Hi all

As a young callow youth I remember, in one of the electronic mags, an LED based oscilloscope. I especially remember wanting to build one and the fact that the LED's were far to expensive at the time.

Well I stumbled across the circuit again recently and then many more variations on a theme. So the time has come to build one!! Firstly let me state this is a fun project just to keep me occupied and I am well aware of the limitations of using the end product as an oscilloscope.

Now the original has a 10 X 10 LED matrix as the display. An LM3914 x CD4017 are used to create the matrix. An LM555 IC acts as the clock. Now I fancy a 10 x 18 LED matrix so I need to cascade a pair of 4017's to do this. I am also using a 4011 IC as the clock and as an internal trigger. a second 4011 is to be used to 'join' the 4017's together so that they cycle from 1 to 18 repeatedly. I have drawn up 'my circuit' based on various schematics but just need an OK on the way I have setup the 4017's. Have I done it correctly?

I will also be using a 741 op amp circuit as an amplifier for the input to the 3914.

I attach the picture of the circuit.

regards

Fenris

 

Each column of LEDs will be lighted for only 1/18th of the total time so they will look very dim. The max current from the 4017 is only about 22mA so each LED will look as dim as one with a continuous current of only 1.2mA.

 

Nope, your cascading won't work with a single 4011, as those are NAND gates, not AND gates which is what is required. However, you could cascade a couple of NAND gates to emulate an AND gate:

Also, note that in your schematic, you have pins 11 and 12 wired together. They are both outputs; this is a "no-no". Pin 12 is not used in this application. Pin 13 is an input.

[eta]
Audioguru's correct; it'll be a pretty dimly-lit display. You should be able to see it in a dimly lit room, though.

My version was derived from this datasheet: http://www.synthdiy.com/files/2002/hef4017b.pdf

 

Hi there

Thanks for the heads up guys. I am aware the LED's won't be particularly bright, a bit like me judging by my cock-up on the drawings . Any hoo here is the corrected version for your approval!

regards

Fenris

 

Fenris,
Your clock generator won't work as you have it drawn (lower left corner).
Instead of using 4011 NANDs for the clock and the 4017 cascade AND, use 4093 NANDs; they have Schmitt-trigger inputs.

Then move the cap from between the two NAND gates, connect one end to the input of the leftmost NAND gate and the other end to ground.

 

Hi Sgt

It's more a question of what I have to hand component wise at the moment. But heeding your warning about the 4011 I have made a 10 X 10 unit that uses 1 3914 and 1 4017. Also the clock os a faithful ol' 555.

The LED matrix is constructed from 2 squares of vero board. These have had alternate strips removed and are now face to face with the tracks at right angles to each other. Each LED is mounted with their legs in adjacent holes which puts them at 45 degrees.

I currently have, roughly, a 33hz sine wave displayed on the matrix. I am also using a 741 pre-amp to boost the signal. I have probed my voice mod with it to see the square wave made by the 555 chip . LED brightness isn't a major issue although it will need a shroud in daylight if the sin is directly on it.

I have found another 4017 cascade circuit by Bill Bowden which looks promising. I fancy getting some 5x7 LED display modules maybe 4 for a 140 LED display? it would be easier than making the matrix myself

Again this is just a bit of fun for me. But the one I have now can at least indicate the presence of a signal even if not of any measurable use. I attach Mr Bowdens cascade 4017 circuit.

regards

Fenris

 

Hi there

Well I have found a circuit design that uses 3914, 4017 and 4049 chips. This is used to drive 9 LED matrix modules. Now the spec'd modules are LITE-ON LIP747 5 x 7 LED display modules. Now I am trying to find out what they are rated at, as rapid do not stock this specific make, so I can select a replacement from their stock. I believe I need column common cathode or row common anode depending on how each manufacturer references them (basically they are the same of course)

The Lite on site doesn't want to play ball with me at the moment so I am having trouble getting the spec's and google thinks I am after a boeing 747. Has any one got a recommendation as to what rapidonline stock item would be a suitable replacement. Green types also seems have less of a power requirement if that broadens the choice.

regards

Fenris

 

Hi Guys.

Heres the circuit I am looking at building. I have corrected the LF353 pin out for power but I am wondering where the GND is for it as it is only showing the +ve and -ve rails.

