Maximum reverse voltage of AZ34063A step-down from 24V solar PV to 13V battery ?

[Andy Chee]

Protecting a battery from discharging has nothing to do with the topic 

If battery discharge has nothing to do with the topic, then go ahead and remove the solar panel diode as well.

Yes, your answers resemble AI prompts, while the topic of this thread is reverse voltage protection of the MC34063A family ICs. 

How rude of you to make such unfounded accusations!

Anyway here's my reverse voltage experiment.

As you can see, I have a whole bunch of 34063's to choose from



And this is my breadboard setup



The test schematic is straightforward.

pin 1 - GND
pin 2 - +V
pin 3 - open
pin 4 - GND
pin 5 - +V
pin 6 - GND
pin 7 - GND
pin 8 - GND

I have included a LED and 430 ohm resistor in series with +V, to act as a current limiter and visual indication.  In addition, the bench supply has the current limit set to 200mA.

So starting with the power supply at minimum, I slowly increased voltage.

The LED started glowing faintly at 2.2V.  Increasing the voltage up to 40V, the LED was bright.

I repeated the test with different chips with the same results.

So I guess that's conclusive.... ensure your 34063 circuit is protected against reverse voltage!

[eneuro]

Why wasn't there any mention that the solar panel is 150W while a 34063A might realistically do 15W at most?

I have not only a car but also a motorbike in my garage, and I am also working on a 2A version, because the commercial charger did not work at all with such a current today, the voltage on the motorbike battery suddenly jumped to over 15V with the panel facing the sun, the available current was slightly over 4A when I tested it by directly connecting the PV to the battery without this system 
However, this simple dc-dc based on the mentioned AZ3406A worked great, at the beginning it was almost 500mA, then at a voltage of 12.7V on the battery the current dropped to 300mA, probably because I used the original cable from this converter 12v-14v to 5v@500ma, but I have MC3406AB systems from ST, better 100uH inductors and the current will be around 0.666A at a frequency of about 69Hz, because in this Chinese converter a Ct of probably 1nF was used 

[eneuro]

I noticed that the input power source for MC34063A is a voltage source. Let us recall that a solar panel is a current source.

Even this noname AZ34063A based circuit converted from 5V@500mA to 13V@500mA works better than the purchased commercial charger set to charge a 12V motorbike battery, because it could not maintain a voltage of 13V, and the specification of this motorbike includes a 13V alternator, so it looks much better when using a 150W 24Voc -> 13V PV panel as a power source than what is available in stores.
The PV panel has a large potential, the maximum short circuit current is 5A, as I connected it today for testing without any regulators at 1:00 p.m. in Europe with a beautiful sun, cloudless sky, even set at a certain angle, it gave a charging current of around 4A.
It is always possible to connect several 500mA dc-dc in parallel, or design your own 2Amax dc-dc converter, because this is enough to charge small motorcycle batteries without removing 

[eneuro]

And this is my breadboard setup

In my free time I will connect the VEC of the BJT NPN BC547 transistor to a 12V car battery, let's see what happens, because breadboard tests are good for YT influencers, it's not a real circuit 

[NiHaoMike]

What about have a single MPPT controller charging a 12V lead acid or LiFePO4 battery, then have the smaller 12V batteries connected in parallel with PTCs to limit the current?

[eneuro]

So starting with the power supply at minimum, I slowly increased voltage.

I repeated the test with different chips with the same results.

How long was the MC34063A family IC exposed to this reverse power supply >=13V in your tests ? 
This is a typical example of tests being done without understanding the subject 
Leave the reverse power supply connected MC34063A not for a moment, but for 12 hours, 30 days a week, 365 days a year, then it will make sense 
These cheap systems as DC-DC 12V-24V to 5V can work great precisely because the output voltage is 5V, not 13V which is what I am interested in, so I soldered a bypass diode and that closes the topic. 

[Andy Chee]

So starting with the power supply at minimum, I slowly increased voltage.

I repeated the test with different chips with the same results.

