FSP300-60THA - FSP 3528 datasheet equivalent

[retiredcaps]

The FSP3528 chip is in the FSP300-60THA. It looks like the FSP 3528 is a proprietary chip and I can't find a datasheet on it to see how it works.

There are some schematics on the Internet, but I'm not that skilled at reading them (and they are in Russian).

BTW, I have a good FSP300-60THA and a bad one. I took voltage readings of both.

The bad one only has 5.0V standby on the green wire. When I jumper green to black, I get no secondary voltages and the gray wire is 0.0V.

Good			Bad
pin 1 - 4.99		5.03		VCC
pin 2 - 2.354		0.129		COMP
pin 3 - 1.221		1.223		EA-
pin 4 - 1.224		0		EA+
pin 5 - 4.98		5.02		TREM
pin 6 - 0.398		0.403		REM
pin 7 - 1.226		1.225		RT
pin 8 - 1.829		1.837		CT
pin 9 - 2.058		0.958		DET
pin10 - 4.98		0		TPG
pin11 - 4.87		0		PG
pin12 - 3.534		3.526		VREF
pin13 - 3.339		0		V3.3
pin14 - 4.88		0		V5
pin15 - 3.914		0		V12
pin16 - 0.353		0.360		PT
pin17 - 0        0		GND
pin18 - 0.185		0.189		DTC
pin19 - 1.813		1.268		C2
pin20 - 1.805		1.268		C1

QUESTION: Is there a popular equivalent chip to the FSP 3528 where I can download a datasheet to at least understand the high level working of that chip? Once I understand how the "popular" chip works, I can be more informed about the FSP 3528.

Nick over at the CubeComputerChannel took a good pic so I will just reference his at

https://www.badcaps.net/forum/attach...9&d=1343774433

edit: Found the answer

http://www.hardwaresecrets.com/print...ply-Review/975

"SD6109 and SG6105"

[mockingbird]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Delta's custom ICs on their PSUs are manufactured by OnSemi with different part numbers I believe. It's just a matter of corellating the differing part numbers. FSP probably has the same shtick going. Who manufactures it for them is the question.

[selldoor]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Missed the edit - -just checking -you do know that you can copy a web page link into google translate and it will translate the whole page and also some(most) of the links on the page so if
it is a foreign "badcaps" you can read whole threads.

http://translate.google.com/translat...2.html&act=url

Try this one
http://translate.google.com/translat...260695&act=url

Your Russian seems pretty fluent

[Duncanbest]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

You can download the datasheet of KA3511 as reference.

It's similar function with different pin numbers.

google KA3511 .

[retiredcaps]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

You can download the datasheet of KA3511 as reference.

Thanks for that lead. Downloaded and reading now.

[everell]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Did you get your power supply working?

I have a FSP300-60THA(1). Mine does work. It has the FSP3528 pwm chip and has DM311 for 5vsb rather than the old two transistor circuit. Do your power supplies have the DM311 chip?

I looked at AN4003 (Fairchild Application Note, PC Power Supply Design with KA3511). It seems to relate rather well with the schematics on the Russian websites. I am thinking that perhaps there is (or was) a 35xx series of pwm chips, and FSP3528 is the Fortron proprietary chip whereas the KA3511 is the one available to the public. If you haven't seen AN4003, just google Fairchild AN4003 and download you a copy. It has a lot of good circuit explanations.

[retiredcaps]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Did you get your power supply working?

Not yet. I'm trying to understand the various pin functions and their thresholds. That way I can try and narrow down where the problem is.

Both power supplies have the DM311 chip.

I don't want to replace any of the caps even though they are crap until I have evidence pointing to them. Unfortunately, I don't have an ESR meter.

My best guess right now based on my voltage readings in post #1 is the PS is in some sort of protection mode.

I have tested a lot of diodes, transistors and resistors and so far I haven't found anything out of the ordinary.

From reading the datasheets, the functionality of a multimeter may be limited. I made need a scope to capture all the different signals and timing parameters.

If you have any suggestions, I'm listening.

Slight off topic.

The thing that is really slowly me down is the cramped space in which to take voltage measurements. It also doesn't help that my spatial mapping skills aren't the best. I'm also trying not to short out the PS when I probe the underside so I keep a plastic shield between the heatsinks and the chassis.

I have no idea how you work on these things. I have admired your work in the past, but now appreciate the difficulties in dealing with this form factor. I prefer working the flat lcd power supplies where everything is relatively easy to probe and some pcbs are marked front and back!!