Sgt's resistor is Sgt Wookie's addition from an earlier question about the 7555 clock.

The 4011 U5-A is for setting the trigger as internal or external but I don't quite get how it works as it looks to me like if pin 1 and pin 2 are both getting +12V then the reset on the 4017's is going to be firing permanently?

The VSS and VDD pins of the 4049's seem a bit odd. On the ones that drive the rows the VDD pin 8 goes to +12V and VSS to GND on the 4049's that drive the column's the VDD pin 8 goes to GND and VSS goes to +12V. Is this Correct?

I have tried e-mailing the guy who designed the circuit but his website hasn't been updated in quite some time. There are images of the built unit and screens showing various wave forms.

http://www.geocities.com/rlaude2000/ledscope.htm

regards

Fenris

 

I have corrected the LF353 pin out for power but I am wondering where the GND is for it as it is only showing the +ve and -ve rails.

The LF353 doesn't need a ground reference itself. It's a dual-supply op amp.

The 4011 U5-A is for setting the trigger as internal or external but I don't quite get how it works as it looks to me like if pin 1 and pin 2 are both getting +12V then the reset on the 4017's is going to be firing permanently?

This schematic has a number of problems. I stumbled over it a few years ago. This is just one of them.

The VSS and VDD pins of the 4049's seem a bit odd. On the ones that drive the rows the VDD pin 8 goes to +12V and VSS to GND on the 4049's that drive the column's the VDD pin 8 goes to GND and VSS goes to +12V. Is this Correct?

No. For the 4049, Vdd is always on pin 1, and Vss is always on pin 8.

I have tried e-mailing the guy who designed the circuit but his website hasn't been updated in quite some time. There are images of the built unit and screens showing various wave forms.

http://www.geocities.com/rlaude2000/ledscope.htm

Hard to say if you'll ever get a reply. I imagine that the person has been inundated with E-mails over the years; they may have gone so far as to change E-mail addresses just to get away from all of it.

 

Hi Sgt

OK then so the LF353 Is OK as is.

The trigger. Consider the Forrest Mimms version. Pin 1 of the 4011 to row 1 of the matrix, pin 2 of the 4011 to column 1 of the matrix that would make more sense I think.

The VSS VDD problem. Wire these correctly then as they should be. I have read a bit about the 4049 and think I understand. If a positive signal is sent in it comes out negative and vice versa?

Is this circuit salvageable? the 4049 is the chip you mentioned as a candidate to cascade 4017's. Is this part of the circuit correct?

I think the 4011 trigger can be sorted as per Mimms. Or left out.

If you consider it pointless due to inherent design problems where do I go from here? joining 3914's to expand a matrix display seems a hell of a lot easier than cascading 4017's. I would really like to pursue 'knocking up' a solid state oscilloscope. The one I have at the moment is bloody great

Having removed the 741 pre-amp circuit The wave forms presented on the 10 x 10 display are much better formed. Obviously the pre-amp I was using was dirtying the signals I was investigating. I have it strapped to my voice mod at the moment and am watching the changing square wave on the display of the 7555 chip that controls the modulation rate. For proof of presence of a signal for basic fault finding I think it could be very useful in this instance. I have also been changing the capacitor of the scopes timer chip to be able to view frequency's outside of the fixed range with great success. I hope I am learning something

Your guidance would be much appreciated.Heres a pic of it in action. The Flash killed the display a tad but you can just make out the illuminated leds top and bottom forming the squarewave.

regards

Fenris

 

OK then so the LF353 Is OK as is.

Yes.

The trigger. Consider the Forrest Mimms version. Pin 1 of the 4011 to row 1 of the matrix, pin 2 of the 4011 to column 1 of the matrix that would make more sense I think.

That would have it continuously resetting the 4017s. It wouldn't work. As soon as the 4017 was clocked to column 2, the column 1 output would go low, which would bring the reset lines high. I do not know which of Forrest Mimms' circuits you are referring to, but you cannot post Forrest Mimms' circuit without his permission, as it is copyrighted material.

As the trigger is wired now, there are problems.
There needs to be a 10k resistor between the output of U5-A (pin 3) and the junction of the diode, the RESET line and the 15k pull-down resistor. This eliminates the output contention between U5-A pin 3 and U11, Q5 (you cannot wire the outputs of two logic gates together if there is a possibility that they will have different logic levels on them.) Also, the 15k pull-down resistor on the output of U5-A needs to be removed.

