Pulse Current Limiting With The SG3525A / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Pulse Current Limiting With The SG3525A

1 2 next

Move Thread

LAN_403

jpsmith123

Fri Apr 17 2009, 05:10AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Hello,

Does anyone have any examples of the use of the "shutdown" pin (pin 10) of the SG3525A switching regulator, for current limiting purposes?

On all the various manufacturers' data sheets I've seen, this feature is hardly mentioned/poorly characterized, and I can't find any application notes that shed much light on it either.

I suppose I could build a prototype and experiment with it, but I'd like to narrow it down a little first if possible...e.g., with regard to pin 10, what is the highest voltage for which there will be no shutdown, and what is the lowest voltage for which shutdown is guaranteed (this is the type of info I was expecting to see on the data sheet).

Regards,
jpsmith123

Steve Conner

Fri Apr 17 2009, 10:19AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Murphy's Law of datasheets: If you don't see a parameter specified on the datasheet, it will be worse than you can possibly imagine. It was left off the datasheet because it's bad enough that the marketing department want to hide it.

In this case, the threshold is specified between 0.6 and 1V on page 4 of the SGS-Thomson datasheet, and that means it probably varies over that range with chip temperature. The pin is just a transistor base. So if you want accurate peak current control, it would be better to use an external comparator.

Other PWM ICs are available that have proper peak current control, and an onboard comparator suitable for direct connection to a current sense resistor.

jpsmith123

Fri Apr 17 2009, 02:13PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Thanks for the info...that data sheet is the only one I've seen so far that puts any kind of number (semi-useless though it may be) on the shutdown pin.

I had assumed that it would be in the vicinity of 0.7 V, based on the block diagram, but many of the data sheets' test circuits show the application of the 5V reference voltage to the pin...which doesn't instill confidence.

Anyway, what I'm doing is building something similar to Uzzors' "multipurpose inverter", but I wanted to add some kind of pulse current limiting to protect the MOSFETs, so perhaps the SG3525A with its shutdown pin, crude as it is, may still be useful for my purposes.

GeordieBoy

Wed Apr 22 2009, 11:16PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Try looking at some of the Unitrode parts like UC3525A and more recent voltage-mode parts. (The SG3525A is really a cheap and not-so-cheerful alternative.)

Often the "shutdown pin" serves a number of functions on proper SMPS controller ICs. Firstly, when the voltage here momentarily exceeds some threshold it will terminate individual pulses on a pulse-by-pulse basis to implement peak-current limiting. Secondly, it will often have another slightly higher threshold which, if reached, will cause several cycles to be skipped, or it will cause the soft-start capacitor to be discharged invoking a new soft start cycle.

The first cycle-by-cycle disable is good for limiting peak-current in the switches, provided that the "shutdown" delay is sufficiently short. The propagation time from this pin going high to gate drive being removed can often cause problems particularly in high frequency designs that use a small choke output choke where current rises quickly.

The second higher shutdown threshold is useful for keeping average current under control when the output of the SMPS has a bolted short. The minimum duty-ratio afforded by the delay time mentioned above is often too long to keep the output current safely bounded, and this additional shutdown mechanism is very useful for preventing output current skyrocketing out of control.

I hope this helps,

-Richie,

jpsmith123

Thu Apr 23 2009, 06:41PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Hello Richie,

Thanks for the reply. In looking at the data sheets, however, I don't see much difference between the UC3525a and the SG3525a (except that the UC3525a data sheet at least has a little info about the shutdown pin).

Regards,
jpsmith123

Steve Conner

Fri Apr 24 2009, 07:44AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

He probably meant UC3526. The UC3525 is the same as the SG3525.

GeordieBoy

Fri Apr 24 2009, 11:10AM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

The UC3525A is what I would call the first "proper SMPS" controller in that it includes a proper fast shutdown pin to terminate gate drive on a cycle-by-cycle basis (although it isn't particularly fast by today's standards). It also incorporates proper gate-drive outputs capable of sourcing and sinking some real current. (Again, not much current by today's standards, and most external gate-drive chips will do much better.)

I'm not familiar with SG3525. If things are left unspecified on the datasheet then it is likely a copy-cat of the Unitrode part with possibly inferior spec.

You should still be able to get the UC3525A and that might satisfy your requirements. If it doesn't, then there are a whole range of more modern voltage-mode controller chips in the UC and UCC family that will surely do what you want.

-Richie,

jpsmith123

Fri Apr 24 2009, 05:02PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

The UC3525a (spec here:

is apparently identical to the SG3525a
(spec here:

I was a little confused at first, because all I found were data sheets like the one from ON Semiconductor, which did not characterize the functioning of the shutdown pin; this is in contrast to the Unitrode (now TI) data sheet which has a "SHUTDOWN" section with some numbers (for voltage threshold and delay time).

I agree with you that by today's standards, this chip is nothing to write home about, but for my intended use, with switching frequency somewhere between 30 and 50 khz, and MOSFETs with some spare capacity, I think this chip will suffice.

Dr. Dark Current

Sat Apr 25 2009, 09:33AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Somewhat on topic, could anyone tell me if the 3525's error amp input common mode voltage includes ground?

The datasheet doesnt say anything about this.

Steve Conner

Sat Apr 25 2009, 10:12AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

If the datasheet doesn't say anything, that means it will work on all your prototypes in the lab, but fail in production, bankrupting you.

Looking at the TI datasheet posted by Richie, on page 4 we see that the common mode rejection is specified over the range 1.5 to 5.2V. To me, that suggests that the common mode range doesn't include ground.

On page 5, there is a schematic of the error amp front end, and again, you can see NPN transistors, which implies it won't work at ground. The transistors require at least 0.7V to work, and the 200uA current source will need about the same, which ties in with the 1.5V we saw above.

The initial letters of a chip part number identify the maker, so all of these xx3525 should be equivalent to each other. Many popular chips are "second-sourced" like this. The original examples are the 741 and 555, so widely copied that they lost their maker's letters altogether. The TL494 and UC3525 are on the way there, second-sourced by the likes of Daewoo, Samsung and NEC.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob