shdesigns charger problem

[poulsbobill]

I have used the shdesigns charger for over a year and love it. I use the simple one. ( http://www.shdesigns.org/lionchg.html) The last few days though i have had a weird problem and hope someone here has an idea.

Now whenver i turn on/off a light switch on the same outlet as the power supply the charger stops working until i unplug the connection to the charging battery and reconnect it. Then it works fine until i turn on or off a light again. I can do this repeatedly. I have replaced the lm317 and the transistor. I have checked the power supply and is not affected, and i dont see a spike of voltage, but my dvm may not be getting a quick spike.

My guess is a small peak or trough of electricity is causing something? to reset? This is a pretty simple charger with no resets or on/offs as far as i know.

What part is resetting? and why in the first place?

What gives? I'm stumped!!!!

Thanks,

Bill

Update: I was able to use a diff power supply and not have the problem. So i guess the question is why does my power supply change voltage? or what? when things on the circuit are turned off and on?

[Dan Baldwin]

what kind of power supply are you using (the original)?

Dan

[poulsbobill]

Dan,

Its an old DTK computer power supply. It has worked fine for over a year. I have another charger i use with this power supply and when i switch lights on/off they are not affected. Still stumped by what on the circuit is turning "off" on the shdesigns charger and why? I still have voltage so ia think it is not the lm317.

Bill

[Dan Baldwin]

it might be helpful if you could check some voltages while the charger is not working. I would like to see the input voltage (from the power supply), output voltage(to the battery), and voltage across the current limit resistor (1 ohm in the original schematic).


Dan

[poulsbobill]

it might be helpful if you could check some voltages while the charger is not working. I would like to see the input voltage (from the power supply), output voltage(to the battery), and voltage across the current limit resistor (1 ohm in the original schematic).


Dan
Dan,

1. 12.25
2. drops about .04v from wherever it was in the charging process (~3.8-4.2 for a one cell)
3. 0

Thank you for taking an interest

Bill

[Chippie]

The voltages indicate that there is no output from the regulator and no current being drawn..fairly obvious I know....to me it looks like the 317 is hanging up....

What would I do?....change the 317 for a start...by pass the current limiting cct by shorting R1 and removing the transistor....it now becomes a voltage reg...place a load across the output and measure V out....then try and replicate the conditions that caused the problem...

Try this and let us know what happens please........

[poulsbobill]

The voltages indicate that there is no output from the regulator and no current being drawn..fairly obvious I know....to me it looks like the 317 is hanging up....

What would I do?....change the 317 for a start...by pass the current limiting cct by shorting R1 and removing the transistor....it now becomes a voltage reg...place a load across the output and measure V out....then try and replicate the conditions that caused the problem...

Try this and let us know what happens please........
Well there is voltage just no current (and thus no voltage across the resistor-and i have replaced the lm317). But i will try to bypass the transitor and see what happens.

Thanks for the idea!!

Bill

[RX5]

why not TRY a linear power supply to power the SHcharger.. thats MY setup...

[poulsbobill]

why not TRY a linear power supply to power the SHcharger.. thats MY setup...
I got what i got. But explain what that is please.

Bill

[poulsbobill]

The voltages indicate that there is no output from the regulator and no current being drawn..fairly obvious I know....to me it looks like the 317 is hanging up....

What would I do?....change the 317 for a start...by pass the current limiting cct by shorting R1 and removing the transistor....it now becomes a voltage reg...place a load across the output and measure V out....then try and replicate the conditions that caused the problem...

Try this and let us know what happens please........
Ok, I bypassed the transistor and it does not cut out now so it appears the problem is the transistor. So I replaced the transistor and the same thing happens. I have added a 2.2µF 50V 20% Nonpolarized Electrolytic Capacitor and a 22 uF Electrolytic Capacitor because thats what i had where the .1uf's go and that seems to help some but not every time. The plan calls for ceramic caps...i am not sure what the diffrences are.

This is maddening.

Any other ideas?

Bill

[xtal]

I would suspect the 1 ohm resistor has changed value, or another resistor..
then again capacitors can become leaky.....

As much time as you have probably wasted on this you could of built
up a new on.....,,, but you are not going to let it get the best of you,,
I've been there ......