The VSS VDD problem. Wire these correctly then as they should be. I have read a bit about the 4049 and think I understand. If a positive signal is sent in it comes out negative and vice versa?

If a logic "1" is on the input, a logic "0" is output and vice versa. 4049s are logic inverters.

Is this circuit salvageable? the 4049 is the chip you mentioned as a candidate to cascade 4017's. Is this part of the circuit correct?

I don't remember mentioning 4049s as a candidate to cascade 4017 ICs; I did mention 4093 quad Schmitt-trigger NAND gates.

But when you're mentioning the 4049 ICs here, you're not talking about cascading the 4017 ICs; you actually mean the inverters on the outputs of the 4017 ICs and the LM3914s.

That part of the circuit is an ugly kind of kludge, as pull-up resistors had to be used on the outputs of the LM3914s in order to drive the 4049 inverters, and the 4049 inverters are expected to source and sink current to/from the LEDs without current limiting resistors. It'll work, but it's ugly; the 4049 ICs weren't designed to be LED drivers. To do it properly would require more work than I'm willing to invest.

I think the 4011 trigger can be sorted as per Mimms. Or left out.

Already discussed above.

If you consider it pointless due to inherent design problems where do I go from here? joining 3914's to expand a matrix display seems a heck of a lot easier than cascading 4017's. I would really like to pursue 'knocking up' a solid state oscilloscope. The one I have at the moment is bloody great

You could try it with the modifications already discussed. The trigger won't be great, as there is no protection for the input of U5-A, or provision for adjusting the slope, or compensating for variations in the input signal voltage level. The external trigger needs some additional work.

Having removed the 741 pre-amp circuit The wave forms presented on the 10 x 10 display are much better formed. Obviously the pre-amp I was using was dirtying the signals I was investigating.

741 op amps have poor characteristics. They are very old, very slow, and have a very limited input/output range along with being rather non-linear. They are not good enough for amplifying audio signals.

 

Hi Sgt

Thanks for your critique . OK then Let's get back to basics. I'll drop this design and go for tried and tested circuits. I have somewhere the data sheet for the 3914 which explains how to cascade them, pin 9 of the 1st chip to pin 1 of the next and pin 9 of the last one to pin 11 as per a single unit. But obviously there are a few other pins that need interfacing as well.

The 4017's cascade using the suggested 4093. This type of IC has a 'hard' switches between states unlike the 4011 I was originally going to use which wanders up and down slightly between states which may cause problems. Forget the trigger this is only a very simple 'oscilloscope'.

The 7555 timer circuit will be used so I can have the ranging option.

The LF353 I will use so I can control any input. Right I am going to attempt to draw it up.

I did breadboard the 4011 trigger circuit this morning and added it to the working circuit.......... I can't say it actually did anything much

thanks again for your input.

regards

Fenris

 

OK then here is my attempt at drawing up what I would like to build. The rows, driven by the 3914's are set for 15 rows. I think all that needs doing is adjusting R1 to account for this change in the number of pins used. The 4017's are setup to reset on the 15 count (pin 1 of the second IC). So if I use LED display matrix modules I will need 6 which will be set up 3 x 2 giving a resolution of 15 x 14 diodes.

I include the reference materials I have used to draw the circuit up. Have I got it right?

regards

Fenris

 

I really don't have time to look at this today. I took a quick look at it, but something in the 4017 area doesn't look quite right.

 

Hi Sgt

Possibly the Q0 of the second 4017 being present when in fact it isn't shown in the data sheet diagram. I should remove it and move the reset line up 1 to keep the 14 column count.

I shall await your approval, or not, before proceeding.

cheers

Fenris

 

You didn't use the reference designators properly, so everything is labeled "U" instead of "Un", where "n" is 1, 2, 3, etc.

At any rate, the right-side 4017 has Q5 tied to the left-side 4017 reset input. Therefore, as soon as Q5 goes high, it will reset the left-side 4017 which will in turn reset the right-side 4017. So, you'll never see an output from Q5, which is OK.

The part that is not correct is that the 4093 needs to be an AND function instead of a NAND function. You can get that from a 4093 like I showed in the previous schematic post by cascading two of the NAND gates in series. Otherwise, there will be a "glitch" when the signal goes from the 9th column to the 10th column.

