Dual CMOS based timer problems...

[Kain]

I have built a timer circuit based on CMOS 4538 which consists of 2 monovibrators. I am getting some sort of problems, such as bouncing and wrong timing. Do you see anything wrong with the schematic that i posted because I don't see it. Maybe i overlooked something... There is no noise in the device it is used for. If I test it out of the circuit it works fine, but once I connect it in and it freaks out. The other thing I noticed is that if I use 2 chips on separate boards for the purpose it stops doing it. I am lost...

View attachment 37495

 

[Anotherforummember]

Difficult to see how the circuit is functioning.

 

[ante1]

I agree, and the power supply part sure needs some explanation?  ???

 

[Kain]

Ok now. I do not know if you saw the 680nF capacitor which you should note is connected before the bridge rectifier. As hopefully we all know capacitors have reactive resistance to AC, which means that we limit the current to way lower value, and this is why the zener is still not blown up  ;D. This is very basic AC mains power supply which is used when you need only small ammount of power, as in this case since we power a CMOS and basically 2 LEDs that are limitet to 5mA each if I remember correctly. I will hook up 1 and 15 to ground just for the heck of it. The manufactor says that those are already internaly connected to ground and I actually measured them - they are indeed. The AC1 and AC2 notations in the schematic are only for the CAD to know where the other parts of the schematic connect; this is why you see them on more than one place. The AC mains is plugged in pins 1 and 2 of the connector.

 

[Kain]

I tried to connect 1 and 15 to ground and as expected there was no change. It acts in the exactly same way. Any other ideas? I hope the PSU part pf the schematic is clear already 

 

[Staigen1]

Hi there Kain

This will surely not work, take a close look at the PSU:s rectifier! The output to the Zenerdiode is taken from the input to the rectifier bridge and the input to the bridge is connected to its output. Also, small, lowpower PSU:s for things like this usally not have rectifier bridge. As you see, the catodes of Q1 and Q2 is related to AC1, then i belive the "ground" of D2 and IC1 also should be connected to AC1! If you then connect the output of the bridge to VCC and move the connection of the 680 nF condenser to both inputs of the rectifier i belive that it works. Usally you also have a little resistor in series with the 680 nF condenser, in the range 100 to 300 ohms, to reduce inrush current, usally 1 to 2 wattage. Hope this help.

//Staigen

 

[ante1]

Hi Kain,

I don’t recommend this kind of psu solution, it’s not entirely safe. If you get (AC1) to live side the circuit ground will be live minus the diode drop of one diode in the bridge which is not healthy! 

 

[Alun]

I agree Ante, this kind of power supply is fine for small devices with no external connections so in this case it's a very big safety hazzard.

 

[Kain]

Ok now. Do not pay this much attention to the PSU part. I know it's not correctly drawn, but I know how to build it correctly and it does it's job well -  I am getting stable 13V out of it. As about safety I know that it is not safe to have mains all over, but you didn't know that this device is in fact enclosed in a box, not to mention that practically in this box there is as much unsafe AC as you want ;D. I am not the one who chosed this type of PSU anyways. Assuming that the bridge is connected correctly (not as in the schematic print), and knowing that I am having stable 13V to the chip, where do you think this thing fails? I mean the way the CMOS schematic is - is there anything wrong with this? This is what I am asking mainly. 

 

[Staigen1]

Hi there Kain

Shame of you, sending up a shematic with such a serious error in it, and asking us to analyze it, and why it is not working as expected ! Shame shame shame
Back to the CAD software!

Ok now. Do not pay this much attention to the PSU part. I know it's not correctly drawn, but I know how to build it correctly and it does it's job well -

 

[Kain]

Ok. There is the schematic with the bridge rectifier connected correctly. By the way, if this sort of power supply is not using bridge rectifier I wonder why do we have this published in here:

http://www.electronics-lab.com/projects/power/022/index.html

This is exactly what I used without much change (maybe little change in the current limiting capacitor so I can get little more current).

View attachment 37536

 

[Staigen1]

Hi there Kain

So you dont see the error in the shematic? The thyristors Q1 and Q2 are connected to AC1, but the "ground"(the negative connection of C2) of IC1 is not! The gates of Q1 and Q2 shall have their steering pulses related to their catodes! The output of IC1 is related to the "ground", and not to the catodes of the tyristors, i belive that is the error! A simple solution would be to connect the "ground"(the negative connection of C2) to AC1( a short over the left diode in the rectifier bridge)

//Staigen

 

[Kain]

Ok, what happens if the gate of those thyristors is floating? I will try shorting though to see if this fixes it. Thanks for the reply.

 

[Staigen1]

Hi there again Kain

I'm not shure what happens, when you connect the thyristor the way you did, but it's not correct, thats for sure! BTW, isn't the resistors in the gates( R18 and R19) a little bit large(100K), maybe you should try smaller ones here(10K or so). I know the thyristors are highsensitive, but....?

//Staigen

 

[Kain]

In this case there is not need of lower resistance actually. The thyristors that are on the schematic are sensitive gate ones and require very little current to triger. The funny thing is that it actually works even the way it's made right now.  The things go wrong when I wire up the timer in some machine. Sometimes one of the timers retrigers (bounces) or sometimes both of them time wrong. I understand that the gates are floating, but the reason why the thyristors fire up is because they are sensitive gate ones. I believe that if they were regular it's likely they won't fire up. Just a guess though... So let me guess - you said in these cases we do not use bridge rectifier because we leave components float with respect to ground. Correct?  ;D

 

[Staigen1]

Yes, in similar constructions i have seen, the rectifier is only halfwave, with only two diodes, the anode of one diode connected to the common negative and the catode of the other diode connected to the powersupply's positive terminal!

//Staigen

 

[ante1]

Hi Kain,

I would try with pull down resistors for the gates to prevent false trigging!

 

[Kain]

Hi Ante! Yes I was thinking about that too since these are too sensitive, so I am guessing they might trigger from noise sources arround if there are any. I'll do this too and let you know if there is improvement  ;D

 

[ante1]

Hi Kain,

Yes, exactly my thought ! Please let us know!  ;D