[poulsbobill]

I would suspect the 1 ohm resistor has changed value, or another resistor..
then again capacitors can become leaky.....

As much time as you have probably wasted on this you could of built
up a new on.....,,, but you are not going to let it get the best of you,,
I've been there ......
Oh yeah...the point is to figure it out...

Latest...i have replaced the transistor and the .1caps and no change...what seems to help the most is the 22uf cap next to the transistor no doubt softening the voltage spike.

Do you know the difference s in caps? cermaic vs electrolitc vs...?

Bill

[JohnMuchow]

Do you know the difference s in caps? cermaic vs electrolitc vs...?

If the circuit calls for a 0.1uF ceramic and you're using 2.2-22uF electrolytics, that might still be the problem.

A smaller value ceramic cap is going to do a MUCH better job at "swallowing" high-frequency interference. The larger electrolytic will be "slower" not only due to its larger capacitance value but also its higher ESR (equivalent series resistance) than the small ceramic. The higher the ESR, the slower the cap charges and discharges. If it's too slow, the interference just bypasses the cap. or, if the caps in your circuit aren't there to just swallow interference or votlage pulses, the larger electrolytics you're using can affect the timing of a circuit compared to the smaller cap called for in the circuit.

I'd use a small ceramic cap before trying anything else. You probably don't need to match that 0.1uF value, but it's a really common value and easy to find.

[Edit] Just checked the schematic. Those 0.1uF caps are to suppress noise and those large electrolytics you're using can't come close to the performance of a small ceramic cap on high-frequency noise. You can keep the electrolytics in the circuit but definitely add the 0.1uF caps as close as you can to the regulator and the transistor. Keep the lead lengths of the caps as short as possible too. The longer the leads, the less effective the capacitor is at "swallowing" electrical noise and keeping that regulator from oscillating (which it will do without the caps).

You can check the datasheet for the regulator for its min/max capacitance ratings and the type of caps that are recommended.

[RX5]

bill,

WHY NOT re-build it using parts as-per-schematic?? I NEVER made the simpler one but I opted for the later(the one with -charging indicator) and it worked FIRST TIME.. I cant see why you cant... :)

[poulsbobill]

bill,

WHY NOT re-build it using parts as-per-schematic?? I NEVER made the simpler one but I opted for the later(the one with -charging indicator) and it worked FIRST TIME.. I cant see why you cant... :)
Obviously i can...as i have been using this for a year...again the point is to find out what has caused the problem and a fix.

Replacement will be the final result if i cant fix it.

Bill

[poulsbobill]

If the circuit calls for a 0.1uF ceramic and you're using 2.2-22uF electrolytics, that might still be the problem.

A smaller value ceramic cap is going to do a MUCH better job at "swallowing" high-frequency interference. The larger electrolytic will be "slower" not only due to its larger capacitance value but also its higher ESR (equivalent series resistance) than the small ceramic. The higher the ESR, the slower the cap charges and discharges. If it's too slow, the interference just bypasses the cap. or, if the caps in your circuit aren't there to just swallow interference or votlage pulses, the larger electrolytics you're using can affect the timing of a circuit compared to the smaller cap called for in the circuit.

I'd use a small ceramic cap before trying anything else. You probably don't need to match that 0.1uF value, but it's a really common value and easy to find.

[Edit] Just checked the schematic. Those 0.1uF caps are to suppress noise and those large electrolytics you're using can't come close to the performance of a small ceramic cap on high-frequency noise. You can keep the electrolytics in the circuit but definitely add the 0.1uF caps as close as you can to the regulator and the transistor. Keep the lead lengths of the caps as short as possible too. The longer the leads, the less effective the capacitor is at "swallowing" electrical noise and keeping that regulator from oscillating (which it will do without the caps).

You can check the datasheet for the regulator for its min/max capacitance ratings and the type of caps that are recommended.
JohnMuchow,

Now i am learning something! Thanks!! I built it to original specs and have had no problems for over a year. I added the larger caps as a lark. (frustration actually)

I thought that a higher uf value would help smooth the spike as it "holds" more juice? It does work better since i have added this 22uf cap though. What is high frequency interference vs low?

An interesting aside is...if i just turn on the charger and flip the light switch it shuts off...if i wait about 10 minutes and flip the light switch it will reduce the current draw by a few milliamps and then go back to normal... switch the light again within a few minutes and off again. I assume the cap is slowly charging?

How do you know it is high frequency interference and not a voltage spike? or?...
Is it possible to put a very large cap on the power supply terminals to help dampen the spike i am apparently getting...or are the .1uf as close as possible to the transistor the way to go?...i can definately shorten the leads to them. Is this interference causing the transitor to "turn off"?

Bill

[JohnMuchow]

JohnMuchow,

Now i am learning something! Thanks!! I built it to original specs and have had no problems for over a year. I added the larger caps as a lark. (frustration actually)

I thought that a higher uf value would help smooth the spike as it "holds" more juice? It does work better since i have added this 22uf cap though. What is high frequency interference vs low?

An interesting aside is...if i just turn on the charger and flip the light switch it shuts off...if i wait about 10 minutes and flip the light switch it will reduce the current draw by a few milliamps and then go back to normal... switch the light again within a few minutes and off again. I assume the cap is slowly charging?

How do you know it is high frequency interference and not a voltage spike? or?...
Is it possible to put a very large cap on the power supply terminals to help dampen the spike i am apparently getting...or are the .1uf as close as possible to the transistor the way to go?...i can definately shorten the leads to them. Is this interference causing the transitor to "turn off"?

BillTo be honest I haven't really read the thread closely enough to understand what your problem might be. Sorry. But, a scope is really the only way to know if you're having noise or spike problems.

High frequency vs. low totally depends on the circuit. For audio, it might be at several to dozens of kilohertz. For radar, it's gigahertz.

Don't forget that the light bulb, if still warm, can draw significantly less current during its startup surge than when it's cold. This can affect the size of any spike or voltage drop you might see.

If I was having this problem, I'd go back to the original schematic/specs and see if it then worked. If not, then a component is shot (probably the regulator or transistor). If it does work, then all your problems are solved. Trying to understand why a modified circuit doesn't work when you don't fully understand what the modification is doing (or how the original circuit worked and requires from its components) is a fast track to blown neurons.

Starting from a known point (the original schematic/specs which you know will work) can really help you. It often seems like it's unnecessary, but after a very short time racking your brain with quick fixes and all kinds of testing, it's time to step back and start from the beginning.

Other stuff...
- You MUST have a 0.1uF ceramic or 1uF tantalum cap directly across the input of the regulator. SHORT leads is a MUST here.
- A 1uF or larger cap on the output helps the regulator handle transient current draw but isn't absolutely necessary. Keep one on the output anyway.
- Adding more caps to the input and can help (a bit) to filter any noise coming into the regulator. It's better to have 2 caps and an inductor to filter that kind of noise though (on the V+ line): Incoming power>cap to ground>series inductor>cap to ground>regulator input cap>regulator.
- If you reversed polarity for even a fraction of a second, EVER, replace all components and start over. it's just not worth seeing what blew or not with a circuit like this.

Start over and get the circuit working from the original specs. If it doesn't work, find a scope and see what your problem really is. If that doesn't work, fix all possible causes and find another circuit. :)

[poulsbobill]

To be honest I haven't really read the thread closely enough to understand what your problem might be. Sorry. But, a scope is really the only way to know if you're having noise or spike problems.

High frequency vs. low totally depends on the circuit. For audio, it might be at several to dozens of kilohertz. For radar, it's gigahertz.

Don't forget that the light bulb, if still warm, can draw significantly less current during its startup surge than when it's cold. This can affect the size of any spike or voltage drop you might see.

If I was having this problem, I'd go back to the original schematic/specs and see if it then worked. If not, then a component is shot (probably the regulator or transistor). If it does work, then all your problems are solved. Trying to understand why a modified circuit doesn't work when you don't fully understand what the modification is doing (or how the original circuit worked and requires from its components) is a fast track to blown neurons.

Starting from a known point (the original schematic/specs which you know will work) can really help you. It often seems like it's unnecessary, but after a very short time racking your brain with quick fixes and all kinds of testing, it's time to step back and start from the beginning.

