even knowing it was made obsolete many years ago i thought i may be able to find original chips somewhere but as yo can now see that didnt work out to well.
 Yes i have considered  cad design and used it still having Eagle and Kicad on my pc but i find them cumbersum to use and to much to learn now. Its just easier to pick up pencil and paper and i can assure you designing and drawing circuits this way certainly keeps your mind active. Not saying i wont give Eagle and Kicad another go but at the moment i just want to get this psu over the line.

indeterminate

Yes i have taken voltage readings but when you have 4 chips and each one gives very different readings you know the chips suspect to say the least. two were from ebay two from Ali. craftily worded on ebay as untested ( and thats because there,s no point testing this carp).

xavier60

Who mentioned pin 7. I know every circuit i have seen on the net including Moterolas shows that pin as connected to 0v.

coromonadalix

If the chips you have are genuine, contact me. The circuit i seen originally was on this forum. But for my next build i am going to try the discrete version of the MC1466L as shown on chemelec website.

[coppercone2]

I strongly suggest making a traditional circuit diagram (boring kind). The one you have is a useful layer of abstraction but its not a design document.

That one is good if you are very familiar with the circuit and you want to do mods. But its not a design trouble shooting document

[xavier60]

@multime. I mentioned the possibility of pin 7 being mistakenly connected to GND. Where did the MC1466L come from? I'm willing to test on a breadboard if not too expensive.

[multime]

But thats were Motorola tells you to connect pin 7 ?.  Tell me more about I'm willing to test on a breadboard if not too expensive.  I am in the uk

[multime]

coppercone2

If i get the time i will do that

[xavier60]

In the schematic in post #3, Gnd is the PSU's negative output terminal.  The reason for my comments is that the mistake connecting pin 7 to this Gnd could be easily made. I have not looked at your layout.
I aslo mentioned that pin 7 should always measure about -7V with respect to the PSU's positive output terminal.

[Doctorandus_P]

I have now after checking the pcb again today for about the tenth time and ...

I thought for about 10s to put your drawings into KiCad and both put symbols over the schematic, and footprints over the PCB and turn the whole thing into a KiCad project. I thought about this again (and a bit longer) when you responded you do have / use KiCad.  But an obsolete schematic as this is just not worth it.

Once you are more familiar with KiCad, doing something like this does not take very much time, You get ERC and DRC, much magnified and clearer pictures on your monitor, Footprints that fit (for home etching). Even for a one-off build on matrix board, you can plan the footprint placement, and shift things around in KiCad until everything fits. The advantages of modern EDA tools are just to big to ignore or dismiss.

[multime]

xavier60
Ok i understand now

 Doctorandus_P

Just to put the record straight, yes i have kicad and eagle on my pc and my use of these has consisted of me opening both of them just to see if i could get the feel for them. I didnt, and have never designed any circuits or pcbs, with them because as i have said i found them cumbersome so my adventure with them ended there, yes i may take another look at them at some point.

 

[Doctorandus_P]

I understand your trouble. A PCB design program is a quite complicated piece of software and it takes a considerable amount of time and effort before you can work with it efficiently. Once you are past the leaning curve you would not want to go back. It is your choice whether you want to invest that amount of time & effort.

There is also no need to learn two different PCB design programs. Development of eagle was (as far as I know) on a quite low level for 10 or so years before autodesk bought it, and after that support has become more expensive and difficult to obtain. Autodesk has declared it will stop all support for eagle within a few years. As a result, eagle is mostly a dead end.

KiCad is still growing, both in capabilities of the software itself, the amount of people using it, and support from 3rd parties such as libraries and PCB manufacturers. A lot of users of the "free" version of eagle switched to KiCad after autodesk bought eagle. I am quite biased myself, I like KiCad and never liked eagle and barely know it, but I am still quite sure that KiCad is a much better choice then eagle if you want to go into PCB design.