What part is this?

This is from a chinese pirate game (som2), and it trips the voltage down from 5V to 3.6V. They have a 29L3211 chip in there for the rom, which is 3.3V (+-10%). It looks like a diode, but I have no idea what value it would be...



I made a voltage divider using a 47ohm and 100ohm resistor to get to 3.3 volts, and made a Star Ocean cart using a 29L3211 chip. It works for a few minutes but stalls at the same spot during the intro every time. I then connected the cart (ran long wires from the Vcc on my 29L3211) to these diode-looking things and the game worked perfectly.

Anyone have any ideas? Is my voltage regulator made incorrectly?


EDIT: before anyone asks, my voltage divider is: +5V------R1-50ohm------+3.3V------R2-100ohm------GND

They're diodes as you guessed, just lacking anything beyond the end terminals in glass. Their benefit is that the voltage drop is mostly independent of the load put on them, unlike your resistive divider, whose voltage varies based on the load (and dissipates lots of power regardless of load, the tradeoff to make it "stiffer" in the face of a varying load).

so is there any way to tell what type of diode they are? The 2nd and 3rd on the right are cascaded together, and then go directly to the 3.3v chip.

I am interested in this because I dont understand why my voltage regulator causes the game to freeze... but these dont. Does it handle things differently?


Also, the game works fine in my FCtwin, no crashes there using my voltage divider. But its on my SnesJR that it stalls every time, but not when using these diodes.

Hook a volt meter and monitor the output of your divider the whole time, and the output of this divider. Also monitor the +5V into your divider, in case you're severely loading the SNES power supply. Hmmm, maybe a capacitor in parallel with your divider's output would help?

I actually just did that before you posted... I had the multimeter hooked up and watched the voltage to the chip, and it was within 3.29-3.33 V. I also checked the snes and it was putting out a steady voltage as well.

As for the cap, something like a 22uf?

Sure, maybe a 22uF electrolytic and a 0.1uF ceramic (I've read that electrolytics aren't as good for filtering high-frequency junk).

Well the cap fixed it. I used a 47uf cap between the 2 resistors, works perfectly now. So you were right, there was probably some extra voltage garbage going on in the circuit.

Thanks for the help!

They are in fact diodes, and because a diode's forward voltage is almost always the same, you've basically got 5V - 2×0.7V, or somewhere around 3.6V. As long as anything is sinking ~any current at all through it, it'll be about the same voltage.

OK, that makes sense that the capacitor helped. CMOS makes little blips of high current draw when gates change state, especially a lot at once. Since your divider's output voltage depends on load, the voltage would be dropping momentarily during these high-current blips. The capacitor gives a strong output even during these. I had forgotten this important reason for small caps across the power rails of CMOS chips.

Well the way its hooked up with the cap is:

+5V-----47ohm----3.3V out-----47uf cap------100ohm-----GND

So being that its on the input line of the cap, im not sure that its leveling out the voltage or not...i just figured it was filtering out the noise, at least.

A cap between Vcc and GND is called a "bypass capacitor" or "decoupling capacitor" and is used to smooth out power supplies. See also the parable of the balloon.

You should have the cap in parallel with the 100 ohm resistor, not in series.

so if I was to hook it up in parallel...how would I do that exactly? Sorry if thats a dumb question, but if I understand it correctly, rather than having the cap hooked up directly to the resistors I should have it before or after the resistors?

In addition to all the information Google will tell you about series and parallel circuits, I'll say that two components in series have only one of each of their leads connected together, while parallel has both pairs of leads connected.

Well...to update, my game stopped working. I did a voltage test and found it putting out 4.5v instead of the required 3.3V. So, apparently the way I had the cap in the voltage divider caused the divider not to work. crap.

Can you draw a diagram of how this cap should be hooked up in parallel to this voltage divider? I googled it, and I found a bunch of lightbulb diagrams and things like that but nothing to really apply to my cart.

If you need 3.3v don't mess around with diodes, voltage dividers or other such nonsense. Drop the $0.50 to $1 and buy yourself a 5v to 3.3v linear voltage regulator.

