Author Topic: 2033 transistor replacement (Read 3188 times)

spilihps · « **on:** March 18, 2018, 06:41:49 pm »

Hello!

Almost one more "identify this" thread.

I have a ic marked with 2033, it is what i believe a SOIC-8 package.

I think that ive found that it is somekind of MOSFET. However several things seems to be designated 2033..

So ive found one called IRF7468PbF at my dealer. Do you think that will be a safe replacement?
The Drain, source,gate pin layout is the same atleast.

What i know about the circuit:

It is used for switching a heating circuit for a O2 sensor to ground. the load is about 6-10ohm @ 12v

The old one is shorted between Drain-gate..

spilihps · « **Reply #1 on:** March 19, 2018, 05:32:35 pm »

So i have just some additional information,

a picture of the IC. (its actually cracked?)

And also a crude schematic. I can confirm that i have +5v signal on (What i think is the gate) when it tries to initiate the heating.

https://ibb.co/b1HoNx

https://ibb.co/kzd1aH

Rog520 · « **Reply #2 on:** March 19, 2018, 05:41:17 pm »

TPS2033?

http://www.ti.com/lit/ds/symlink/tps2033.pdf

spilihps · « **Reply #3 on:** March 19, 2018, 05:43:58 pm »

Ive seen that. But would really the OC be connected directly to the rest of the legs?

Rog520 · « **Reply #4 on:** March 19, 2018, 05:46:34 pm »

Yeah, I was wondering that, too. But the other pins seems to line up.

spilihps · « **Reply #5 on:** March 19, 2018, 07:23:07 pm »

And also, it does not mention that 12v would be OK?

This is one strange IC

spilihps · « **Reply #6 on:** March 19, 2018, 07:36:38 pm »

After some image search, i did find something that some chinese site called HAT2033 (that have same line before 2033)
The datasheet: http://pdf1.alldatasheet.jp/datasheet-pdf/view/63552/HITACHI/HAT2033.html

But hard to tell if their image is the correct one. However the datasheet pins seem to make more sense..

Rog520 · « **Reply #7 on:** March 19, 2018, 07:44:28 pm »

Ah, much better match by the look of it!

spilihps · « **Reply #8 on:** March 19, 2018, 07:56:04 pm »

Yeah, and the data seem to be within spec with the one that ive found. However, i do see some differences in the "internal" circuitry that im very unsure of. Any ideas about that?

spilihps · « **Reply #9 on:** March 19, 2018, 08:34:35 pm »

Here is the internal circuitry of the two:

The "old" is the original one.

Cant seem to find anyone that have the exactly same diode layout internally?

Rog520 · « **Reply #10 on:** March 19, 2018, 08:47:15 pm »

The back-to-back diodes from gate to source are meant to protect the gate from overvoltage. Just an added bonus if your new part has them.

Oh, I see your new part (assume that's the IRF7468) *doesn't* have them. You can always add them externally (or a proper TVS rated to keep the gate below its 20v breakdown). Or buy the HAT2033. Looks like they're available on eBay from China.

spilihps · « **Reply #11 on:** March 19, 2018, 09:45:49 pm »

So i guess its needed then?

I was hoping to get a new compatible part from a known source. Dont really trust the ebay ones

wraper · « **Reply #12 on:** March 19, 2018, 10:54:15 pm »

IRF7468 is not a good replacement. It has several times higher gate charge/capacitance and only 12V Vgs instead of 20V, lower Vds. I don't think you should bother about gate protection unless your circuit has something specific in it what could damage gate. It's there for ESD protection, nothing else, and there should not be ESD within circuit.

Quote

It is used for switching a heating circuit for a O2 sensor to ground. the load is about 6-10ohm @ 12v

Is it just occasional on/off or PWM?

spilihps · « **Reply #13 on:** March 19, 2018, 11:33:02 pm »

Thanks very much for your input!

My selection is about this:
https://m.elfa.se/en/semiconductors/discrete-semiconductors/transistors/transistors-mosfets-single/c/cat-DNAV_PL_0101030501?q=*&filter_Category4=Transistorer+%26ndash%3B+MOSFET-transistorer%2C+enkel&filter_Category4=Transistors+-+MOSFETs+Single&filter_Category3=Transistorer&filter_Category3=Transistors&filter_Housing+type=SO-8&filter_Polarity=N

The search filters aren’t that good

As I understand it is just occasional on/off. However I cannot confirm this 100%.
When listening to the gate signal it was constantly high, but it have some surveillance circuitry built in, so after about 2-3 seconds it turns it of and notes a fault code.

