PWM Circuit not working at all - Need Help

I am using a TL5001 to step down ~20V to 5V, datasheet here: https://www.electro-tech-online.com/custompdfs/2006/06/tl5001a.pdf

Here is the circuit I am using:

**broken link removed**

For some reason, this circuit just plain doesn't work at all. And I mean no part of it works. I was trying it on a breadboard and it wouldnt work, so I actually bought a general pc board and soldered all the components on and I get the same results. Nothing works.

My output switching frequency should be around 100khz and should look like a triangle wave that should have a min at 0.7 and a max at 1.3, but instead the output looks like this:

**broken link removed**

When I measure the output voltage (5V) I just get 20V (same as the input voltage). I have no idea what I am doing wrong and could use some help. I have worked with other PWM controllers and I have at least be able to get the switch output to work fine (even on a breadboard), nothing is working with this IC.

Also, I ordered 3 TL5001's and they all work the same way, so I doubt I got a defective copy.

The output of the IC is an open collector. Your circuit needs a resistor from gate to source of the Mosfet to turn it off. The value of the resistor must be calculated to discharge the high capacitance of the Mosfet quickly for the Mosfet to remain cool.

audioguru said:

The output of the IC is an open collector. Your circuit needs a resistor from gate to source of the Mosfet to turn it off. The value of the resistor must be calculated to discharge the high capacitance of the Mosfet quickly for the Mosfet to remain cool.

Click to expand...

I tried putting the 470 ohm resistor from the gate to the source but when I do that, the value of the gate voltage is just a constant 20V (the input voltage). As you noticed, the output isnt even ~100khz like its suppose to be. Something else is wrong here.

What's the voltage at Vfb? it should be between 0 t0 1.5V

Paul Obrien said:

What's the voltage at Vfb?

Click to expand...

0.986V after about 1 second

It starts off very erradic, then settles at the 0.986V. This should be the right value since I put the feedback network voltage divider to be the same as the 1V reference voltage for a 5V output.

470 ohms at the Mosfet is with a 5V input in the datasheet. With 20V input then the output transistor in the IC will fry with the high current. Use 2.2k.

Your input voltage is 20V and so is the absolute max gate voltage of the Mosfet. Does the Mosfet still work? (Is it able to be shut off?)

Your voltage at FB is correct if the output voltage is 4.99V for your resistor values so the output voltage shouldn't be too high. Check the values of those divider resistors.

A capacitor in parallel with R8 may solve the problem. I think the 180uF filter cap looks like a short circuit at turn on causing the short circuit protection to latch the power mosfet off. A capacitor from pin 6 to gnd will give it soft start.

I was just thinking right now, how in the hell can Vfb be 0.986V if my circuit isnt working and is 20V at the output.

I just checked now it is working, my output is 4.88V now. Weird how it works now, I didnt change anything.

Anyway, do you know what the proper calculation is for the resistor between the gate and the source? It's not listed in the datasheet and I'm just using the 470 ohm resistor they used.

FusionITR said:

Anyway, do you know what the proper calculation is for the resistor between the gate and the source? It's not listed in the datasheet and I'm just using the 470 ohm resistor they used.

Click to expand...

The absolute max output current is 21mA and 100mA peak.
Their 470 ohms with a 5V input gives a current of about 10mA.
2.2k with your 20V input is a current of 9mA.

Ok I changed the gate to souce resistor, and added at 0.1uF softstar capacitor, and now my output is about 5.04V.

Problem is, my output isnt stable at all, it is fluxating between 4.99 and 5.15V, so im guessing either my feedback compensation is crappy. Do you guys know any guides out there that explains feedback compensation? My book does a crappy job of explaining it.

Heres how unstable the voltage is. This is the 5V output when ac coupled and viewed over a 250ms period:

**broken link removed**

Is that with any load? I use the National Simple Switchers and the output is usually very good, maybe a load is needed.

What is the waveform at C1? Also, try removing all the compensation. If C2 is large enuf, it generally will be stable.