Well the problem is that I've tried that, and it doesnt work for some reason. I used a AMS1117 3.3 voltage regulator and the game will not boot, no matter what. I tested the voltage and it comes through at an even 3.30v, and the game wont boot. If I bypass that and hook up 5v to the chip, it will load, but obviously fry in a short time. With the voltage divider it boots, but stalls.

If the game wont boot, but is getting the required 3.3v, what could the cause of the problem be? Am I missing something?

Looks like you've got 3 diodes for two different supplies. The one on the left looks like it's suppling that glob top with ~4v. But the two on the right are in series and suppling ~3.3v to everything else. Make sure you're not neglecting that glob top's power supply. I could be wrong, that's just what I see from the pic.

Right but thats not what I just mentioned. When using a 3.3v voltage regulator it wont load. Using a voltage divider it will load, and then stall. But when I used the diodes pictured the game worked fine, which is why I had wanted to know what kind of diodes they are.

EDIT: Well, i just went out and bought some 1V forward diodes, and those work. I put 2 in series, and it dropped the voltage by 1.8, bringing me down to 3.4V (which works thanks to the 10% tolerance).

So, I figure I will just use these as at least this works. I still have no clue why the voltage regulators I have dont work, and why the voltage divider stalls the game....but the diodes do the trick!

They are almost certainly 3.6V zener diodes, reverse-biased (hooked up "backwards" between Vcc and Gnd, utilizing the clean breakdown properties of a zener diode to get the ~3V).

Edit: The two on the right are in series, so maybe they are not zeners... But a reverse-biased zener diode would work (get one rated for your target voltage, so a 3.6V zener diode reverse-biased across your voltage source and ground will result in a ~3.6V Vcc, which will be fine for a 3.3V chip).

They can't be zeners; they're connected to the wrong supply (to +5V, not ground), the band is pointing the wrong way, there's no load resistor to bias them.

Yeah i think thats what im using now (1N4148T). The guy told me they were zeners, and that generally they will knock it down by about .7v-.9v each. So two in series works out. I have played it for about an hour now and it seems to be fine.

<weep>

No.
Wikipedia explains. The person at the store was confused. All silicon diodes have a forward voltage somewhere around 0.6-1V.

1N4148s are conventional "signal" diodes. They have a well-characterized voltage-current relationship in forward conduction, and are assumed to basically never allow any significant current in reverse.

A zener diode, unlike a conventional diode, is designed so that 1- the reverse breakdown voltage is known and doesn't move, and 2- it can withstand continuous reverse current for an indefinite period of time.

Ok, sorry if i'm making people cry on the inside, just trying to wrap my head around this stuff...... So, if I was to buy a 3.3V zener (1N4728) would that take the voltage DOWN to 3.3V, or take OUT 3.3V from the Vcc (giving me 1.7v)?

Either, but I'd rather talk about the 3.3V regulator you tried using instead. Which one did you get? How is it labelled? How did you hook it up? Did you add capacitors?

And to be honest, you should be able to use the conventional already-purchased diodes in the same way as they're on the StarOcean pirate. (I've got a SATA power cable that uses that technique, even if it weirds me out)

Chances are unfortunately high that you could have already destroyed the 3.3V part you had when you put the capacitor in series.

AMS1117 3.3v voltage regulator, labelled "AMS1117 3.3 011TF". No i didnt try adding a capacitor, as the poster from Racketboy that I linked to did have any caps there for his cart. It was hooked up exactly as he had it done (please check out his thread for pics).

I bet you don't have any pictures of your own assembly?

I'd arbitrarily guess you either needed a little more capacitance, and/or a little shorter of a supply line between the regulator and the load IC. Try adding capacitors between Ground and Vin or Ground and Vout, near to the regulator and/or the IC you're powering off it.

So, at long last I found out what my issue was... It was my SNESjr. I borrowed a friends model 1 SNES and things worked perfectly. So apparently there is something about the snesjr that doesnt like diodes and regulators. Who knew?