Other stuff...
- You MUST have a 0.1uF ceramic or 1uF tantalum cap directly across the input of the regulator. SHORT leads is a MUST here.
- A 1uF or larger cap on the output helps the regulator handle transient current draw but isn't absolutely necessary. Keep one on the output anyway.
- Adding more caps to the input and can help (a bit) to filter any noise coming into the regulator. It's better to have 2 caps and an inductor to filter that kind of noise though (on the V+ line): Incoming power>cap to ground>series inductor>cap to ground>regulator input cap>regulator.
- If you reversed polarity for even a fraction of a second, EVER, replace all components and start over. it's just not worth seeing what blew or not with a circuit like this.

Start over and get the circuit working from the original specs. If it doesn't work, find a scope and see what your problem really is. If that doesn't work, fix all possible causes and find another circuit. :)

JohnMuchow,

The circuit does work. It just shuts itself off when a voltage spike/lag hits the power supply and thus a change in voltage? to the charger causes it to shut itself off. It has worked for over a year with no problems. I have replaced everything but the resistors. I have not replaced the power supply, i am trying to work around this probably failing power supply.

As the other poster said i could just start over...i am trying to learn, explore, and figure out some electronics here. Thats why all the questions and i do appreciate the answers.

So to clarify...you think it is a regulator problem (lm317) and not the transistor --or do i misunderstand?

Can i make the inductor? or...

Bill

[JohnMuchow]

JohnMuchow,

The circuit does work. It just shuts itself off when a voltage spike/lag hits the power supply and thus a change in voltage? to the charger causes it to shut itself off. It has worked for over a year with no problems. I have replaced everything but the resistors. I have not replaced the power supply, i am trying to work around this probably failing power supply.

As the other poster said i could just start over...i am trying to learn, explore, and figure out some electronics here. Thats why all the questions and i do appreciate the answers.

So to clarify...you think it is a regulator problem (lm317) and not the transistor --or do i misunderstand?

Can i make the inductor? or...

BillYou can make the inductor but they're widely available already made. PM me your address and I'll send over one with the two caps that will form the noise filter. It will be useless though if noise isn't your problem (and a voltage drop to the supply is).

A couple of questions:
- Does the charger stop when you turn the light on or off? Both?

- If the problem occurs when the bulb is turned on, keep the bulb on for a couple of minutes to heat it up. Then VERY quickly turn it off, turn on your power supply and and charger, and turn the bulb back on. Does the charger still stop? If no, then the cool bulb is drawing too much current at start-up and is messing up the power supply and/or charger when it causes the AC line voltage drop. if yes, then I have no idea what's causing the problem.

- Replace the bulb with a much lower wattage one, like 25W (to reduce the startup current it requires which will reduce any voltage drop on the AC line). Does the charger still stop? If yes, then it's probably not a voltage drop to the power supply. If no, the I have no idea what's causing the problem. :)

- Put a LED (with 500-100 ohm resistor in series) or other indicating device on the put of the power supply. Does it flicker or go off when you switch the light?

- Move the power supply to another outlet (not at the same wall plate). Does the charger turn off when you turn the light off/on?

- But a new power supply from a place you can return it (CompUSA, wherever). Does it work with the new supply when the light is turned off/on?


If it is a voltage drop to the power supply that's causing the problem you might be able to reduce its affect on the charger (assuming the power suply doesn't shut off its output when the input voltage drops too low) by connecting every capacitor you can find to the input to the regulator to hold up the voltage to the regulator when it dips. The regulator will draw from the caps until the power supply comes back up to the proper voltage again. This might require thousands of uF though....tens of thousands even.

John

[Dan Baldwin]

Bill

Have you checked the 1 ohm resistor to make sure it's not opening up? You could also check to see if you have any bad solder joints around that resistor. What transistor are you using for Q1?

It should not be necessary to use any larger than the .1 mfd capacitors in the current regulation circuit. Electrolytics can have significant leakage, and that can affect the final voltage.

Have you tried hooking the charger up with a load (light bulb or resistor) instead of the battery pack? Does it still stop working? Have you tried tapping on the board and the components to see if you can make the charger fail that way?

