Update:

Made some progress of sorts..... started playing with my multimeter (MM) and a variable power supply (VPS) which led me to a problem.

Started by just testing the battery pack output (which I hadn't done b4) and the MM said it was putting out 5.5V. (not sure if this is normal for two bridged 3V batteries?) Then I connected the leads to the BB and tested the voltage at the next pin holes..... MM said 2.9V. "That's strange" I thought.... would the components on the circuit be causing this for some reason?

So I repeated the experiment with my VPS set at 5.5V..... (MM said 5.63). Used it to power the circuit and again tested voltage at the next pin holes..... MM said 4.9V..... exactly what it should be after going across a single IN4001 diode.

The circuit was still behaving erratic..... doing all kinds of stuff. So I started turning down the VPS and when I got close to 4.5V everything suddenly worked fine. (The 8 LED's latched one by one, and then shut off one by one.) At 4V the LED's that still did light where barely on. I guess the 74HC164 is very finicky about how much juice you pump into it?

So, even though the batteries had not seen that much use, I inserted new ones..... MM said output was now 6.7V. Attached it to the BB and had to go thru 2 diodes b4 the circuit functioned properly again. I left it running for 100 minutes and then decided to check what the voltage was actually running the board.

While I was attaching the MM the circuit suddenly froze up with 7 LED's lit..... Did I do that?

Went ahead and checked the next pin holes and MM said there was 3.1V going into the BB. Checked the battery pack and MM said output was now down to 5.6V. Plugged the battery pack back into the BB by-passing one diode and the circuit began working as b4. MM reports fluctuating between 3.6V and 4.2V going into the board.

These batteries are rated at 220mAh..... if I have to change batteries every 2 hours that's not cool. Kinda hard to by-pass a diode once the finished circuit board is soldered too.

I'm using pretty much the exact circuit that Allen offered previously, and I'm using a 3909 to clock the 74164..... I'm open to suggestions on how to improve battery life and stability.

 

Coin cells are designed for low current applications, such as watches. At Vcc=5V, your current is about 70mA peak, 35mA average. The internal resistance is too high for this current level, and the amp-hour rating also drops at these current levels. Check out some CR2032 datasheets.
I would use AAA or AA cells.

EDIT:
Why are you using a 3909 for a clock source? Are you also flashing an LED with it?
Are you using an HC164, or a TTL part?

 

Coin cells are designed for low current applications, such as watches. At Vcc=5V, your current is about 70mA peak, 35mA average. The internal resistance is too high for this current level, and the amp-hour rating also drops at these current levels. Check out some CR2032 datasheets.
I would use AAA or AA cells.

EDIT:
Why are you using a 3909 for a clock source? Are you also flashing an LED with it?
Are you using an HC164, or a TTL part?

Hello again Ron....

Looked at trying to squeeze 4 AAA cells into the model kit, and that will never happen without major surgery.... thus I decided to see if the two CR2032 cells would work for me.

When finished, the light show won't be running continuously..... it's just for occasional "eye candy". I was hoping however that my battery life would be greater as I'll also need to light up some additional LED's and possibly a couple of strips in the model.

I went with the 3909 to reduce the size of the finished circuit..... although that might not be the best choice as far as current usage goes? I won't be blinking an LED with it, just sending the signal to the 74HC164, which it does achieve. Would my previous AMV clock require less battery draw?

Thanks,
Wolf

 

I miscalculated your current. The average current, if running continuously, should be about 25mA.
I would change R10 to 100k. Paradoxically, this causes a tiny increase in average current, because, although the peak base current drops from 3.7mA to 40uA (averages are half those values), the current in LED #8 goes up significantly, due to the reduced loading by R10 on the HC174. If you don't make this change, LED #8 will be dimmer than the other 7 LEDs.
I would probably use a CMOS 555 for the clock generator. Why is the 3909 any smaller?
I don't have a solution to your battery problem.
What kind of model does this go in? You may have answered this already.

 

So, even though the batteries had not seen that much use, I inserted new ones..... MM said output was now 6.7V. Attached it to the BB and had to go thru 2 diodes b4 the circuit functioned properly again. I left it running for 100 minutes and then decided to check what the voltage was actually running the board.