[everell]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

The method I use is to remove the pc board completely out of the chassis. This requires unsoldering the two wires that go to the AC outlet, unsoldering the two wires that go to the 115v/230v switch, and unsoldering the two fan wires. Trying to keep pc board components from touching the chassis and shorting things is beyond my clumsy ability.

Then solder a wire where the two wires were for the 115v/230v switch. This will configure the pc board for 115 volts while it is out of the chassis. If possible, I also solder a two pin socket for the AC line. It is easier to use an AC cord with a plug than to solder an AC cord to the pc board. Then you plug the cord in when you want to power up......and unplug the cord when you want to solder parts. I use a power strip with on/off switch to plug the power cord into.......that way I have a way to turn the power supply on and off.

I have a home made power supply "checker" with LEDs for 3.3v, 5v, 12v, 5vsb, and power good. This gives me a good visual indication of what is going on. It also has a toggle switch to simulate the computer on/off switch controlling the PS-ON function.

I usually use one HDD and one CDROM as a load. Results on my Fortron were surprising. So here are the voltages I got.

With NO LOAD, 3.37v, 5.06v, 12.19v, and 5.07vsb

With HDD and CDROM, 3.36v, 4.84v, 12.84v, and 5.07vsb

With special load I built using power resistors: 2.5 ohms for 5v, 12 ohms for 12v
gives 3.36v, 5.07v, 12.13v, and 5.07vsb

As you can see, loading from the HDD and CDROM really messed up the regulation. Perhaps my HDD and CDROM are in worse shape than I thought! From this you can see that what kind of loading you use can have adverse effects on loading and possibly even the protection circuits.

When troubleshooting a dead psu, one thing I look for is whether or not the LEDs on my checker flicker. Also look to see if the fan spins a turn or two. It points in the direction you need to be looking.

Hope this helps to get you started.

[retiredcaps]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Hope this helps to get you started.

Hi everell,

So I got a bit of movtiation after reading your response and decided that I would take another crack at it with the existing tools I have. I didn't desolder the wires, but instead, pried the cases back a bit so I would have more room to work.

Since I wasn't going to build a LED power checker, I decided to try and use min/max and manual range mode on my multimeter. The Fluke also has a bar graph that updates 40/seconds.

So with all those settings, I was hoping to capture any brief under and over voltage condition on the schottky rectifiers on the secondary side.

Even with all those settings, the multimeter never moved from 0.0V.

So I started probing the transistors and found that Q6 (C945 NPN transistor) had 15.76V DC on the collector pin. The good power supply had a voltage of 11.76V.

I then traced the source back to a Teapo SEK capacitor rated at 100uF 25V (the pcb designation is covered up), but I circled (red) its location in the picture. I found this out at 1:30am in the morning, but decided not to solder that late at night.

I didn't have an exact replacement so I used a Rubycon 220uF 16V ZLH and soldered it in this afternoon. As soon as I hit the power switch the fan started to spin. With an old hard drive and fan acting as load, I got voltages on the rails: 12.717V, 5.0675V standby, 4.9581V, and 3.33544V.

The readings for the FSP3528 chip look like the same as the "good" PS.

I measured the capacitance of the Teapo and as you can see, it is exactly 100uF. I don't have an ESR meter to measure the ESR.

Thanks to everell, Duncanbest, selldoor and mockingbird for helping me learn how to fish in the world of ATX power supplies. I wanted to find root cause rather than blanket and blindly replace all the caps even though they are crap brands.

PS. The "good" power supply had a bad 2200uF 10V capacitor on the 5V rail. The "bad" one, in addition, to the bad Teapo had a bulged Capxon 3300uF 16V cap.

I can't really complain though because both these power supplies came with an Acer quad core CPU system (with DDR2) for free from the previous owners. I also gained valuable knowledge in the process.

[selldoor]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

I also gained valuable knowledge in the process.

So did anyone watching! I've managed to repair a couple of PC power supplies but only in so far as replacing caps - wouldnt really want to get in any deeper, but now we have a few more clues - thanks.

Well done - good fix

[everell]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

More than a year ago I started trying to draw a schematic of the FSP3528 while I was working on a Fortron FSP300-60ATV. So to try to help, I was motivated to try to clean up and finish the schematic I started. So here is the first drawing I made, and the completed masterpiece I just finished. Perhaps this will give you some answers. Although the drawing is from an earlier model, the FSP3528 circuit should be very similar.

