Go Down

Topic: Daisy chained shift register problem :( (Read 2657 times) previous topic - next topic

CrossRoads

Perhaps add a termination resistor to the end of the clock and latch lines. 500 ohm, 270 ohm, something in that range, to Gnd.


Designing & building electrical circuits for over 25 years.  Screw Shield for Mega/Due/Uno,  Bobuino with ATMega1284P, & other '328P & '1284P creations & offerings at  my website.

TomGeorge

Hi,
Post#31 and we still have no schematic or PCB pattern.
If we had them when asked, the cap to the signal line would have been picked up long ago. :o


Please  some diagrams....


Thanks.. Tom.. :)
Everything runs on smoke, let the smoke out, it stops running....

nurimo

Hi,
Post#31 and we still have no schematic or PCB pattern.
If we had them when asked, the cap to the signal line would have been picked up long ago. :o


Please  some diagrams....


Thanks.. Tom.. :)
As you know, it happened just after pcb fry incident. I just assumed it is connected to that. But it seems the problem is some noize in latch pin. Some suggestions were made on forum but not every one of them has good explaination. My problem originates from instant change of power supply load. Since i open 32 motors per power supply with different order in 10 ms interval. But here is what i cannot understand, cant i just have a big cap on vcc of the 74hc595 so that big power changes do not affect the ic ? I already have 0.1 uf

nurimo

Hi, problem was my power supplies. I have changed them. Problem is mostly gone. I havent set all 4 power supply to same voltage by their pot but they are parallel connectable mean well brand 12 V 29 A power supplys. Do I need to set each of their voltages to the same or approximate values ? Can this be the cause of a little parasitic behavior "not setting them all to the same voltace "

TomGeorge

Hi, problem was my power supplies. I have changed them. Problem is mostly gone. I havent set all 4 power supply to same voltage by their pot but they are parallel connectable mean well brand 12 V 29 A power supplys. Do I need to set each of their voltages to the same or approximate values ? Can this be the cause of a little parasitic behavior "not setting them all to the same voltace "
Hi, what do you mean  "parallel connectable" ?
Please at least a power supply connection diagram!!!  :o :o


PLEASE.....Tom.. :)
Everything runs on smoke, let the smoke out, it stops running....

nurimo

#35
Mar 15, 2018, 09:43 pm Last Edit: Mar 15, 2018, 09:50 pm by nurimo
Hi, what do you mean  "parallel connectable" ?
Please at least a power supply connection diagram!!!  :o :o


PLEASE.....Tom.. :)
I can connect all V+ outputs of power supplies together and do the same for V-. I can do better, brand is meanwell and model is LRS-350-12KO

Check out the link below.
https://www.power-supplies-australia.com.au/blog/connecting-mean-well-power-supplies-parallel

TomGeorge

#36
Mar 15, 2018, 10:36 pm Last Edit: Mar 15, 2018, 10:41 pm by TomGeorge
I can connect all V+ outputs of power supplies together and do the same for V-. I can do better, brand is meanwell and model is LRS-350-12KO

Check out the link below.
https://www.power-supplies-australia.com.au/blog/connecting-mean-well-power-supplies-parallel
You told us this;

Quote
Every 5 of them connected to a 40 A 12 V switching power supply

Your powersupply is not included in the list of current sharing SWMP that the link supplies!!!!!!

You do not connect all the V+ together, the gnds YES, but NOT the V+.

You connect the V+ to groups of your boards.
As in;

Quote
Every 5 of them connected to a 40 A 12 V switching power supply

The sharing diagram there will not work unless you have them ALL exactly the same voltage, not possible.

As they do not monitor each others current, they cannot evenly share current.

The noise could quite easily be the supplies interacting with each other.

Please at least a power supply connection diagram!!!  :o :o

PLEASE.....Tom.. :)
PS. Are aware that if you supplied the schematics we requested 7 DAYS and over 30 posts ago, your problem could have been solved.
Everything runs on smoke, let the smoke out, it stops running....

nurimo

#37
Mar 15, 2018, 10:40 pm Last Edit: Mar 15, 2018, 10:48 pm by nurimo
I want to share the connection diagram but its too complicated to understand from pictures and I didnt use wires to connect the components, I used NETs to do it. So I will send pics of pcb instead

nurimo

You told us this;
Your powersupply is not included in the list of current sharing SWMP that the link supplies!!!!!!

You do not connect all the V+ together, the gnds YES, but NOT the V+.

You connect the V+ to groups of your boards.
As in;

The sharing diagram there will not work unless you have them ALL exactly the same voltage, not possible.