Dan

[poulsbobill]

You can make the inductor but they're widely available already made. PM me your address and I'll send over one with the two caps that will form the noise filter. It will be useless though if noise isn't your problem (and a voltage drop to the supply is).

A couple of questions:
- Does the charger stop when you turn the light on or off? Both?

- If the problem occurs when the bulb is turned on, keep the bulb on for a couple of minutes to heat it up. Then VERY quickly turn it off, turn on your power supply and and charger, and turn the bulb back on. Does the charger still stop? If no, then the cool bulb is drawing too much current at start-up and is messing up the power supply and/or charger when it causes the AC line voltage drop. if yes, then I have no idea what's causing the problem.

- Replace the bulb with a much lower wattage one, like 25W (to reduce the startup current it requires which will reduce any voltage drop on the AC line). Does the charger still stop? If yes, then it's probably not a voltage drop to the power supply. If no, the I have no idea what's causing the problem. :)

- Put a LED (with 500-100 ohm resistor in series) or other indicating device on the put of the power supply. Does it flicker or go off when you switch the light?

- Move the power supply to another outlet (not at the same wall plate). Does the charger turn off when you turn the light off/on?

- But a new power supply from a place you can return it (CompUSA, wherever). Does it work with the new supply when the light is turned off/on?


If it is a voltage drop to the power supply that's causing the problem you might be able to reduce its affect on the charger (assuming the power suply doesn't shut off its output when the input voltage drops too low) by connecting every capacitor you can find to the input to the regulator to hold up the voltage to the regulator when it dips. The regulator will draw from the caps until the power supply comes back up to the proper voltage again. This might require thousands of uF though....tens of thousands even.

John

John,

The charger does stop whenthe lights are on or off. With the 22uf now only when turned on.

The stopping and starting is not affected by how long the bulb is on.

I have used a different power supply with no problems...but it is higher voltage thus more waste heat ect for the lmt317.

I have seen this ( http://www.radioshack.com/product/index.jsp?productId=2102508&cp=&sr=1&origkw=capacitor&kw=capacitor&parentPage=search) at RS. Enough?

I will try some of your other ideas. Especially the location of the caps.

I am still confused what would cause the transistor to turn off? Noise or spike?

Thank you,

Bill

[poulsbobill]

Bill

Have you checked the 1 ohm resistor to make sure it's not opening up? You could also check to see if you have any bad solder joints around that resistor. What transistor are you using for Q1?

It should not be necessary to use any larger than the .1 mfd capacitors in the current regulation circuit. Electrolytics can have significant leakage, and that can affect the final voltage.

Have you tried hooking the charger up with a load (light bulb or resistor) instead of the battery pack? Does it still stop working? Have you tried tapping on the board and the components to see if you can make the charger fail that way?

Dan

Dan,

I have checked the 1 ohm resitor and used other values as well with the same results. i have also checked all the connections.

The transistor is the rec one mps2222a.

I will try a diffrent load and see what happens.

I have tried tapping on things with no change.

The transistor just turns off. Seems odd that a spike or noise would do this. Not sure which.

Bill

[xtal]

I may have missed it, but what type of light are you flicking on/off...
flouresent????

The circuit may not be the problem,,,It may be your power source.or
very noisy light or somethhing else getting switched on and spiking the
line...

Try the circuit in another location thats on a different power circuit.

I have a simple [LM317 regulator connected to a transformer] on a hand terminal that shuts off with random voltage line spikes....

[poulsbobill]

I may have missed it, but what type of light are you flicking on/off...
flouresent????

The circuit may not be the problem,,,It may be your power source.or
very noisy light or somethhing else getting switched on and spiking the
line...

Try the circuit in another location thats on a different power circuit.

I have a simple [LM317 regulator connected to a transformer] on a hand terminal that shuts off with random voltage line spikes....
It is flourescent. I am somewhat stucjk where i plug this in at. Are you sure it is the lm317 that shuts down.

Bill

[xtal]

Yes---- The circuit is LM317 + transformer + 2 caps ---- have to cycle power to get it to turn on..... I think the 317 trips its protection circuit....

try using a quality surge protector. then put it on you big screen lcd when
you are done with it....

