I made a tiny 3W mono amplifier using the Texas Instruments TPA6211. As the maximum voltage is a little bit above 5 Volts, I implemented a buck regulator using the MAX5035 (see link at the end) . This is an SOIC 8 pin package . I made it on a perforated prototyping board so the chip is over Kapton tape and I just wired each of the leads.
My available input voltage is 25 VDC , when I connected it, the Vout of the MAX5035 is 5.03 Volts. To my surprise, when I loaded the circuit with a 50 Ohms resistor to draw about 100 mA the output drop down to 1Volt. If I remove the load , it comes back to 5V. The Datasheet says it should handle 1 Ampere. The waveform at the LX pin ( which goes to the inductor) is exactly as shown on the datasheet with no load However, the loaded waveforms in the datasheet show the pulses and in my case Vout just drops flat to 1 V or so with no oscillation at all.
I've replaced the MAX5035, I measured the capacitors, I've also tried a bigger inductor as 220 and 470 uH with just obtaining a little higher output like 1.5 V and about 2 V when loaded.
The assembly in the protoboard is the IC in the center and the inductor and freewheeling diode on the same side of the IC lead everything is in a 1 x 2 inches little protoboard.
I've also moved the wire connecting the LX pin on different positions to avoid it making a a feedback loop with other connections.
The resistors for the ON/OFF pin are as recommended for the input voltage range.
Does anyone had this experience or have an idea what is going on?? For sure, the pulses do not start , but what could be so wrong on the layout for so few connections?? What could be the reason the regulator can not start oscillating ? ( in spite it seems to be oscillating with no load)
It if is pulsing at an extremely low duty cicle at open circuit; why the pulsing vanish when loaded with just 10% of the maximum output current?
I've Googled for MAX5035 bad, no work, layout, problem , start with ZERO hits, just the datasheet and projects using it.
I've lost a whole afternoon trying to get this working to no avail. I do not dare to make a PCB as I believe I may have 50% chances of getting it working. I have ordered equivalent IC from Texas instruments and I will test them in exact same layout. If they work, then good bye Maxim and will not loose more time with them.
Thanks a million in advance , I perhaps had posted too many questions but I think one answer fits all. I am curious on whether I have two bad chips or am I ignoring something very important? In fact, the only thing mentioned on the datasheet is the need of a "ground plane" but I believe this is related to avoid RF interference and comply with regulations than to allow the circuit to start the oscillation, yet I could be wrong.
The schematic is in this link.
MAX5035 1A, 76V, High-Efficiency MAXPower Step-Down DC-DC Converter - Overview
PS :When I used the word "oscillating" above I meant (this is an SMPS) that there is a variable duty cycle waveform and changing the duty cycle with the different loads.
My available input voltage is 25 VDC , when I connected it, the Vout of the MAX5035 is 5.03 Volts. To my surprise, when I loaded the circuit with a 50 Ohms resistor to draw about 100 mA the output drop down to 1Volt. If I remove the load , it comes back to 5V. The Datasheet says it should handle 1 Ampere. The waveform at the LX pin ( which goes to the inductor) is exactly as shown on the datasheet with no load However, the loaded waveforms in the datasheet show the pulses and in my case Vout just drops flat to 1 V or so with no oscillation at all.
I've replaced the MAX5035, I measured the capacitors, I've also tried a bigger inductor as 220 and 470 uH with just obtaining a little higher output like 1.5 V and about 2 V when loaded.
The assembly in the protoboard is the IC in the center and the inductor and freewheeling diode on the same side of the IC lead everything is in a 1 x 2 inches little protoboard.
I've also moved the wire connecting the LX pin on different positions to avoid it making a a feedback loop with other connections.
The resistors for the ON/OFF pin are as recommended for the input voltage range.
