Stumped :-( large voltage drop and logic not working - Analog electronics :-)

[aries1470]

Hi,

I am having issues with the power staying at 5V.

These are the steps that I have taken:

the timing circuit has been confirmed to be giving out 0.999 Hz. Confirmed by Hz counter on multimeter. It also gives 32.768KHz :-)

Voltage is supplied by a USB 2.5A adapter.
There is sag when all IC's are on the main board, and drops to around 4V depending on which supply I use.
I removed all IC's except for timing, Logic AND gates, and just 3 '192's.
I also then supplied the display separately, and that also drops the voltage to around 4v4

When both were powered from a single supply, it even dropped down to 3.8V

Now I have removed all the display’s except for 2, but still no joy.

What I can get, is IF I attach a short piece of Kynar wire wrap wire, to the Latch, it will jump around on the numbers but faint and as soon as I remove it, a number will show and even the dot point.

I do not have a logic probe or CRO to test, only a multimeter a Jaycar QM1535.

I have attached the Schematic and pcb files and images of the build.

I have checked for shorts, but couldn’t find any.
When there are less IC’s on the board, the power supply as counted at various IC sockets shows close to 5V, around 4.85V

Any help would be appreciated.

Are the traces the cause? That is the only thing I can think off, since I have changed over the 192’s and even left only 1 in its socket, but it still could not drive the 5082-7300. I had not had these type of problems when I was having it on a solderless bread board, but then again, I only had it attached as a counter ;-) not a whole system and had tried it with both the 5082-7300 and TIL311.

Thank you.

[PA0PBZ]

So both the display and the logic pcb make the voltage drop when powered separately? Maybe the USB power supply is not what it looks like, not capable of supplying 2.5A?
Use your multimeter to measure the current to the logic board and the display, let's see what we can learn from that.

[Gyro]

It may be might be my eyesight, but I can't see any supply decoupling caps on those boards (unless there is smd on the back.

Try adding some 100nF capacitors between the VCC and GND pins of the ICs (easiest do it on the back of the board). Those HP displays contain logic too and need the same treatment.

[ovnr]

I strongly suggest you try a different power supply, then.

[aries1470]

It may be might be my eyesight, but I can't see any supply decoupling caps on those boards (unless there is smd on the back.

Try adding some 100nF capacitors between the VCC and GND pins of the ICs (easiest do it on the back of the board). Those HP displays contain logic too and need the same treatment.

Hi Gyro,

No, no decoupling, but there is a chunky 1000uF on the small strip board and a 100uF on the main pcb.

I will try your suggestion, as I have a large amount of 0.01uF cap's at hand, similar to the one for the timing of the MC14490. :-)

I'll get to it sortly, and solder directly on the underside.
Bugger, I SHOULD HAVE stuck to my original design of having decoupling caps under the IC sockets.. mutter mutter....

[aries1470]

So both the display and the logic pcb make the voltage drop when powered separately? Maybe the USB power supply is not what it looks like, not capable of supplying 2.5A?
Use your multimeter to measure the current to the logic board and the display, let's see what we can learn from that.

I will try it right now    and report back shortly.

I have the following that I have tried for power supplies:
"Move" by DSE (Dick Smith Electronics"
Model: PW0053
Input: 100 - 240V 50/60Hz, 0,68A (Max)
Output: 5Vdc @ 2400mA (max)

Original Samsung (comes with SG S7E)
Adaptive Fast Charging:
Model: E-TA20HWE
Input: 100 - 240 50-60Hz 0.50A
Output: 9.0 V@ 1.67 A or 5.0V @ 2.0A

Samsung Original Charger (comes with SG TAB 2)
Model: ETA-P11JBE (not 100% certain as the writing is extra small!)
100 - 240V @ 50/60Hz @ 0.35A
5V @ 2A.

and last, but not least:
Motorola - Original power supply:
Model: GMT--TR20
Input: 100V - 240V 50/60Hz @0.1A
output: 5.0V @550mA
I have tried this one for logic only 

I have even fed the boards separately, still get similar resaults.

Will check how much they are drawing now :-)

[aries1470]

I strongly suggest you try a different power supply, then.

Thank you  I have tried a few of them. 
I will implement the other suggestions as mentioned by previous posters shorlty.

[bktemp]

A open TTL input gives a high level, therefore your diode network connected to the clear inputs will not work. At least some pulldowns are necessary at each input to force them to low level.