[poulsbobill]

I popped another transistor while i was testing things...I replaced the transistor and rewired things so i could get the caps closer. I tested it again and I still had the problem. I added the 22uf cap between the lm317 and neg (like it shows on the schematic for the .1uf cap---so now there is both caps) and it would not work any better.

So, I moved the 22uf cap between the transistor legs along with that .1 cap and now all seems ok. If i turn off/on the light within about 5 seconds then it goes off...But, If i wait a few seconds between on/off cycles it works now. My guess is, since you all believe it is the lm317 that is cycling that since this cap is closer to the lm317 it must be doing its thing that way charging in between my turning it on/off.

I appreciate all the help and all the ideas!!!! I will continue to play with it.

Bill

[Dan Baldwin]

What happens if you remove the 22 mfd capacitor, leaving only the .1 mfd?


Dan

[poulsbobill]

What happens if you remove the 22 mfd capacitor, leaving only the .1 mfd?


DanDan,
The problem comes back. Same old on or off and it resets itself. Very odd. I must have the cap electronically closer to the reg in this configuration.

Bill

[poulsbobill]

Other stuff...
- You MUST have a 0.1uF ceramic or 1uF tantalum cap directly across the input of the regulator. SHORT leads is a MUST here.


JohnMuchow, and to the group,

Why must caps be very close to the regulator? You stress this. Why?

Thanks,

Bill

[Dan Baldwin]

I haven't stessed it. My impression is that the cap across the emitter and collector of the transistor is only necessary to stabilize the current loop, not for noise. I don't think it makes any difference how far away from the transistor or the LM317 you put it. The capacitors on the input and output are to stabilize the LM317. Without theses caps, the circuit might oscillate. If you use very long leads, the LM317 might oscillate at very high frequency. I can't imagine how any of them could make the circuit "reset" and stop working unless the capacitor that's across the transistor has an intermittent short. You might try removing the .1 mfd cap and see if it has any effect.

How much current are you drawing when the circuit shuts down?

Dan

[xtal]

It appears to me that on/off switching that flouresent light is generating a LOT of electrical noise..and the louder it buzzes the more noise they make ..and that this noise is somehow passing through your powersupply into the charger circuit..I have seen bad ground connections cause similiar prob's.....then again spikes could be on + line...

What happens if you remove the flouresent bulbs and flip the switch????

Try a 100UF cap on the regulator input + the .1uf ceramic[HF bypass]

[Probedude]

Sorry if I missed it, but:

- what is your input voltage?
- What is your set output voltage?
- what cell count is your pack?
- What voltage is your pack at when it stops charging?

Dave

Edit: Nevermind, I found this

1. 12.25
2. drops about .04v from wherever it was in the charging process (~3.8-4.2 for a one cell)
3. 0


I was thinking you might be running out of headroom.

Just to confirm,
- you are charging a single Lipo cell?
- the circuit works fine with a different power supply?
- when you 'bypassed' the transistor you just removed it?

Can you make this modification for me?
See this document
http://cache.national.com/ds/LM/LM117.pdf
go to page 9, fig. 2. Add in these protection diodes.

[poulsbobill]

I haven't stessed it.
Dan
Dan,

When i say stress i mean "must". Sorry the word "must" be different to you than me. Sorry! My point was asking why you said - You MUST have a 0.1uF ceramic or 1uF tantalum cap directly across the input of the regulator. SHORT leads is a MUST here.

I am running about 700 mah of current.
Thank you,

Bill

[poulsbobill]

Sorry if I missed it, but:

- what is your input voltage?
- What is your set output voltage?
- what cell count is your pack?
- What voltage is your pack at when it stops charging?

Dave

Edit: Nevermind, I found this


I was thinking you might be running out of headroom.

Just to confirm,
- you are charging a single Lipo cell?
- the circuit works fine with a different power supply?
- when you 'bypassed' the transistor you just removed it?

Can you make this modification for me?
See this document
http://cache.national.com/ds/LM/LM117.pdf
go to page 9, fig. 2. Add in these protection diodes.
I am charging a double lipo

The circuit did work fine with a diff power supply. Which of course seems to point out a noisy fluctuating ps.

I ran a wire around the transistor leads to bypass it.