I've read most of the replies from here as well as ETO. You're lucky that the circuit can last for 100 minutes without failing. For a circuit using 8 LED and all lighted at some instances and run only by 2 coin cells like CR2032 is just not practical.

I'm using pretty much the exact circuit that Allen offered previously, and I'm using a 3909 to clock the 74164..... I'm open to suggestions on how to improve battery life and stability.

Most have pointed out that you have to solve the supply problem. Me too don't have a solution for that. If you look at the proteus schematic I attached. You can see that the major current consumers are the LEDs. The 555 only consumes around 7 mA. The current used by the transistor is negligible. I have also added a RC circuit to delay the reset of 164 so the 8th LED can be seen lighted. Haven't tried on the BB so I dont know how well it would work.

Others also suggested using the CMOS 555 (aka 7555) which should consume less current than the AMV circuit.

I also attached the circuit that simulates the LM3909. I used that to flash an LED for months with 2 AA battery.

Allen

 

I miscalculated your current. The average current, if running continuously, should be about 25mA.
I would change R10 to 100k. Paradoxically, this causes a tiny increase in average current, because, although the peak base current drops from 3.7mA to 40uA (averages are half those values), the current in LED #8 goes up significantly, due to the reduced loading by R10 on the HC174. If you don't make this change, LED #8 will be dimmer than the other 7 LEDs.

Most of what you're posting here is Greek to me Ron... I'm very new to this field..... but I am learning a lot from you guys. :b

I would probably use a CMOS 555 for the clock generator. Why is the 3909 any smaller?

The only 555's I have are NE555P which MrChips explained to me are not CMOS and are very power hungry. I'll be doing some shopping soon and the CMOS version is on my list.

I don't have a solution to your battery problem.
What kind of model does this go in? You may have answered this already.

The model I'm building is the Leif Ericson Galactic Cruiser. The first circuit board I built was a fun learning experience, but I had to try and make both the circuit and the power supply smaller because of the cramped space I have to deal with in the models hull.

I would prefer to display the finished model hanging, so I would need a self contained power supply to accomplish this. If all else fails, I'll build landing gear or a stand and go with a wall wart AC adapter.

 

Hi Allen.....

I tried building that 3909 facsimile in Proteus a while back but I could never get it to run.... I've seen quite a few cases now where things don't work in the software but do work with hardware, and vice-verse.

My intention when buying these little coin battery packs was never to use them to power this much junk, but space constraints in the current project forced me in that direction. I'm a bit amazed too that I've gotten this far with them.... and that's mostly due to you folks answering a lot of my Q's. (The circuit actually ran longer than 100 minutes if you continue reading).

As I posted recently I had installed new batteries and the circuit only ran for 100 minutes. What I did not post is that I then by-passed one of the diodes and the circuit ran for another 45 minutes. Then I by-passed both diodes and the circuit ran for a last 15 minutes. In all three cases voltage tested at the BB entry had dropped to 3.1V.

I left it laying there and turned it back on this evening (about 24 hours later) and the circuit began running fine.... which surprised me since I thought those batteries were toast. Voltage at the BB was back up (fluctuating) between 4V & 5V, and testing the battery pack alone was back up to 5.85V. (previously 5.5V).

Are these batteries draining fast, or over heating?

Someone at another forum I've been visiting offered up a schematic to make the 74164 circuit behave just like the previous 4017/thyristor version. I tested it on my BB a while ago but using 10K instead of 33K resistors and it worked great.... (with "dead" batteries). He said if I go to 33K that the current draw would be much reduced. What do you think?

Thanks,
Wolf

 

I have not had the time to breadboard a 74HC164 circuit for you. If you wish to minimize the power consumption I would use CMOS inverters instead. The added advantage is that you can use the spare inverters to make a CLOCK circuit. This eliminates the 555 timer.

 

I have not had the time to breadboard a 74HC164 circuit for you. If you wish to minimize the power consumption I would use CMOS inverters instead. The added advantage is that you can use the spare inverters to make a CLOCK circuit. This eliminates the 555 timer.

