[SOLVED] DC DC regulator: Voltage drop

Thread Starter

phoefler

Joined May 10, 2020
9
Hi,

I am currently working on developing a circuit that will control an external system using a PWM signal. The PWM signal will oscillate between -12V and 12V. To facilitate this, I have an external voltage supply rated at 12V with a maximum output of 1.5A.

I am utilizing an AZDelivery ESP-32 Dev Kit C V4 to generate the PWM signal. Additionally, I have a breakout board based on the MAX485 chip to convert UART to RS485.

My intention is to power both the ESP32 and the RS485 converter using the same voltage source. Both breakout boards can be powered with 5V. Therefore, I attempted to use a DC-DC converter to step down the 12V to 5V.

Here is the datasheet for the IC I am using: https://atta.szlcsc.com/upload/public/pdf/source/20160514/1463194849923.pdfSCR-20240427-moub.png

Please find attached the circuit diagram, which should be identical to the typical application circuit shown in the datasheet.

I have observed a peculiar behavior. When neither the MAX485 nor the ESP32 breakout boards are connected, I can measure 5V - which is perfect. However, when I attach one of these breakout boards, the voltage drops to approximately 1.7V.

I have tried changing the power supply and increasing the maximum current to 3A at 12V, but the behavior remains the same. It appears to be related to the drawn current. When I remove the breakout boards and replace them with a resistor, I observe similar behavior. With a high resistance (1MOhm), I measure a consistent 5V. With a 22k resistor, I measure 4.3V. And with a 10 Ohm resistor, I measure 0V.

As I am not an electronics expert, this situation is quite perplexing to me. I initially believed that a DC-DC converter would provide a stable voltage output, so this discrepancy is unexpected.

Any explanation is welcome :)
Thanks
 

crutschow

Joined Mar 14, 2008
34,855
Does the 12V remain stable when you load the output?

Is the inductor rated for 3A?

Double check all the wiring.

C14 and C15 should be ceramic type capacitors
 

Thread Starter

phoefler

Joined May 10, 2020
9
Thanks for your answer.

Yes, 12V remains stable, no matter which load/resistor is connected.

Yes, I think the inductor is rated for 3A.
C7 is this part: https://www.lcsc.com/product-detail/Power-Inductors_Sunltech-Tech-SLO1040H470MTT_C182153.html

C14 and C15 are this part: https://www.lcsc.com/product-detail...Electro-Mechanics-CL10A105KB8NNNC_C15849.html
I think this is a ceramic capacitor.

I double checked the wiring multiple times.
This is actually the 2nd prototype with the same behaviour. The first one used another DC DC converter.
I though, that it is a problem with the IC in the first place, that's why I switched to XL2012E1.

If I got you right, you do not see any obvious error, right?
Are there typical mistakes, that could lead to such behaviour?
 

MrChips

Joined Oct 2, 2009
31,106
Let's do some quick calculations using Ohm's Law.
With 4.3V across a 22kΩ resistor, the current is

I = V / R = 4.3V / 22kΩ = 0.2mA

Hence the supply is having difficulty supplying 0.2 mA at 5V.

There is a loss of 0.7V @ 0.2mA at the power supply.

R = V / I = 0.7V / 0.2mA = 3.5 kΩ

This indicates that the internal resistance of the power supply is 3.5 kΩ which is very high.

If you need to supply 100mA at 5V, this would represent a load of 50Ω. You want the internal resistance of the 5V power supply to be much lower than 5Ω.

Having said this, I have no idea what is wrong with the 5V power supply. A simple solution is to use an LM7805 linear regulator.
 

Thread Starter

phoefler

Joined May 10, 2020
9
Let's do some quick calculations using Ohm's Law.
With 4.3V across a 22kΩ resistor, the current is

I = V / R = 4.3V / 22kΩ = 0.2mA

Hence the supply is having difficulty supplying 0.2 mA at 5V.

There is a loss of 0.7V @ 0.2mA at the power supply.

R = V / I = 0.7V / 0.2mA = 3.5 kΩ

This indicates that the internal resistance of the power supply is 3.5 kΩ which is very high.

If you need to supply 100mA at 5V, this would represent a load of 50Ω. You want the internal resistance of the 5V power supply to be much lower than 5Ω.

Having said this, I have no idea what is wrong with the 5V power supply. A simple solution is to use an LM7805 linear regulator.
Thanks for your thoughts.
the first prototype used a linear regulator, as far as I can tell.
https://www.lcsc.com/product-detail...DO_UTC-Unisonic-Tech-78L05G-AB3-R_C71136.html

I had the exact same behavior. That probably allows the guess, that it is not directly related to the voltage regulation circuit, right?
(Though, I just noticed that the output current ist not very high)
 

ElectricSpidey

Joined Dec 2, 2017
2,889
If the 12-volt supply is capable of supplying the current, then I would suspect bad parts from that supplier.

Or perhaps a high resistance somewhere between the supply and the regulator.
 

Thread Starter

phoefler

Joined May 10, 2020
9
If the 12-volt supply is capable of suppling the current, then I would suspect bad parts from that supplier.

Or perhaps a high resistance somewhere between the supply and the regulator.
Yes I think so. I used two different power supplies.
the first one has a fixed max current rating at 1.5A, the second power supply is adjustable and I tested it with up to 3A.

I can’t believe that it’s a problem of a few bad parts, as I’ve seen the exact same behavior with a LDO in the predecessor prototype.
Even though I noticed a few mins ago that the LDO has only 100mA output current.
 

Thread Starter

phoefler

Joined May 10, 2020
9
Unfortunately, in the world we live in, bad parts are the first thing that comes to my mind.
I changed the part, but with the same result.
So, it seems that the part was ok and somethings different is going on ... ?

It really drives me nut, as I do not understand what's going on.
When attaching the ESP32 and the MAX485 the voltage drops to ~1,5V.
 

Thread Starter

phoefler

Joined May 10, 2020
9
Yes, that's fine.
Can you measure it to see if it's below an ohm?
Yes, certainly. However, I'm not entirely certain about the correct procedure. Could you please provide some guidance? I disconnected the 12V power source and attempted to measure the resistance using a multimeter. However, the reading was unexpectedly high, exceeding 1 megaohm. Then, I tried reconnecting the 12V power and repeated the measurement, which yielded a lower value but still far from the expected 145 milliohms. Is it possible that this discrepancy is the issue, or am I overlooking something in my approach?
 

crutschow

Joined Mar 14, 2008
34,855
Yes, certainly. However, I'm not entirely certain about the correct procedure. Could you please provide some guidance? I disconnected the 12V power source and attempted to measure the resistance using a multimeter. However, the reading was unexpectedly high, exceeding 1 megaohm. Then, I tried reconnecting the 12V power and repeated the measurement, which yielded a lower value but still far from the expected 145 milliohms. Is it possible that this discrepancy is the issue, or am I overlooking something in my approach?
Yes, you should be able to measure less than 1 ohm across the inductor with no power applied.
If you don't than something is wrong.
 

Thread Starter

phoefler

Joined May 10, 2020
9
Thanks for the explanation.
During measurement, I found the problem.
It seems, that the inductor had a bad solder joint. I noticed it when I gentle pressed with the measuring point on inductor.
I tried measuring without and with power and in the latter case it suddenly start working.

I re-soldered the inductor and it's now working as expected :)

Thank you all for your help!
 

BobTPH

Joined Jun 5, 2013
9,300
Never try to measure resistance on a powered circuit. The Ohmmeter is supplying a current if its own and it will interact with the powered circuit to give incorrect results if not worse, i.e. damage either the circuit or the meter.
 
Top