Also the hat2033 says high frequency..

wraper · « **Reply #14 on:** March 20, 2018, 12:18:33 am »

You could use, say, Si4436DY, Si4850EY, RSS065N06FRATB, IRF7478
If not PWMed or if gate driver is strong enough to drive higher gate charge, then FDS5672, IRF7855

You should check if package and pinout is right for those as I looked mainly on parameters and might miss something.

spilihps · « **Reply #15 on:** March 20, 2018, 09:55:52 am »

Thanks wraper for your help.

Distrelec did only have the IRF7478PBF from your list. (Dont know what is going on with them these days)
However, they also had IRF7854PbF. It seems pretty close to the 55. What do you say about that one?

The gate is driven from the mcu through a (B11) wich seem to be a PNP transistor. And when i looked at it with the scope, the signal was 5v.

So, MCU -> small pnp -> resistor -> base. If that makes any sense?

Would you say that the IR7478PBF is safe both if PWM or not?

Again thanks.

wraper · « **Reply #16 on:** March 20, 2018, 11:05:01 am »

IRF7854PbF has high gate threshold voltage, so it's unsuitable to be driven by 5V. IR7478PBF has higher gate charge than HAT2033 but I guess it should be fine even if PWMed.

spilihps · « **Reply #17 on:** March 20, 2018, 03:11:45 pm »

Alright, so then the 7854 is no go!

That leaves me IR7478PBF.

No documents for the control unit states that there should be any PWM.
Also, it seem to be very uncommon for this to be PWM controlled. It turns on until the engine heats up, then there should be no heating happening.
I will however check the signal again once i have mounted the 7478 just to see.

About the gate charge, am i reading the diagram correctly if i say they compare like this:
Vgs: about 5
Qg: 25 (7478)
Qg: 12 (HAT2033)

What is the recommended value here for a PWM application?

Also, i do hope that the 2033 is really the HAT2033. Otherwise im in the deep end!

wraper · « **Reply #18 on:** March 20, 2018, 03:33:02 pm »

Quote from: spilihps on March 20, 2018, 03:11:45 pm

What is the recommended value here for a PWM application?

That depends on PWM frequency and gate driver strength. Without details, it's just safer to use closest part.

Quote

I will however check the signal again once i have mounted the 7478 just to see.

Just note that it might 100% duty cycle = DC before target temperature is reached.

spilihps · « **Reply #19 on:** March 20, 2018, 03:37:06 pm »

Alright i see.

Will have to read up on that..

But would you say we are on a safe side atleast with the 7478?

Again thank you..

wraper · « **Reply #20 on:** March 20, 2018, 03:39:38 pm »

Quote from: spilihps on March 20, 2018, 03:37:06 pm

But would you say we are on a safe side atleast with the 7478?

I guess so, I don't think that gate driver does not have any margin to not being able handling higher gate charge.

spilihps · « **Reply #21 on:** March 20, 2018, 03:50:19 pm »

So it all depends on the B11 component, then?

I will mount this, and report back to!

Would be fun to know what made it blow in the first place..
Could a heater circuit that for some reason, got to low impedence do this?

mikerj · « **Reply #22 on:** March 20, 2018, 04:16:44 pm »

A short on the heater circuit could do it, assuming there is no protection for this.

Rog520 · « **Reply #23 on:** March 20, 2018, 04:34:08 pm »

Depending on what's available to you, there are compatible FETs that have lower Qg. For example:

https://www.diodes.com/assets/Datasheets/DMT6017LSS.pdf

http://www.vishay.com/docs/75489/si4850bdy.pdf

Go to Digi-Key or Mouser or the like and do a parametric search, which will turn up all sorts of options...none of which have the gate protection that I can find, however. You can easily add a TVS externally.

If the FET isn't being PWM'd I wouldn't worry so much about gate charge.

If it were me (and you're not in a rush $:=\$ ) I would get the exact replacement off of eBay. Qty 5 for something like USD$9 plus a few bucks shipping.

And then of course, like others have mentioned, find out why the original shorted before blasting through replacements.

spilihps · « **Reply #24 on:** March 20, 2018, 04:48:33 pm »

The O2 sensor is replaced with a new one. So the problem should not still be there.

I think i will go with the one wraper helped me find.

I feel more confident with that, comparing to some ebay stuff from china that i have no idea of the origin of..


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: 2033 transistor replacement (Read 3188 times)

Share me