Hey Chips,

Even though I recently (all but) begged Allen to post a pic of his BB, that's probably not the best thing for me in the long run..... we all retain our lessons best when they are worked for. Probably better I just keep looking at schematics for now and stumble through them till I get it right?

I imagine that a CMOS inverter is some sort of IC that replaces the transistor/resistor/resistor method? I'll need to look that up.....

Wolf

 

The two transistors in your circuit are serving the function of inverters. You can replace these with a single 14-pin DIP IC, CD4049, CD4069 or 74HC04.

You can replace the 555 timer circuit with an oscillator circuit using three of the remaining four inverters on the same IC.

 

The two transistors in your circuit are serving the function of inverters. You can replace these with a single 14-pin DIP IC, CD4049, CD4069 or 74HC04.

You can replace the 555 timer circuit with an oscillator circuit using three of the remaining four inverters on the same IC.

It might save you some space, but no matter what you do, over 90% of your power will be going to the LEDs.
You could save a LOT of power by sequencing one LED at a time.

 

@Chips: How on earth do you remember all these part designations let alone what each component is capable of??? Each time I'm done looking up the last number you put out there you fire off 2 or 3 more..... :b

@RonH: I'm sure you're correct Ron..... but I do like the effect of a bar light over the blinking LED's. I actually like the "sequential on/sequential off" effect even more, but that circuit probably uses at least 50% additional current so I'll save that for a bigger power supply down the road.


Wolf

 

@Chips: How on earth do you remember all these part designations let alone what each component is capable of??? Each time I'm done looking up the last number you put out there you fire off 2 or 3 more..... :b

@RonH: I'm sure you're correct Ron..... but I do like the effect of a bar light over the blinking LED's. I actually like the "sequential on/sequential off" effect even more, but that circuit probably uses at least 50% additional current so I'll save that for a bigger power supply down the road.


Wolf

Sequential on/sequential off takes the same amount of average current as sequential on/all off/sequential on/...

 

That's why they call me MrChips.

 

Did you try the circuit designed by alec_t ? He used HC132 instead of inverters. I just add a switch on the MR input so you can select the LED sequence in one of the 2 modes you want.

Attachment 1: alec_t original design
Attachment 2: simulated on ISIS with a switch added to select modes.

Allen

 

Sequential on/sequential off takes the same amount of average current as sequential on/all off/sequential on/...

Ron.... are you sure about that? Cuz it doesn't seem practical.

If the clock is one tick per second, then with Seq/on-Seq/off the last LED is lit 8 seconds longer than it would be, 7 seconds for the next, etc....


Wolf

(Oh hell... you are right again Ron.... of course the power consumption after 8 clock ticks would be the same whether the LED's are being fired, or shut down.)

 

Did you try the circuit designed by alec_t ? He used HC132 instead of inverters.

No Allen.... I didn't try Alec's 2nd schematic using HC132 because;

A) I don't have any of those.

B) I have no clue what they are or how they actually work.

Are they like a transistor or a thyristor?


Wolf

 

Hey guys.... got another Q for ya.....

If I go ahead with using the 3909 to clock the circuit, I figured I won't need the flashing LED.... just the clock pulse.

So why is it that when I replace the LED with a jumper across 6 & 8 the circuit quits working?

Actual "clock" schematic I'm using below with pin 8 feeding pin 8 on the 164.

Thanks,
Wolf

 

No Allen.... I didn't try Alec's 2nd schematic using HC132 because;

A) I don't have any of those.

B) I have no clue what they are or how they actually work.

Are they like a transistor or a thyristor?


Wolf

a) You can get them here
http://www.digikey.com/product-detail/en/SN74HC132N/296-1573-5-ND/277219

b) It is a quad-NAND schmitt trigger gates. If you have 4093 or 4011, you can try them and they might work too.

Attached is its datasheet.

Allen

 

Hey guys.... got another Q for ya.....

If I go ahead with using the 3909 to clock the circuit, I figured I won't need the flashing LED.... just the clock pulse.

So why is it that when I replace the LED with a jumper across 6 & 8 the circuit quits working?

Actual "clock" schematic I'm using below with pin 8 feeding pin 8 on the 164.

Thanks,
Wolf

The LED is an integral part of the oscillator. As you discovered, the oscillator won't run without it.