555 timer pulse width problem

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Hi,

some time ago I designed a circuit which is supposed to generate a short pulse (about 1 ms) upon pressing a tactile switch. In addition, the pause between two successive pulses should be about 2 seconds. I came up with the circuit attached, using one 555 (U2) to generate the pulse and a second
555 (U1) to hold the trigger of the first 555 low for about two seconds. Maybe there are more optimal solutions for that (I'm pretty sure there are) but
the circuit worked as expected when I built it up using THT components (on a PCB).

However, I wanted to get more experienced with SMD soldering. So I decided to create an SMD version of it which is based on the same
circuit.

After I completed the SMD version, it showed some unexpected behaviour in terms of the pulse width generated by U2. Instead
of having a pulse width within the theoretical limits of about 0.5 ms and 1.6 ms (depending on how RV1 is adjusted), I get a pulse width
of about 6.5 ms - no matter how I adjust the pot RV1. Since the timing of U1 is working as expected I suspected that it may be
a problem with U2 or the components connected to it.

I then measured some voltages in the THT and the SMD version and I came up with the following roughly equal measurements for both versions:

VCC: 6.2 V
CV: 4.1 V

So, these values should be fine.

In addition I had a look at the charging curve of cap C9 and there I noticed the significant difference between the two
circuit versions: in the THT version the cap charges up to roughly 66% and then discharges again as expected after about 1 ms (see,
Cap_Charge_THT.png). But in the SMD version the cap seems to get fully charged. And after a while the cap gets discharged a bit (after about 6.5 ms).
It then seems to charge up again for a short period and after reaching the threshold voltage, it discharges completely (see
Cap_Charge_SMD.png).

So to me it seems like the 555 U1 is not detecting correctly the point where the cap is charged above the threshold voltage.

For your information - the timer chips used:
THT: TLC551CP
SMD: TLC555QDRQ1

And as a side note: at first I suspected that there might be a problem with R2 and/or RV1. Hence, I also measured the resistance over them
and it seems to be fine (between 470 and 1470, depending on the settings on RV1). The charging curve of the cap also changes as expected
when adjusting RV1.

Maybe it's some difference between the two timer chips. But I can't see a hint in the datasheets.

Does someone have an explanation for that behaviour I see? It's not an important circuit for me since I'm a beginner and I'm still
learning. But it would be good to understand what's going on in the SMD version.

Best,
Michael
 

Attachments

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Hi,

both versions are based on the same circuit. Just the PCBs are naturally different.
In addition, I measured everything I could think about around the pulse generating 555 and
everything seems to be connected correctly.

And I probably should mention that at first I thought that I maybe fried U2 during soldering.
So I assembled a second PCB but the problem remained. So I can conclude that the 555 is
most probably not fried and there is also no short due to excessive solder.

Best,
Michael
 

dl324

Joined Mar 30, 2015
14,908
Welcome to AAC!
In addition I had a look at the charging curve of cap C9 and there I noticed the significant difference between the two
circuit versions: in the THT version the cap charges up to roughly 66% and then discharges again as expected after about 1 ms (see,
Cap_Charge_THT.png). But in the SMD version the cap seems to get fully charged. And after a while the cap gets discharged a bit (after about 6.5 ms).
It then seems to charge up again for a short period and after reaching the threshold voltage, it discharges completely (see
Cap_Charge_SMD.png).
Examine the wiring/soldering on the anode of C9. When the threshold input gets to 0.67Vcc, the cap should get discharged.

You should avoid drawing unnecessarily long ground nets (that would avoid some unnecessary wire crossings), avoid unnecessary wire bends, and print snapshots from PDF so we don't have to look at colored schematics or the grid.
 

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Welcome to AAC!
Examine the wiring/soldering on the anode of C9. When the threshold input gets to 0.67Vcc, the cap should get discharged.

You should avoid drawing unnecessarily long ground nets (that would avoid some unnecessary wire crossings), avoid unnecessary wire bends, and print snapshots from PDF so we don't have to look at colored schematics or the grid.
The wiring/soldering seems to be ok. There is a definitely a connection between the anode of C9 and
the 555 discharge and threshold pins. I also desoldered the cap and resoldered it, but this didn‘t change
anything. I also just measured through an empty PCB to verify that the routing is correct and I found
no problem.

Regarding your tips: thanks, I will keep them in mind when posting a circuit the next time.
 

dl324

Joined Mar 30, 2015
14,908
There is a definitely a connection between the anode of C9 and
the 555 discharge and threshold pins. I also desoldered the cap and resoldered it, but this didn‘t change
anything.
Instead of doing things willy-nilly, determine what's not working right and make fact-based choices.