[retiredcaps]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

More than a year ago I started trying to draw a schematic of the FSP3528 while I was working on a Fortron FSP300-60ATV.

Your efforts to help are greatly appreciated.

I did fine your original schematic in my earlier search and downloaded it. My schematic reading and spatial mapping skills require a lot of work. The "equivalent" datasheet really helped though. It helped me understand the 2 letter codes and conditions that cause the protection.

I'm working on another power supply that is intermittently dying right now. I'll post more details and open a new thread once I can get some decent readings and reproduce the problem. At first glance, it also uses a proprietary or rebranded pwm chip so I cannot find a datasheet.

[mockingbird]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

I have an earlier revision of this PSU without the DM311 chip...

The problem is that the PSU went up with a puff of smoke and I think I traced it down to an N-Channel MOSFET (Datasheet attached) on the primary heatsink. I'm not entirely sure that this is what burnt, but would it make sense that this transistor would prevent the primaries from charging up or would it be the bridge rectifier or something smaller like a diode or resistor?

The other two transistors on the primary heatsink are E13009.

edit: Just to clarify, the supposedly busted FET is the CEF02N7.

[everell]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Your bad FET transistor is one of the two transistors in the two transistor 5vsb circuit. Until you get the 5vsb circuit fixed, the power supply is not going to work! The other transistor in that circuit is probably also blown. So this would be a good candidate for a DM311 (or other pwm chip) mod.

[mockingbird]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

But what would cause the primaries not to charge up at all, specifically? (0 volts on both of them).

Like I stated the prmary heatsink has three TO-220 packages. Two E13009 and one CEF02N7. I suspect the CEF02N7 is busted (Though I have no way to test it because my DMM has no diode function). My question is, does the role the CEF02N7 plays have something to do with the primaries charging up, and would its defective state cause them not to charge up?

Yes, it's a prime candidate for a DM311 mod... If I can get a newer revision FSP300-60THA, I think I could possibly mimic it (Such is the limit of my current technical ability - so that I am only able to implement circuits extrapolatively for the most part).

[cubecompmtdx]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

If the primary side isn't charging up, then I would guess that it could be the bridge rectifier Or if the unit has a main on/off switch on the back (not the voltage switch), then there's a possibility the even the switch could be bad in terms of staying open circuit. Also, just in case you haven't done so already, check the fuse. When primary switchers fail, the can pull a lot of mains current through them. I'm guessing since your unit lacks the DM311 chip, then it might use a two-transistor circuit. Older units, including FSP, tend to have these. That CEF02N7 might be the primary side 5vsb transistor.

[mockingbird]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Checked the fuse, it's fine. It's probably not the switch because the person who gave it to me said that he saw smoke coming out of the PSU. The CEF02N7 is a 700V 1.9A part... That's why I suspect it has something to do with the primaries.

[cubecompmtdx]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

that component works with primary side voltages, so that why it's rated for 700v. The voltages get stepped down by the transformers. The way a SMPS works is it takes in 50/60Hz AC, rectifies it to DC, and stores it in the primary capacitors. Active PFC units work slightly different at this stage. The switching transistors take the DC power and turn it into a high frequency AC supply. I've heard that in some units the frequency is around 60kHz. With a high frequency, smaller transformers can be used to output more amps. This HF AC is stepped down by the transformers and is then fed to the secondary side rectifiers, which convert it back to DC again. From there, the outputs are cleaned by inductors and capacitors, etc and fed to the load. This is just a general idea of how computer power supplies work.

[lti]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

Are you getting voltage on the AC side of the bridge rectifier?

There is also a thermistor on the input that could have failed.

Make sure the transistors aren't shorted.

[gordontwist]

Re: FSP300-60THA - FSP 3528 datasheet equivalent

...

I then traced the source back to a Teapo SEK capacitor rated at 100uF 25V (the pcb designation is covered up), but I circled (red) its location in the picture. I found this out at 1:30am in the morning, but decided not to solder that late at night.

I didn't have an exact replacement so I used a Rubycon 220uF 16V ZLH and soldered it in this afternoon. As soon as I hit the power switch the fan started to spin. With an old hard drive and fan acting as load, I got voltages on the rails: 12.717V, 5.0675V standby, 4.9581V, and 3.33544V.

The readings for the FSP3528 chip look like the same as the "good" PS.
...

That'll be C37...

I read with interest - is the conclusion that this cap just has bad ESR even though it's capacitance is in spec - and this was causing the problem?

Gordon