[aries1470]

Powered by the Samsung 2.0A:
122mA and 4.53V all ic except the 74ls42( as I didn't have that on hand even though I thought I did...)


Motorola has a hardwired cable:
 145mA @ 5.05V! Yep, performs better for some reason.
Changed thecable to another one, low and behold, it doesn't drop as much voltage:
482mA @ 4.74V with the DSE power supply - all 0's.
400mA all 0's with the Motorola.
4.82V for the 2.0A Samsung


But I am still not getting anything ticking over, even when I press the button to advance, it doesn't respond.
I powered the IC board with the motorola one, voltage is coming up nicely, at 5V.

I have also tested pressing the button for the minutes, it gets processed okay from the MC14490 and goes to PIN 5 of IC3, but the minutes do not change

Will adding the bbypass capacitors fix the problem, do I need to do anything on the Display board with the IC's? Not sure if the "LE" pin is needed or not
I will solder tomorrow morning the by-pass caps as it is getting very late here, and see if that will help

[aries1470]

A open TTL input gives a high level, therefore your diode network connected to the clear inputs will not work. At least some pulldowns are necessary at each input to force them to low level.

Can you recommend the value for the resistor to use and location please, if not to much to ask for  ?
How many would I need? Would one be enough?
I will need to solder them on the bottom of the pcb directly.

p.s. Here is a picture of how it lights up 

[aries1470]

Just remembered, am I suppose to add one end of the resistor to +5 or 0V/GND?

I know that many of you would be going  but, I am not that advanced yet  I am self teaching my self.

[bktemp]

A open TTL input gives a high level, therefore your diode network connected to the clear inputs will not work. At least some pulldowns are necessary at each input to force them to low level.

Can you recommend the value for the resistor to use and location please, if not to much to ask for  ?
How many would I need? Would one be enough?
I will need to solder them on the bottom of the pcb directly.

For a guaranteed low the voltage must be max 0.8V. The input low current is specified with max 0.4mA. 0.8V/0.4mA=2k. So 1k should be fine.
But wait, there is another problem: The guaranteed high level output current is only 0.4mA. A high level requires at least 2V. 2V/1k=2mA. That does not work. Even using 2k pulldown requires 1mA high current. It can work, but you are outside of guaranteed operating range. Maybe you could replace the LS TTL with some more modern HCMOS parts. They don't have any significant input currents. The also have a much lower current consumption.

[aries1470]

Hi,
Yep, back again.
I have added the bypass capacitors, but no joy. I ran it with only 3 (edit: 3 192's, and I was getting now 5.2v with the by-pass caps  )
I will now be trying to use a TIL311 and a single 5082-7300 on a breadboard to see if it is the PCB for the display or another issue all together 

Any other ideas? I am open to them

Oh, and I have now attached the bottom view of the PCB for those that asked

Also, once this is working, just wanting to make sure, that I will be adding the Pull-down resistors to GND and I will just solder on the underside direct to pin 8 (GND) and pin 14 (CLR), is that correct? 

[rstofer]

Yes, pull-down resistors go to ground.

The problem is, as pointed out above, pull down resistors need to have a fairly low value and, as a result, the logic gate might not be able to pull the line up to the high level threshold.  To compound the problem, you have several chips that all see the same signal so all of the pull-down resistors will be in parallel, reducing the combined value even further.

I don't think I have ever seen diodes mixed with TTL so I don't really have a solution.  It's the nature of things that we use pull-up resistors and switch contacts pull to ground.

[eetech00]

Hi

I see a few things that don't look right.

1. No bypass caps across each IC power supply pins.

2. No current limiting resistors for the LED displays.

3. Both S and R pins of the 4013's are tied together, then shorted across the power supply pins at IC10.

I didn't check everything but these items stuck out...

[Gary350z]

1. No bypass caps across each IC power supply pins.
2. No current limiting resistors for the LED displays.

The display datasheet says the displays have decoder/drivers built in, so they don't need current limiting resistors.

You need the bypass caps for reliable operation, but this won't pull down the power supply.

Post a picture of the project when it is done.

[Chris Mr]

I downloaded the data sheet for the displays which may be wrong HP5082-7400?

I couldn't find one on the data sheet I downloaded that matched the pin-out on your schematic.

The data sheet I downloaded says they are LEDs only with no current limiting so I probably downloaded the wrong data sheet.

