La Crosse BC-900 Charger Test

[nikiwind]

I'm analysing it with scope and plotter (it takes time) and I'll post later scope and plotter screens with some conclusions.
May be I'll find the way for modifications to make it even better and I'll post it as well.

[nikiwind]

I have a glitch and I think I now why.
I'll try to do some modification on a next few days and I'll post suggestions pictures and a scope screens.

[Archangel]

I for one am looking forward to it.

[nikiwind]

I did the modification and now power (3V on the board) looks much better.
Now I have to run a few tests to make sure everything fine and then I'll post results.
Just one problem. It will be board pictures with modification and scope screens but looks like I can't upload it to the forum and I have no idea how to include it

[jtr1962]

Interestingly, my two chargers (the original and the replacement LaCrosse sent me) seem to be doing fine on a different power supply which can be obtained here. The chargers don't seem to mind 3.3V instead of 3V. Apparently the original supply is very glitchy which comes as no surprise given its small size (no room for large filter capacitors).

I also noticed that some components on the PC boards get very hot when running above 500 mA so I always have a small fan blowing on the chargers when running either 700 mA or 1000 mA. At first I thought this might be due to the higher voltage power supply but the chargers get just as hot on the stock supply.

I'm very interested in seeing what you came up with. If you have a place to host your pictures you can include them by using img and /img tags around the url of the image.

[Flash_Gordon]

Originally Posted by jtr1962

Interestingly, my two chargers (the original and the replacement LaCrosse sent me) seem to be doing fine on a different power supply which can be obtained here. The chargers don't seem to mind 3.3V instead of 3V. Apparently the original supply is very glitchy which comes as no surprise given its small size (no room for large filter capacitors).

Good find. Regulated switching power supply at over 4.5 amps ought to do the trick.

Is the connector the same size or did you have to change it?

Mark

[jtr1962]

Originally Posted by Flash_Gordon

Is the connector the same size or did you have to change it?

Same size and polarity. The only drawback is that the new power supply is physically larger than the charger and needs to sit on a flat surface. On the plus side since the plug is normal size you don't take up an extra space on a power strip as you would with the stock supply.

[Archangel]

jtr -
What issues were you having that you decided to try a different adapter?

[nikiwind]

I have a lot scope screens (TEK3000) and a voltage plots (FLUKE 196C) which will illustrate what's going on and how to fix it plus actual board before and after modification.
And I don't think another power supply can completely eliminate the problem. It is some PCB design problem.
But unfortunately I don't have a place to host all these pictures.
Any idea how I can post them?

[balazer]

imageshack.us? Or your ISP?

[Archangel]

Many people use ImageStation for a free site, but i'm not one of them, so can't give you any specifics.

[jtr1962]

Originally Posted by Archangel

jtr -
What issues were you having that you decided to try a different adapter?

The units were doing a hard reset when in refresh mode once two or more cells finished their refresh cycles. So far, so good in that both chargers haven't yet done this with the new power supplies.

[jtr1962]

Originally Posted by nikiwind

But unfortunately I don't have a place to host all these pictures.
Any idea how I can post them?

You might have a look at this thread. A lot of people, myself included, are very interested in your results.

[N162E]

Originally Posted by nikiwind

I have a lot scope screens (TEK3000) and a voltage plots (FLUKE 196C) which will illustrate what's going on and how to fix it plus actual board before and after modification.

But unfortunately I don't have a place to host all these pictures.
Any idea how I can post them?

http://www.pbase.com Its a few dollars a year to be able to post a lot of pictures.

[nikiwind]

Test post for the image.

This is a scope screen of the power before modification:

3V power before mod.

[nikiwind]

OK it works

This is the power after mod.


http://i.pbase.com/v3/54/594354/1/48935242.TEK00008.PNG

As you can see there is no glitches any more.
And these glitches were responsible for the "selfreset" and other funny behaviour when you have to charge more then one battery.

Within next few days I'll upload detailed picturestory with test results and explanation what should be done.

[SilverFox]

Hello Nikiwind,

Very interesting...

Tom

[nikiwind]

This is the first part of the story:

Test setup.

[nikiwind]

This is what happened.

When one battery was fully discharged and goes to the charge mode second one almost on the same time before actual full discharge occurs switch to the charge mode too.