One of the ideas of using the 4093 is that the 555 timer can be done away with altogether. Two 4093's in series with a resistor, a pot, and a cap are perfectly capable of generating a decent clock signal.

 

Hi Sgt

Here is the revision having read your post. Designators added! 4093 U5A-U5B set up as per your guidance. The remaining gates, U5C-U5D, are now set up as the clock. I have to figure out what frequency I should run at. Perhaps note the original 7555 clock circuits RC network and transpose the values. Capacitor C could be set up as a switched trio of caps to change the range or, if I understand it, change the timebase IE 10uS, 100uS and 500uS.

If the variable resistor of the clock is a 1M would it be correct to fit a smaller value variable resistor in series with the wiper to add fine tuning control?
The fixed resistor is sets the minimum frequency.

I have seen another variation where there is a resistor coming from U5C pins 8 and 9. Any benefit doing it that way?

U4 has had the matrix line removed from Pin 3(Q0) as per the data sheet diagram. The reset line from U4 has been moved up to the next pin and the matrix line removed from the same. As it stands this gives me 15 rows x 14 Columns. This is expandable to 20 Rows x 17 Columns with the components as is.

With regards to pin 3 of U4. why is it not used in each subsequent module?

thanks for your help

regards

Fenris

 

Hi Sgt

Here is the revision having read your post. Designators added! 4093 U5A-U5B set up as per your guidance. The remaining gates, U5C-U5D, are now set up as the clock.

Oops - that's not going to work. It needs to be more like this:

You can't use this circuit with a 4011, as that IC doesn't have Schmitt trigger inputs, where the 4093 does.

S1 shows how you could use a switch to change the range of frequencies by switching in different values of capacitance.

I have to figure out what frequency I should run at. Perhaps note the original 7555 clock circuits RC network and transpose the values. Capacitor C could be set up as a switched trio of caps to change the range or, if I understand it, change the timebase IE 10uS, 100uS and 500uS.

You could do that, I suppose. Keep in mind that you're not going to get very high frequencies out of this thing. 10uS is more than a bit optimistic.

If the variable resistor of the clock is a 1M would it be correct to fit a smaller value variable resistor in series with the wiper to add fine tuning control?

You could certainly do that if you wished.

The fixed resistor is sets the minimum frequency.

It sets the maximum frequency. The main purpose of the fixed resistor is to prevent overloading the output circuitry of the IC if the 1MEG pot is set to 0 Ohms.

I have seen another variation where there is a resistor coming from U5C pins 8 and 9. Any benefit doing it that way?

I suppose you could experiment with it and see which works better for you. Please keep in mind that I simply don't have time to explore all possible permutations of interconnections with this circuit. I suggest that you download some logic simulation program and experiment with various combinations. It's easier than blowing the lids off of your IC's.

U4 has had the matrix line removed from Pin 3(Q0) as per the data sheet diagram. The reset line from U4 has been moved up to the next pin and the matrix line removed from the same. As it stands this gives me 15 rows x 14 Columns. This is expandable to 20 Rows x 17 Columns with the components as is.

With regards to pin 3 of U4. why is it not used in each subsequent module?

Because it would be on the entire time U3 was counting.

 

Hi Sgt

Thanks very much for your help. I will get parts and bread board the circuit first. Which will allow me to iron out any unknowns and fine tune the setup. I can then fiddle with those 'variations' and see what happens just so I know .

Next time I post on this one hopefully it will be with a working design

Thanks again for your help. I'll look into the simulation software (running linux I should have access to loads of free stuff )

regards

Fenris

 

Hi guys

Parts are order to make the large matrix LED oscilloscope. But until then I put the original circuit from bread board to veroboard. I need only find a suitable container and its all done. 500K and 5K potentiometers dont seem to exist this side of the pond so I am using a 470K and 4.7K pot for sweep speed and controlling the vertical axis.

The main circuit board is piggy backed under the LED matrix using single core mains wire with its sheathing in place to act as a spacer. The wire is soldered to the bottom board and bent over the matrix board. In the event of a problem I should be able to part the 2 pieces easily.

I also bought some mini banana sockets and plugs and a pair of grasping probes. Heres some pictures of the item and you can just make out the sine wave it's being fed

regards

Fenris