Can you attach the data sheet you used for this project please.

[PA0PBZ]

http://www.s100computers.com/My System Pages/ZFDC Board/HP5082-73xx.pdf

[Chris Mr]

The data sheet for the displays says the latch input has to be low to make the input value get to the display and high to keep it there (if you want to make the data change for another device) and the LE pin is left open on the schematic.  The data sheet says the inputs are TTL / DTL compatible which means if left open the assumption is that the input is high - meaning changing the inputs won't alter the outputs.

You can connect the LE inputs to ground if the intention is that the input data is the required output (I haven't gone through the whole schematic to find out how it works).

That won't stop the current being high but might help to make it work.

If you connect your meter in series with the power supply what current reading do you get?  (make sure the probes of the meter are in the correct connectors for current)

[eetech00]

1. No bypass caps across each IC power supply pins.
2. No current limiting resistors for the LED displays.

The display datasheet says the displays have decoder/drivers built in, so they don't need current limiting resistors.

You need the bypass caps for reliable operation, but this won't pull down the power supply.

Post a picture of the project when it is done.

Look at figure 7, page 3-191, "Typical driving circuit for 5082-7304" on the datasheet.

[Gary350z]

1. No bypass caps across each IC power supply pins.
2. No current limiting resistors for the LED displays.

The display datasheet says the displays have decoder/drivers built in, so they don't need current limiting resistors.

You need the bypass caps for reliable operation, but this won't pull down the power supply.

Post a picture of the project when it is done.

Look at figure 7, page 3-191, "Typical driving circuit for 5082-7304" on the datasheet.

Looking at the datasheet:

Figure 2 "Block Diagram of 5082-7300 Series Logic" says pins 1,2,3,8 are logic inputs going into a latch. It also shows a "LED matric driver" driving the LEDs.

Figure 7 "Typical Driving Circuit for 5082-7304" shows pins 1,2,3,8 being driven by external LED drivers.

They can't be logic inputs and LED drivers at the same time.
I don't get it.

More investigation is needed to see exactly what these displays do.

[Chris Mr]

Gary350z,

Inside the display is a lot more than just an input connected to an LED.  The four inputs go into a decoder which selects the right LEDs to turn on to make the number 'look' right (for humans).  It also has a latch so that you can fool the decoder into thinking the number is still there when it isn't.

When the latch pin is low (zero volts) the decoder works from the pins.  When the latch input goes high whatever was on the pins is 'remembered' whilst the latch input stays high.

Much more complicated than just an LED display in a package.

Hope that helps

[Gary350z]

Gary350z,
Inside the display is a lot more than just an input connected to an LED.

Yes I know that. I was just stating that in one place the datasheet states that it has latches and drivers like you said, but in another place it shows the LEDs being driven by a an external driver. I'm sure with a thorough read of the whole datasheet it could be figured out.
 

[Chris Mr]

There is also a difference between the devices in the data sheet.

On the 7300 and 7302 the DP input is for the decimal point, on the 7340 that pin is a blanking control input.

It'd be nice to know which device you are using 

That doesn't get us to the point where it stops taking lots of current though, just helps to make it work.

As previously asked, how much current does it take when it drops the supply to 4v?

Oh, and as an engineering point, does anything get hot when that happens 

[eetech00]

The display datasheet says the displays have decoder/drivers built in, so they don't need current limiting resistors.

Actually, they do need current limiting resistors.
If you read a little lower on the last page, it explicitly states this:

"Note: LED current must be externally limited. Refer to Figure 7 for recommended resistor values."

You need the bypass caps for reliable operation, but this won't pull down the power supply.

The bypass caps have nothing to do with the excessive current draw.
Did you investigate the short circuit I mentioned in my earlier post?

Looking at the datasheet:

Figure 2 "Block Diagram of 5082-7300 Series Logic" says pins 1,2,3,8 are logic inputs going into a latch. It also shows a "LED matric driver" driving the LEDs.

Figure 7 "Typical Driving Circuit for 5082-7304" shows pins 1,2,3,8 being driven by external LED drivers.

They can't be logic inputs and LED drivers at the same time.
I don't get it.

More investigation is needed to see exactly what these displays do.

Actually, they ARE both logic inputs AND LED driver inputs. The logic inputs are same inputs providing current to the LED's as well. The input are decoded in BCD or HEX depending on the chip.