TRI1461, another tiny chip for building a DC-DC converter

Last month I reviewed the chip MP2359DJ for building a DC-DC converter, and in the comments people suggested a few other similar chips.
A little while later, I ordered a couple of different ones for my “assortment,” and recently I received and tested another one.

Yes, I know that reviewing a chip is funny—what is there to review, anyway?
But, for example, I was interested in finding out what else is available for sale besides the parts I use, since I can't keep up with everything. That's why comments on reviews like this can be very useful.

I hadn't planned to include photos of the packaging in this review, but in this case the seller managed to surprise me.
The chips arrived in a large cardboard box. When I picked up another batch of small packets at the post office, I didn't even realize at first what it was or who it was from.
When I opened the envelope and then the box, I was surprised. I've received much more expensive and fragile items in ordinary envelopes.


No bonus chip this time; I got exactly what I ordered: 10 ordered, 10 received. Too bad :)


First, here are the specs of the new chip compared with the previous one.
There are some clear differences in the specs. The key ones are:
The on-resistance is half as high, the frequency is more than twice as low, and the input voltage range is slightly narrower: 23 V instead of 24 V.
On the other hand, the rated current is 2 A.


The chip comes in the same package as the previous one, but as it turned out, the similarities don't end there.


The block diagrams look different, although in essence they're the same.
By the way, I think the frequency was lowered because they used a different MOSFET with lower on-resistance.
Yes, I found one more difference: the reference voltage of the chip under review is 0.805 V, while the previous one has a reference voltage of 0.810 V.
It's a very small difference, so you can use the same resistor values in the divider (they were listed in the previous review).


And here's another similarity.
When I received the chips, I decided to check them before confirming receipt, of course. My first thought was to draw and etch a new board.
But when I looked at the typical application circuit, I found that the chips have exactly the same pinout and connections.
In other words, this chip is practically a drop-in replacement for the previous one, except that it can deliver more output current.


For the test version, I decided to use the same component values as last time. The only difference was the input capacitor: I used 10 µF instead of 5.6.


I did consider simply desoldering the previous chip, but I didn't want to damage a working device.
Another incentive was that last time I etched two boards instead of one. I often do that with small boards; it's very convenient.
The 0.1 µF capacitor isn't in the photo. I somehow forgot about it and added it later.


The manufacturer's reference PCB layout is noticeably different from the previous one. But my thinking went like this: the chip is pin- and circuit-compatible with the previous one, so why not try it on the previous chip's board?
Besides, I didn't see any particular obstacles. The layout is basically correct for both chips.


Like two peas in a pod. The old one is on top, the new one below.




I like these little boards; they're convenient to integrate into all sorts of devices.
Though chips like these aren't usually bought to make boards like this. It's easier and cheaper to use off-the-shelf ones.
The chips are usually used as part of a larger device. For example, TV box circuit boards often have lots of similar little converters (I should check whether they happen to use these too).


The following equipment was used for testing:
Power supply
Electronic load
Multimeter
Another multimeter
Oscilloscope
Non-contact thermometer

The similarity in the photo is probably just a coincidence, but the board produced almost the same output voltage: 4.86 V versus 4.84 V with the previous version.
Ripple with no load is practically zero.


Next, I captured ripple waveforms in different operating modes. Then, while disconnecting the oscilloscope, I accidentally shorted the board's output terminals.
I can't remember now; I think the board was under load, but it all ended with a cloud of magic smoke coming out of the chip. The power supply went into output current limiting, and that was that :(
You can see in the photo that the chip was badly burned. Even when I thought I'd just desolder the burnt chip and solder in a new one, I ran into trouble: the chip seemed welded to some of the traces, and I could barely pry it off.


Oh well, the new fighter is back in action. On to more tests. Though these particular tests were still done with the old chip.
First, let's check the no-load current consumption.
The previous chip consumed half as much, which is noticeably better.


The next test covers two things: ripple level and board efficiency. The multimeter on the left shows the input current, while the one on the right shows the voltage at the board's input (including the voltage drop across the wires from the power supply).
I won't go through every test in detail; it's easier to put all the results in a table at the end of the review. I'll just say that the photo on the left shows a 1 A load, and the one on the right a 2 A load.


The ripple level is practically the same in every mode.




At a 2 A load, I got an unexpected result: the chip went into thermal protection.
On the one hand, that's good, of course—I confirmed that the thermal protection works. On the other hand, it's a bit disappointing, since I could only run a proper test at currents up to 1.6 A. I also measured the chip's temperature under load at that current, since even at 1.8 A the protection kicked in, albeit after a longer time.


After that, I checked the minimum input voltage at which the chip can still maintain the set output.
At 1 A, everything looked good: the output started to drop when the input fell below 5.66 V.
A difference of about 0.8 V is a pretty good result, in my opinion.
At 1.6 A, though, there were problems: the chip kept threatening to trip its thermal protection. It worked normally once the input voltage was raised above 9–10 V.


Still, at the end of the test I decided to take a chance and check the short-circuit protection.
This time, I first used the electronic load to set the current to 5 A and then 10 A. The chip handled both just fine.
Then I shorted the output wires, and that went well too. The inductor got quite hot, but the chip didn't enter hiccup mode.
As a reminder, in the previous review the chip would periodically shut down in this situation, twice a second.
I first tried it with an input voltage in the 10–20 V range, then tested it at a very low input voltage as well.
I couldn't burn out the chip this way.


And, as last time, I'll check the chip's efficiency—or, more precisely, the efficiency of the converter built around it.
The manufacturer claims up to 92% efficiency, but the graph shows only 90–91% with 12 V in and 5 V out, so let's see how it does.


So, what can I say?
At a 1 A load, efficiency is more or less close to the claimed figure, though a little lower. At 2 A, though, it's really bad.
Component temperatures are listed for load currents of 1 A and 1.6 A; the other figures are for 1 A and 2 A. That's because I could still measure the current, but could only measure the temperature at an output current of 1.6 A.


Summary.
Pros
Price
Overcurrent and overtemperature protection work correctly
The short-circuit protection behavior is unclear
Decent efficiency at a load current of 1 A.
Simple circuit; no need for large electrolytic capacitors
Very low ripple
Low reference voltage of the built-in voltage reference, at 0.805 V
Good packaging

Cons
Short-circuit protection does not work correctly in some situations.
Maximum output current of 1.6 A.
Cannot achieve a 100% duty cycle because the driver supply capacitor needs time to charge.

My opinion. I was disappointed with this IC. Compared with the previous one, this one has some noticeable drawbacks. The PCB layout does not follow the datasheet, admittedly, but I don't think that could have affected things THIS much.
The short-circuit protection is one thing—different situations can arise, and it's more of a last-resort safeguard—but the fact that the IC simply can't deliver its rated current is bad.
In practice, I was able to get a load current of 1.6 A, but the case temperature reached 84 degrees (this is what a small spot IR thermometer is for). At 100, the thermal protection kicks in. So the actual output current is no more than 1.5 A; I wouldn't push it any higher.
I probably wouldn't recommend using this IC.

I hope you found this review useful. I have another similar, but more interesting, IC on the way. If people think these reviews are worthwhile, I'll write about that one too.
As always, feel free to ask questions in the comments :)

As a supplement, here's the datasheet for the IC — download.