How long was the MC34063A family IC exposed to this reverse power supply >=13V in your tests ? 
This is a typical example of tests being done without understanding the subject 
Leave the reverse power supply connected MC34063A not for a moment, but for 12 hours, 30 days a week, 365 days a year, then it will make sense 

It broke instantly as soon as power was applied, so what's the point of leaving it for longer?

Maybe you didn't understand how the test works?

If the chip could withstand reverse voltage, then the LED should have stayed off from 0V up to 40V.  But the LED turned on immediately with 2V.  Therefore the chip CANNOT withstand reverse voltage.

And in case you still don't understand, it failed reverse voltage immediately!

[eneuro]

I have included a LED and 430 ohm resistor in series with +V, to act as a current limiter and visual indication.  In addition, the bench supply has the current limit set to 200mA.

So starting with the power supply at minimum, I slowly increased voltage.

The LED started glowing faintly at 2.2V.  Increasing the voltage up to 40V, the LED was bright.

I repeated the test with different chips with the same results.

The topic is not the reverse voltage of the LED supply, but the IC MC34063A 

The LED can be easily protected by soldering a second, reversely polarized diode in parallel.
If the LED diode fails, it doesn't matter in my case, because of course I added two 4.7k resistors to 560R which limited the current in the case of the original circuit at 5Vout, so now I have a green 3mm LED diode and 2x4.7k+560R, i.e. at a voltage of 13Vout, a current of about 1mA flows and there is no question of any reverse power supply, because the charged battery has the same polarity at the output. 

[Manul]

What the hell is the argument about? Andy clearly wrote that MC34063 sinks current into it's output while unpowered.

[Andy Chee]

I have included a LED and 430 ohm resistor in series with +V, to act as a current limiter and visual indication.  In addition, the bench supply has the current limit set to 200mA.

So starting with the power supply at minimum, I slowly increased voltage.

The LED started glowing faintly at 2.2V.  Increasing the voltage up to 40V, the LED was bright.

I repeated the test with different chips with the same results.

The topic is not the reverse voltage of the LED supply, but the IC MC34063A 

Clearly you still don't understand the test.

Here is the test setup:



If the LED lights up, that means the 34063A failed reverse voltage.

I don't know how else to explain it to you.

[Manul]

Andy, do you still have that setup? Out of curiosity, have you tried pin 2 and 5 separately? Does pin 2 sinks too?

[Andy Chee]

Just tested again with pin 5 connected to GND instead. 

Same result, the LED lights up.

[KerimF]

But the LED turned on immediately with 2V

This is really weird! It seems that the breakdown voltage of Q2 EB junction is close to zero unless a resistor, on Q2 BE, was omitted on the schematic.

[magic]

Do you mean that the hand-drawn schematic above lights up the LED with 2V instead of 40V?

I suppose it might be the 100Ω resistor, if it's made of a P diffusion into N silicon biased to Vcc. Basically, it's a diode to VCC in such case.
I have that Zeptobars image somewhere but I'm too lazy now to look at it to confirm if that's the case.

edit
There is also BC junction of Q1 in series with 100Ω, which you externally short with VCC.

[Andy Chee]

Do you mean that the hand-drawn schematic above lights up the LED with 2V instead of 40V?

With the schematic shown above, I slowly increased voltage from 0V.  The LED began slightly glowing at 2.2V and only got brighter from there, until I maxxed out at 40V.

And yes, I was also interested in examining the die, to suss out what's going on.  But I'm pretty much a n00b when it comes to reverse engineering dies.

https://zeptobars.com/en/read/MC34063

edit
There is also BC junction of Q1 in series with 100Ω, which you externally short with VCC.

you're onto something there! the BC junction is just a diode PN junction which allows current to flow like this:

[KerimF]

Let us not forget the forward voltage of an LED. Even for a very small current it is not less than 1 V.

[magic]

This makes sense. And Q1 may even operate in reverse, so only some fraction of LED current has to pass through 100Ω.

So I suppose small currents should be relatively harmless to this chip and one might get away without a reverse diode.

[eneuro]

Clearly you still don't understand the test.
...
I don't know how else to explain it to you.

Create your own thread for these tests, because they are useless to me, I've 13vout and IC MC34063A protected with diode FR107G 1000V 1A 

[eneuro]

So I suppose small currents should be relatively harmless to this chip and one might get away without a reverse diode.

Not necessarily, because it may be a matter of time before the reversely connected power supply damages the emitter of the NPN transistor even though a small current flows through it.

These tests are pointless because they do not test a real circuit I have, secured with this cheap fast diode, it charged my motorcycle battery to 12.9V connected to a PV panel, so the only thing I am doing now is my own version with a larger 100uH inductor for a slightly higher current operating at higher frequencies around 70kHz. 

[eneuro]

But the LED turned on immediately with 2V

This is really weird!

Yes, absolute nonsense, there is no information about the color of this LED to estimate its Vf and current limiting resistance values in series, in addition, instead of a resistor ladder at the output, the output is connected to GND 

Keep things simply as shown below 

[Andy Chee]

But the LED turned on immediately with 2V

This is really weird!

Yes, absolute nonsense, there is no information about the color of this LED to estimate its Vf and current limiting resistance values in series, in addition, instead of a resistor ladder at the output, the output is connected to GND 

Keep things simply as shown below 

The colour of the LED is shown in the breadboard photo 

The size of the limit resistor was specified in the post with the breadboard photo 

Clearly you still fail to understand the purpose of the test setup, to investigate maximum reverse voltage of the 34063A.

Yes, your diode works to protect the chip, but you still have not answered the question.... what is the maximum reverse voltage?

[eneuro]

I have included a LED and 430 ohm resistor in series with +V, to act as a current limiter and visual indication.  In addition, the bench supply has the current limit set to 200mA.

I have successfully tested the MC34063A protected by a diode by charging a car battery VARTA Silver Dynamic 12V 63Ah 610A with a 150W solar panel, where the short circuit current is limited only by battery internal resistance, it will definitely exceed 1000A 

Yes, your diode works to protect the chip, but you still have not answered the question.... what is the maximum reverse voltage?

This does not matter, because the IC MC34063A has been tested in real working conditions at a battery voltage of 13Vmax, even if I start a car or motorcycle during charging, nothing will happen because the car alternator can raise this voltage to 14.2Vmax or the documentation for the motorcycle alternator says 13V, that's why I charge to 13Vmax with this circuit. 

[Andy Chee]

I have successfully tested the MC34063A protected by a diode by charging a car battery VARTA Silver Dynamic 12V 63Ah 610A with a 150W solar panel, where the short circuit current is limited only by battery internal resistance, it will definitely exceed 1000A 

I can also test reverse voltage with a car battery and NO PROTECTION DIODE, however my LED test predicts what will happen with the car battery.... fried 34063!

I can make a smoke video, but I am not a YouTuber.  I do not make videos of things blowing up.

[KerimF]

This may be off topic.
I guess the circuit of interest here (which I am not sure of) uses an inductance after the output emitter (or source, if MOSFET). And the input collector (or drain) is connected to the positive output of the solar panel.

If this is the case, when the output transistor (or MOSFET) is off (while driven by PWM), it cuts the current of the solar panel. Naturally, this decreases the efficiency of the solar cells. In this application, the efficiency (speaking relatively) decreases from 100% (using the right topology) to 13V/24V (close to).

This may not be important here. But it has to be considered when designing a step-down converter in case one likes to get all energy provided by his PV. This is achieved by allowing the PV continuous/available current to enter the charger all the time.

[eneuro]

If this is the case, when the output transistor (or MOSFET) is off (while driven by PWM), it cuts the current of the solar panel. Naturally, this decreases the efficiency of the solar cells.

I'd like to have 2A charging current, so I found an easy solution, connect three MC34063A circuits in parallel designed for 0.666A*3~1.998A current  The probability that at the same time none of these systems will draw current from the PV panel is very small, because the input capacitors will be charged even if one of them blocks the current, and the switching frequency will never be identical, because the CT IC MC34063A capacitor will never have the same capacity in each circuit.