Just as an aside, i have charged as much as 2.0 amps with this charger with proper heat sink and fan. lm317 rated at 1.5 amps. i have charged 1-3 cells.

The diode addition sounds interesting.

Bill

[poulsbobill]

It appears to me that on/off switching that flouresent light is generating a LOT of electrical noise..and the louder it buzzes the more noise they make ..and that this noise is somehow passing through your powersupply into the charger circuit..I have seen bad ground connections cause similiar prob's.....then again spikes could be on + line...

What happens if you remove the flouresent bulbs and flip the switch????

Try a 100UF cap on the regulator input + the .1uf ceramic[HF bypass]
Nothing happens if i remove the bulb. I will try a bigger cap.

Bill

[Probedude]

Bill,
I'd like you to try a few other things besides adding in the reverse current protection diodes in the previous email.

- get your charger to 'lock up', then yank out Q1 while it is still powered. I don't want you bypassing it by jumping around it and then powering it up.

- If you can adjust your power supply output to above 12V, give that a try.

[xtal]

PROBLEM SOLVING


A Kansas farm wife called the local phone company to report her
telephone failed to ring when her friends called and that on the few
occasions, when it did ring, her dog always moaned right before the phone
rang.

The telephone repairman proceeded to the scene, curious to see this
psychic dog or senile lady. He climbed a telephone pole, hooked in his
test set, and dialed the subscriber's house. The phone didn't ring right
away, but then the dog moaned and the telephone began to ring.

Climbing down from the pole, the telephone repairman found:

1. The dog was tied to the telephone system's ground wire with a steel
chain and collar.

2. The wire connection to the ground rod was loose.

3. The dog was receiving 90 volts of signaling current when the number
was called.

4. After a couple of jolts, the dog would start moaning and then urinate.

5. The wet ground would complete the circuit, thus causing the phone to
ring.

This demonstrates that some problems CAN be fixed by pissing and moaning.

Just thought you'd like to know.

[JohnMuchow]

Dan,

When i say stress i mean "must". Sorry the word "must" be different to you than me. Sorry! My point was asking why you said - You MUST have a 0.1uF ceramic or 1uF tantalum cap directly across the input of the regulator. SHORT leads is a MUST here.

I am running about 700 mah of current.
Thank you,

BillI think I said that. :)
The cap can prevent oscillation of the regulator (a bad thing) if the regulator is more than 6" from the power supply's large filter cap. IMHO, the cap should always be included. It only costs a few pennies and prevent a lot of troubleshooting problems. This is why I say it's a MUST have item in the circuit.

Short leads are a must for smaller capacitors as they are typically meant to work on the higher frequency interference and long leads have (relatively) high inductance which prevents the caps from doing their job.

The same goes for placing the cap far from the regulator. If there's a couple of inches of cable between the cap and the regulator, that's the same thing as having capacitor leads a couple of inches long.

[poulsbobill]

I think I said that. :)
The cap can prevent oscillation of the regulator (a bad thing) if the regulator is more than 6" from the power supply's large filter cap. IMHO, the cap should always be included. It only costs a few pennies and prevent a lot of troubleshooting problems. This is why I say it's a MUST have item in the circuit.

Short leads are a must for smaller capacitors as they are typically meant to work on the higher frequency interference and long leads have (relatively) high inductance which prevents the caps from doing their job.

The same goes for placing the cap far from the regulator. If there's a couple of inches of cable between the cap and the regulator, that's the same thing as having capacitor leads a couple of inches long.

Ok I get it.

Thank you

Bill

[Dan Baldwin]

- You MUST have a 0.1uF ceramic or 1uF tantalum cap directly across the input of the regulator. SHORT leads is a MUST here.


Right, it was John that said that, not me. I am innocent of all charges.

Dan

[Probedude]

poulsbobill,
Did you ever get this solved?


Dave

[poulsbobill]

poulsbobill,
Did you ever get this solved?


Dave
Dave,

Yes. By adding a 20uf cap near the tx/lmt317 connection. When a light is turned on and off i can see the current draw go down on my dvm for a second and then back up. But it doesn't shut off. Good enuff!!!!

Thanks,

Bill