[nikiwind]

Board (bottom view) before modifications



Board (top view) before modifications



Board (bottom view) after modifications









I think unit has this problem because of bad design.
No high volume low ESR (impedance) capacitors on the board, relatively long wires with some inductance from power supply and PW supply limited reaction time for the step load (overshot and undershot). All of it especially glitches (overshot and undershot) make the measurement system not reliable and this is the reason to selfreseting behaviour and other "glitches".

[balazer]

How common are these glitches? Are all BC-900s affected? Can the problem be reported to the manufacturer?

[nikiwind]

Three 470 uF capacitors are not really necessary. It could be just one tantalum cap > 100 uF rated for voltage higher then 6.3V.
Ceramic caps (very low ESR) are not necessary too but put them close to the switching MOSFETs for sure not make it worse.

For illustration a scope screens before and after modification.
As you can see after modification glitches completely gone.

Before (one battery):









Before (four batteries):









After:







Now it is under continuous test to prove it but what I've seen so fare it works.

Your comments are appreciated

[nikiwind]

Originally Posted by balazer

How common are these glitches? Are all BC-900s affected? Can the problem be reported to the manufacturer?

I have just one unit but I believe it is common problem for all of them.
At least my 20+ years like an electronics Eng. tell me about it

[Archangel]

Not that i'm not clutching at straws here, but what version number is yours?

[jtr1962]

Very interesting results. I also thought that they skimped on bypassing which is why I added a bypass cap at the power supply input. Right now my units have three 4.7 uF caps in parallel at the input plus the new power supply and seem to be doing OK. However, this mod was ineffective with the stock power supply.

I will try your mod also. You said only 1 470 uF cap was needed. Would you put it at the power supply input or elsewhere? Perhaps the 2.2 uF caps at the MOSFETs will work provided you have a lot of bypassing right at the power supply input. I plan to try 4.7 uF at the MOSFETs because I have a lot of them handy and something much larger at the power supply input. Are those 470 uF, 10V caps tantalum?

Another thing to help make the charger completely bullet-proof might be to put a few uF right on the supply lines for the microprocessor, assuming that you can find them.

Thanks for the testing and for sharing your results. You might want to report this to LaCrosse engineering. Maybe they'll send you a few free chargers.

[nikiwind]

Originally Posted by Archangel

Not that i'm not clutching at straws here, but what version number is yours?

32 like the everybody.

[nikiwind]

Originally Posted by jtr1962

Very interesting results. I also thought that they skimped on bypassing which is why I added a bypass cap at the power supply input. Right now my units have three 4.7 uF caps in parallel at the input plus the new power supply and seem to be doing OK. However, this mod was ineffective with the stock power supply.

I will try your mod also. You said only 1 470 uF cap was needed. Would you put it at the power supply input or elsewhere? Perhaps the 2.2 uF caps at the MOSFETs will work provided you have a lot of bypassing right at the power supply input. I plan to try 4.7 uF at the MOSFETs because I have a lot of them handy and something much larger at the power supply input. Are those 470 uF, 10V caps tantalum?

Another thing to help make the charger completely bullet-proof might be to put a few uF right on the supply lines for the microprocessor, assuming that you can find them.

Thanks for the testing and for sharing your results. You might want to report this to LaCrosse engineering. Maybe they'll send you a few free chargers.

I think even 100-200 uF tantalum cap will be sufficient.

I'd like to have schematic of this unit. In this case I definitely can make it perfect

It is almost impossible to catch it unless you sitting and looking on it for a few hours or using some plotting device like I did

[balazer]

This is all very good. It would be nice if you could give a dummie's guide to the simplest fix that would be good enough for most people.

[nikiwind]

Originally Posted by balazer

This is all very good. It would be nice if you could give a dummie's guide to the simplest fix that would be good enough for most people.

Open it from the bottom and solder one tantalum capacitor (>100 uF and > 6.3V) like shown on the pictures. It should be enough.

Go to the

http://www.digikey.com

type "tantalum capacitor"

then choose "Tantalum" under the "Capacitors"

then choose "Voltage-Rated" - 6.3V, "Packaging" - Cut Tape, "Features" - Low ESR, and "Capacitance" - > then 100 uF (for instance 330 uF)

Then "Apply Filters"

Then choose anyone. For example:

B45197A1337K509

Digi-Key P.N. 495-1516-1-ND

[balazer]

In your pictures I see capacitors added in three positions. Adding a single capacitor to any of those three positions will work?