Does anyone had this experience or have an idea what is going on?? For sure, the pulses do not start , but what could be so wrong on the layout for so few connections?? What could be the reason the regulator can not start oscillating ? ( in spite it seems to be oscillating with no load)
It if is pulsing at an extremely low duty cicle at open circuit; why the pulsing vanish when loaded with just 10% of the maximum output current?
I've Googled for MAX5035 bad, no work, layout, problem , start with ZERO hits, just the datasheet and projects using it.
I've lost a whole afternoon trying to get this working to no avail. I do not dare to make a PCB as I believe I may have 50% chances of getting it working. I have ordered equivalent IC from Texas instruments and I will test them in exact same layout. If they work, then good bye Maxim and will not loose more time with them.
Thanks a million in advance , I perhaps had posted too many questions but I think one answer fits all. I am curious on whether I have two bad chips or am I ignoring something very important? In fact, the only thing mentioned on the datasheet is the need of a "ground plane" but I believe this is related to avoid RF interference and comply with regulations than to allow the circuit to start the oscillation, yet I could be wrong.
The schematic is in this link.
MAX5035 1A, 76V, High-Efficiency MAXPower Step-Down DC-DC Converter - Overview
PS :When I used the word "oscillating" above I meant (this is an SMPS) that there is a variable duty cycle waveform and changing the duty cycle with the different loads.
The only things I can suggest are the inductor ! These can be very critical on SMPS designs. You are probably looking at something like this with a very low DC resistance and not a coil from for example a radio,
EPCOS|B82476A1104M|INDUCTOR, 100UH, 1200MA | Farnell United Kingdom
The diode also has to be of the correct fast recovery type.
The IC needs to be decoupled as close as possible to the package.
EPCOS|B82476A1104M|INDUCTOR, 100UH, 1200MA | Farnell United Kingdom
The diode also has to be of the correct fast recovery type.
The IC needs to be decoupled as close as possible to the package.
1. Most likely a bad connection somewhere. Test all your connections with a DMM (resistance) with the circuit disconnected.
2. Use a 100uf. Higher inductances will cause a higher voltage drop with higher current. Most Buck controllers are fixed frequency.
3. for 5V output R1 should be 1M and R2 should be 384K. Make sure they are the right values using your DMM.
4. measure your Input voltage when you apply a load to see if there is an issue with your input voltage supply.
2. Use a 100uf. Higher inductances will cause a higher voltage drop with higher current. Most Buck controllers are fixed frequency.
3. for 5V output R1 should be 1M and R2 should be 384K. Make sure they are the right values using your DMM.
4. measure your Input voltage when you apply a load to see if there is an issue with your input voltage supply.
Thanks you both, it took long time to come back to this that I barely found the prototype board. The problem was a simple issue. The Pin 5 which is the ON/OFF control requires a resistive divisor as per the datasheet must be 1M to Vin with 384K to GND. In normal conditions, this provides about 4.16 Volts with Vi = 15V.
For some reason, instead of 384K, I had a 125K resistor which ended providing 1.66 Volts to the ON/OFF pin. The result is my opening post issue, instead of shutting down completely, the converter keeps producing some output voltage and it goes down when loaded.
I've replaced the resistor and this IC works like a charm up to 1 Ampere.
The ON OFF pin should have a more powerful comparator so it either goes ON or OFF rather than giving a confusing intermediate state. Not sure if this is crazy or simple to accomplish.
Thanks again and as you pointed out , some connection had to be be wrong :=)
For some reason, instead of 384K, I had a 125K resistor which ended providing 1.66 Volts to the ON/OFF pin. The result is my opening post issue, instead of shutting down completely, the converter keeps producing some output voltage and it goes down when loaded.
I've replaced the resistor and this IC works like a charm up to 1 Ampere.
The ON OFF pin should have a more powerful comparator so it either goes ON or OFF rather than giving a confusing intermediate state. Not sure if this is crazy or simple to accomplish.
Thanks again and as you pointed out , some connection had to be be wrong :=)
- Status
- Not open for further replies.