Review of LM2675-5 and MOSFET switch

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
Before I go an order the parts from digikey, I thought maybe someone could go over the design I came up with.

For the LM2675 part, I used the WBench app from TI
http://www.ti.com/product/lm2675/description?keyMatch=lm2675&tisearch=Search-EN-Everything

My supply will be 12V. I want 5V out at 60ma. O f particular concern is the coil choice of 1uh. Does that seem OK?


Also there is the switch. It is intended to switch 12V @ ~1A. It will be switched by a mcu.

upload_2015-12-18_15-0-41.png
 

OBW0549

Joined Mar 2, 2015
3,566
My supply will be 12V. I want 5V out at 60ma. Of particular concern is the coil choice of 1uh. Does that seem OK?
1 μH is WAYYYY too small. Take a look at Fig. 25 in the datasheet; according to that, for 5.0V out at 60 mA, you should be using 100 μH at least.

Also, the output capacitor is too small. From Table 3, it looks like for a 100 μH inductor, you should be using about a 100 μF capacitor.

Also there is the switch. It is intended to switch 12V @ ~1A. It will be switched by a mcu.
Ummm... something's fishy about that switch, and how you've connected its load (is that across the LED?) As you have it, that switch isn't going to do anything at all.

Hope this helps a bit...
 

crutschow

Joined Mar 14, 2008
34,280
1μH seems too small.
What is the PWM switching frequency?
A rough check of the minimum inductance (for not discontinuous operation) is L = V*t/i where L is the inductance, V is the difference between the input and output voltage (7V for your design), t is the ON time of the switching transistor during normal operation (from the PWM signal), and i is the minimum output current.
 

OBW0549

Joined Mar 2, 2015
3,566
But look CAREFULLY at how you've got it connected. The (-) end of the LED is connected to ground. The (+) end is connected to the drain of Q1. The source of Q1 is connected to ground. When Q1 turns on via a positive voltage on the CTRL pin, the only thing that happens is BOTH ends of the LED become grounded. Therefore nothing happens.
 

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
1μH seems too small.
What is the PWM switching frequency?
A rough check of the minimum inductance (for not discontinuous operation) is L = V*t/i where L is the inductance, V is the difference between the input and output voltage (7V for your design), t is the ON time of the switching transistor during normal operation (from the PWM signal), and i is the minimum output current.
1 μH is WAYYYY too small. Take a look at Fig. 25 in the datasheet; according to that, for 5.0V out at 60 mA, you should be using 100 μH at least.
.
What I thought. But my mistake on the schematic. Wizard actually calls for 1mh not uh which seems way to high.

I did look at Figure 25 but it only goes down to .2Amps from what I see. I need .06amps. Do I just use the minimum?

upload_2015-12-18_16-11-47.png

Really glad I posted this review here. Thanks for the help so far.
 

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
But look CAREFULLY at how you've got it connected. The (-) end of the LED is connected to ground. The (+) end is connected to the drain of Q1. The source of Q1 is connected to ground. When Q1 turns on via a positive voltage on the CTRL pin, the only thing that happens is BOTH ends of the LED become grounded. Therefore nothing happens.

Duh! Yep! Again glad I posted! Thanks! what I should have done was drawn the LED in first and it would have been obvious.
 

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
Table 1 is actually a little more complete. They call for 68uh for .32 amps if I am reading the chart correctly.

upload_2015-12-18_16-24-52.png
 

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
Corrected design below. I changed out the coil and ran it through the wizard. It increased efficiency to +92% from around an 87%. I suspect the wizard is not setup for such low currents.

The idea of this whole thing is to replace the buckpuck in the photo. In addition to driving LEDs, the buckpuck also has 5V out to run a mcu and has a control pin to turn on and off the light. The old light had no internal current regulation. The new light has it's own current regulation so I want to make a plug compatible board that will provide 5 VDC and allow me to switch the light.





upload_2015-12-18_16-37-48.png
 

OBW0549

Joined Mar 2, 2015
3,566
Wizard actually calls for 1mh not uh which seems way to high.
I agree; 1 mH does seem high to me...

I did look at Figure 25 but it only goes down to .2Amps from what I see. I need .06amps. Do I just use the minimum?
I'd just use a straightedge and extrapolate the table on downward to the left. Just eyeballing it, it looks like 100, 150 or 220 μH would do the trick.
 

OBW0549

Joined Mar 2, 2015
3,566
Table 1 is actually a little more complete. They call for 68uh for .32 amps if I am reading the chart correctly.
I think that 0.32 Amp figure is the max. current rating of that particular inductor, not the current level for which that inductance value is recommended.
 
Last edited:

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
Looking good. Only thing I noticed (this time) was the input capacitor C1; you have 4.7 μF, and the circuits I see in Figs. 22 and 23 of the datasheet show 22 μF. Is 4.7 μF big enough, I wonder.
Wizard suggested that. Rest of the numbers seemed ok at .5 amp so figured the input cap was good. I will change it out as per the datasheet.

Thanks for the help everyone. Really appreciate it.
 

Picbuster

Joined Dec 2, 2013
1,047
Please start reading datasheet it will produce all information on inductor cap's and diodes.
I do not understand the conversation while all is embedded in datasheet.
Do not get me wrong I do not want to be rude but try to lean people to study information profided by producer first.
 

ronv

Joined Nov 12, 2008
3,770
Wizard suggested that. Rest of the numbers seemed ok at .5 amp so figured the input cap was good. I will change it out as per the datasheet.

Thanks for the help everyone. Really appreciate it.
I'm confused. (Not unusual:D) The buck puck is 700 ma. Is this for your 108 led project?
If so you need to configure the regulator for constant current and much higher.
 

Thread Starter

spinnaker

Joined Oct 29, 2009
7,830
I'm confused. (Not unusual:D) The buck puck is 700 ma. Is this for your 108 led project?
If so you need to configure the regulator for constant current and much higher.

The 108 LEDs are going away. The new light will have it's own constant current source. I did the math way back when and it came to 64oma for that light. It was plenty bright back in the day. The controller it replaced was set way, way lower.
 
Top