As they do not monitor each others current, they cannot evenly share current.

The noise could quite easily be the supplies interacting with each other.

Please at least a power supply connection diagram!!!  :o :o

PLEASE.....Tom.. :)
PS. Are aware that if you supplied the schematics we requested 7 DAYS and over 30 posts ago, your problem could have been solved.




As you can see 7805 regulator has 2 capacitors 0.1 and 33 uf.  The 0.1 cap at the top is not connected anymore. It was connected to Latch line. 12V supply connected to 1000 uf and 100 uf. But I added another one 100 nf. Also changed the smd resistors from 220 to 470 since I need only 2-3 ma per output. But the noise still there not much but there.

As I said in previous post, powersupplies are connected parallel right now. They have a name for this feature " current sharing." mean well powersupplies have this feature. So I connected V+ terminals together. Still every 5 board connected to 1 power supply but since they all connected together pcb boards connected together all ofcourse. But before you say anything, I tried to do it by seperating the power supplies and supply the 5 boards each. It didnt work, a little noise is still there, this is why I used current sharing feature.

I know they should be at exactly same voltage level or the higher one will be overloade and noise will still be there. I just followed the manufacturers suggestion on current sharing.

TomGeorge

Hi,
What don't you understand with this statement?

Quote
Your powersupply is not included in the list of current sharing SWMP that the link supplies!!!!!!
Don't just copy and paste!!

Please read..

Also, why don't you have a fuse on each of your PCBs?

Tom... :)
PS Please at least a power supply connection diagram!!!  :o :o 
An image of your PCB pattern and a schematic will help!!!!!!!

Everything runs on smoke, let the smoke out, it stops running....

nurimo

Hi,
What don't you understand with this statement?

Don't just copy and paste!!

Please read..

Also, why don't you have a fuse on each of your PCBs?

Tom... :)
PS Please at least a power supply connection diagram!!!  :o :o
An image of your PCB pattern and a schematic will help!!!!!!!


I dont have it on this version I added after I burn them :S Fuse free versions of this board with CD4094 works for 2 years now. And technician of the product showed me from catalog that it has the feature but if you say so, I wont argue your experience and remove the connections. So what do you suggest ? I dont want to use capacitors on signal lines but it seems to get better when I connect 1 uf cap on Latch pin at the end of the signal line. It gets better but not completely gone.

lastchancename

#41
Mar 15, 2018, 11:53 pm Last Edit: Mar 15, 2018, 11:54 pm by lastchancename
Quote
Problem is mostly gone.
So you're worse off now than when you started!!

A persistent fault is far preferable to an intermittent issue.
But what would we know?, your project is too complicated for us to follow the schematics.

(( I wasn't gong to follow, but I saw the thread is still alive!
Tom, you're a hero. ))
Experienced responders have a nose for laziness, (they were beginners once)... expecting the poster to contribute to the learning experience.

nurimo

So you're worse off now than when you started!!

A persistent fault is far preferable to an intermittent issue.
But what would we know?, your project is too complicated for us to follow the schematics.

(( I wasn't gong to follow, but I saw the thread is still alive!
Tom, you're a hero. ))
Tom suggested the shift register a while ago we have a connection :) If I had that experience, I would help others too. Untill then, absorb what you can mode on. And Im redicioulisly stuck with no ideas. I practically bought very good and noise free power supply it got better but not completely gone.

nurimo

So you're worse off now than when you started!!

A persistent fault is far preferable to an intermittent issue.
But what would we know?, your project is too complicated for us to follow the schematics.

(( I wasn't gong to follow, but I saw the thread is still alive!
Tom, you're a hero. ))
by the way, its not complicated in schematics, its complicated in pcb board view. and eagle schematic side is a mess. because I didnt use wire connections, I used nets and there is no wire connection that can be seen. This was what I meant when I said complicated. On the contruary its too simple actually. shift register opens and closes optocoupler and optocoupler activates power mosfet thats all.

TomGeorge

#44
Mar 16, 2018, 01:46 am Last Edit: Mar 16, 2018, 01:48 am by TomGeorge
I want to share the connection diagram but its too complicated to understand from pictures and I didnt use wires to connect the components, I used NETs to do it. So I will send pics of pcb instead
  • Do you or do you not have a schematic of the PCB?
  • Do you or do you not have a pattern of the PCB?
  • Have you divided your power supplies up?



If you designed and had this PCB manufactured you should have BOTH or the above.


Tom.. :)
Everything runs on smoke, let the smoke out, it stops running....

Go Up