When the capacitor voltage charges to 4.13V, the discharge transistor should be turned on. Measure the voltage at the anode of the cap and at pin7 pin6. They should be connected, but I can't know if that's the case. If the capacitor doesn't get discharged at 0.67Vcc, check for problems with discharge.
EDIT: corrected typo. Meant to say pin6 for threshold, though pins 6 and 7 should be connected.
Regarding your tips: thanks, I will keep them in mind when posting a circuit the next time.
It would also be helpful if the schematic editor you're using had a more useful 555 timer symbol. Inputs should be primarily on the left and outputs on the right to facilitate having a predominantly left to right flow in schematics.

I use Eagle and the symbol they used wasn't very useful, so I created my own.
 
Last edited:

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Instead of doing things willy-nilly, determine what's not working right and make fact-based choices.
Maybe it sounded like I did things randomly without thinking about it. But one of my thoughts was that there may be
a problem with the cap. So I desoldered it and replaced it with the model I used in the THT version to see if the different caps
may cause the different behaviour. Once I saw that both caps behave the same I resoldered the SMD cap.

When the capacitor voltage charges to 4.13V, the discharge transistor should be turned on. Measure the voltage at the anode of the cap and at pin7. They should be connected, but I can't know if that's the case. If the capacitor doesn't get discharged at 0.67Vcc, check for problems with discharge.
The anode and pin 7 are connected. In fact, the anode of the cap is connected to pins 6 and 7, like it should be.

It would also be helpful if the schematic editor you're using had a more useful 555 timer symbol. Inputs should be primarily on the left and outputs on the right to facilitate having a predominantly left to right flow in schematics.
I use Kicad and I honestly didn't think too much about that. But I'm always thankful for useful tips.
 

dl324

Joined Mar 30, 2015
14,908
I use Kicad and I honestly didn't think too much about that. But I'm always thankful for useful tips.
I finally got around to reading your schematic. I didn't do it initially because your style is messy.

You want a pulse in the millisecond range, but you have the trigger direct coupled. You should AC couple the trigger so the time you have the switch depressed won't affect timing. From Signetics:
1658595867619.png
1658595881792.png
1658595895663.png
 

Thread Starter

hexmax78

Joined Jul 13, 2022
8
I finally got around to reading your schematic. I didn't do it initially because your style is messy.

You want a pulse in the millisecond range, but you have the trigger direct coupled. You should AC couple the trigger so the time you have the switch depressed won't affect timing.
Sorry for my messy circuit style, it's actually the first one I did. Suggestions on how to create less messy circuits are always welcome.

Also the description in my initial post contains a small mistake. U1 and U2 are supposed to go high at the same
time (since the trigger inputs are connected). And as long as the output of U1 is high (about 2 seconds), the
trigger input of both 555 timers is pulled up to VCC through Q5. I also verified that by looking at the trigger voltage which
goes up almost instantly once the output of U1 goes high. So, I guess that my circuit meets the requirement of
the trigger voltage going over 2/3 VCC before retriggering. And as also stated, the THT version of the circuit works
perfectly. So I don't understand how this could be the problem in the SMD version.
 

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Check the pot connections.
Is one pin not connected?
If so, it shouldn’t be, and it should be connected to the wiper pin.
Yes, indeed, the unused pin of the pot is not connected to anything. However, I can not see how this could
change anything? It‘s also not connected in the THT version and that version works as expected.
 

crutschow

Joined Mar 14, 2008
30,455
Yes, indeed, the unused pin of the pot is not connected to anything. However, I can not see how this could
change anything? It‘s also not connected in the THT version and that version works as expected.
Yes, if the pot is working correctly the unused pin can be left open.
The main reason to connect it is, if the pot wiper momentarily loses contact, the maximum circuit resistance becomes the maximum pot resistance, not an open circuit (infinite resistance).
In some circuits that could be a problem, but probably not here.
 

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Yes, if the pot is working correctly the unused pin can be left open.
The main reason to connect it is, if the pot wiper momentarily loses contact, the maximum circuit resistance becomes the maximum pot resistance, not an open circuit (infinite resistance).
In some circuits that could be a problem, but probably not here.
I read about those possible issues in the meanwhile. Unfortunately after designing
that circuit. But in future circuits I will connect the unconnected
pin to the wiper pin. Just to be sure.
 

Thread Starter

hexmax78

Joined Jul 13, 2022
8
Hi again,

I finally found the problem. According to the schematic C5 is a 47uF cap. But it sems like
I exchanged that cap in my THT version with a 1uF cap without changing the schematic accordingly.
Hence, I used the wrong cap in my SMD version. After replacing it with a 1uF cap, everything
works as expected now.

@dl324: thanks again for your hint. I found the problem after probing the trigger voltage again.
It simply didn‘t go up fast enough after triggering.

Best,
Michael
 
Top