Confusing design strategy in a car battery charger

Thread Starter

Spottymaldoon

Joined Dec 4, 2015
88
I've been trying to figure out a defective car battery charger and am confused by the design strategy they've used (I don't have a schematic). So far as I can see the incoming 120V mains is first of all transformed into a 170v DC and then, using a power FET oscillator, converted back to ac and put through a transformer whose output is then rectified to give the approximately 13V needed to charge the battery.

Can somebody please explain why such an expensive route has been taken - rather than, as in the old days, just transforming and rectifying the incoming ac to 13V DC?
 

crutschow

Joined Mar 14, 2008
30,102
Can somebody please explain why such an expensive route has been taken - rather than, as in the old days, just transforming and rectifying the incoming ac to 13V DC?
Likely because it's not more expensive,
In the "old days", there was no other choice to using a line powered transformer.
Today, the cost of a large, low-frequency AC transformer to perform the conversion directly is more than the cost of the small, high-frequency switching transformer and electronics to generate the 13V.
The size and cost of a transformer for a given power rating is inversely related to the operating frequency, and switching converters can operate at a frequency a thousand times the mains frequency or more.
 

Thread Starter

Spottymaldoon

Joined Dec 4, 2015
88
Thanks, Crutschow, for this eminently reasonable explanation! Makes total sense now and explains why I have seen the same approach used in many other modern designs.
 

Ian0

Joined Aug 7, 2020
5,504
I’ll just add to @crutschow ’s excellent explanation that voltage and current regulation can be performed at the same time, as part of the power conversion stage, so you actually get a better battery charger that will charge your battery and won’t destroy it.
Very few of the old fashioned designs had any control at all, and relied on the user disconnecting them before they killed the battery.
 

crutschow

Joined Mar 14, 2008
30,102
Very few of the old fashioned designs had any control at all, and relied on the user disconnecting them before they killed the battery.
Yes, the old Sears one I have has an output of about 15Vdc and uses the transformer winding resistance to limit the current to <6A.
Works fine as long as I don't leave it connected to the battery after it's charged.
 

MisterBill2

Joined Jan 23, 2018
12,399
Certainly the "inverter" arrangement in the charger is less expensive to produce, and much lighter for shipping. it also allows a smaller package. While it is certainly possible for the circuit to regulate the voltage closely and do a better job charging, unless such is claimed it is probably not present. It is easy enough to make a poorly regulated inverter supply, and it costs less as well.
and those cheap inverters are subject to failure with repair parts being unavailable.
 

Thread Starter

Spottymaldoon

Joined Dec 4, 2015
88
Thanks to everyone for responding - it's appreciated.
Unfortunately I don't have a schematic so I'm largely having to guess how to fix it. A 170V DC is produced at the back end using expensive looking RFI suppression toroids and a big electrolytic and there's a power FET which is evidently to do the switching but there's no oscillation so, unless this is switched on from the front end, I am left with changing out the FET for now.
Just in case the circuit looks familiar to somebody I'll attach a picture.Battery charger.jpg
 

Ian0

Joined Aug 7, 2020
5,504
If the FET has failed it will be short-circuit, and the fuse will have blown.
Check the secondary rectifier diode. If that fails and goes short, the circuit stays perpetually in current-limit.
 

MisterBill2

Joined Jan 23, 2018
12,399
The first item to check is the fuse. Then check and verify that the capacitor is charging to 170 volts. Then verify that there is a voltage across the FET source and drain. After that it is an issue of checking components if you do not have a circuit to follow. If you have a meter that has a diode check function that will be very useful. But also, maybe even first, a careful examination for broken connections. That includes the charging cable clips. Not a likely suspect, but a possible guilty item. and is a replacement PCB assembly available?
 

Dodgydave

Joined Jun 22, 2012
10,361
That's a switchmode psu which uses a chip to pulse the transformer and produces a lower voltage output, and regulated by the opto-couplers.
What is the number on the chip?

IMG_20220702_214811.jpg
 

MisterBill2

Joined Jan 23, 2018
12,399
I see four small signal transistors, one power tab transistor, several diodes several capacitors, and many resistors. Also a ribbon cable connecting to something.With no circuit available a logical analysis would be difficult, and so component checking is an alternative. After inspecting all of the connections the next step would be to check the transistors for shorted and open conditions. Then check the capacitors for short circuits, loss of capacitance, and breakdown.
also, it will be worth while to see if there is still any warranty time, or if replacement circuit boards are available.
And a view of the solder side of the PCB can be useful because some failures are caused by failed solder connections.
 
Last edited:

Dodgydave

Joined Jun 22, 2012
10,361
This is the datasheet test circuit diagram, it may be the same on the primary side...

Check the supply at pin 8, and check the fet for shorts, as a temporary measure you can open the diodes on the secondary side to remove the load.
3147cd00002330_1.jpg
 
Last edited:

MisterBill2

Joined Jan 23, 2018
12,399
I do see what might be some suspect solder connections, but t may also be the illumination playing tricks.
I also realize that we did not get a detailed description of the complaint. So that may be useful. Knowing the symptomes often aids tin the diagnosis.
 

R!f@@

Joined Apr 2, 2009
9,801
No need to take out any components yet.
My approach is :
1. Discharge all caps.
2. Do short circuit test on FET, Secondary Diodes and All electrolytics. While soldered.
From than onwards I deep dive
 

Thread Starter

Spottymaldoon

Joined Dec 4, 2015
88
There is 170V DC across the primary smoothing capacitor but the FET isn't switching this across the transformer. The secondary half is without power. FET and secondary diode test OK. Pin 8 Vcc on L6565 is at 12V. I don't see any bad connections although several have been reworked after manufacture. I could systematically try checking all components but, for an $80 item it really isn't worth my time - or yours!

I got excellent replies which taught me why these new fangled methods are used - thanks to all.

Has anybody got a recommendation for a good reliable Walmart battery charger?!
 
You expect something "good" from Walmart? These car battery chargers have, in the race to the bottom for price, just terrible quality out of china. The electrolytic capacitors have a short life. Jay Leno's Garage endorsing CTEK which are notorious for failed capacitors. Schumacher are another old brand that is getting chargers OEM'd for them out of china and they either have buggy software or unreliable hardware. I tried Energizer brand but it was too finicky about initiating charging. Like a battery at 8V it refused to charge. Otherwise the build quality was good.

So car battery chargers are not that great for quality and reliability nowadays. The best brands and technology are still getting contract manufactured in china so the build quality/reliability is quite poor despite the Kardashian name brand.

The L6565 will not start up until over ~13V on it's power pin 8. Either the small cap nearby has gone bad, worth pulling it and testing/replacing it. It's a common failure. Or the startup resistor to that pin has gone high value etc. and a few other parts to check.
Not sure how many amps this charger is rated, but I can see low cost small capacitors 470uF which will get quite hot even pulling 2-3A they are undersized.
 
Top