>>> Troubleshooting Delta Power MAX 500ATX ATX power supply

[dragon3x]

Hi, this is for troubleshooting and, hopefully, repairing a Delta Power psu
that is still quite new, as it worked OK for some months until it failed.
At the time I was not into power supply electronics; it was put
aside and a new power supply installed inside the computer. Also at the time
power supplies were half the price of today's.

On the PCB there are a D339 and 7500B chips.

I have found that schematics with these 2 chips (see included file).
Does it represent the actual circuit of the MAX-500ATX, I cannot tell.

Some measurements, but first: In my country power is 120 V. / 60 Hz.
V standby is OK at 5.0x volts
On the high voltage rectifier diodes I read +166 and -166 V.
On the high voltage heatsink measure is -166 V.
So far so good. But there is no "good" anymore. No 12 V., no 5 V.,
no 3.3 V.

Inside the power supply box itself there is no apparent sign of
burnt/defective part (see photo included).

To debug that circuit what would be the next measurements ?

Thanks for your time and interest.

[dmill89]

Note: that unit was definitely not manufactured by Delta electronics (the Delta most refer to when talking about PSUs), and looks to be in the 200W-300W range at best. Looks like the typical cheapo generic Chinese unit with a somewhat scammy name (since I don't believe Delta Electronics has ever use the "Delta Power" name, either branding their units just "Delta" or the many OE units they produce for other brands/pc makers, it is probably "technically" not a rip-off but definitely makes it sound like the unit is related to Delta Electronics when it is definitely not). Probably an ancient half-bridge design with super-group (note the one output coil) regulation (which tends to be rather poor). The 4-diodes for a bridge rectifier tend to be a weak point on these but if you're seeing voltage on the primary that likely isn't the failure point here. Given this unit likely lacks anything in the way of overcurrent protection if it was ever loaded anywhere near its "rating" it is likely that something was overloaded and failed. You can start with testing the primary switching transistors, then secondary rectifiers, and of course caps. The almost total lack of input filtering means that any kind of surge or "dirty" power could have also cause damage.

[stj]

check the 120/240 switch is working properly,
it has a voltage-doubler design and should be pushing well over 200v

[harp]

What is resistance of 12, 5 and 3v3 output?
If you find it shorted, look the diode on secondary side, often fails.
If it is not shorts then for any reason switching do not work, maybe faulty mosfet or failure of poweron signal circuit, or drivers.

Regardless to Delta, I think that all of them have mandatory their logo on pcb board.
The heatsink looks so thin...

[dragon3x]

Hi, @stj: the voltage switch has been removed and a jumper wire soldered on the pcb for the 120 V. selection.

Here are power transistor voltages on the high-voltage heat sink.
In front of these transistors, from left to right:
Q1-Pin 1: -163 V.
Q1-Pin 2: +163 V.
Q1-Pin 3: -163 V.

Q2-Pin 1: 0 V.
Q2-Pin 2: +163 V.
Q2-Pin 3: 0 V. (connected to Q3-Pin 2)

Q3-Pin 1: -163 V.
Q3-Pin 2: 0 V. (connected to Q2-Pin 3)
Q3-Pin 3: -163 V.

Heat sink vltage: -163 V.

@harp: I will test resistance of 12, 5 and 3v3 output in my next post.

[dragon3x]

Hi, harp : With power supply not connected to 120 V., resistance to ground is:
for 12 V. : 230 ohms
for 5 V. : 22 ohms
for 3.3 V. : 15 ohms

Hoping this will help somewhat.

[harp]

It looks ok, I have simmilar atx 450w with 12v/250R, 5v/25.5R and 3.3v/25.0R...

Ok, offline check first:
- main mosfet, to confirm thre is no short on them
- 7500 and 339, check supply rail to this chip there is no short


Online mains check, with 60w current limit bulb in series:
-can you describe exactly whats going on output when it is connected to mains
-fan spins, a even little?
-if you grounded power-on is something changes? Is steady 0.00v or fluctating from where to where?

So you say (purple) standby 5v is ok, what is voltage on green wire?
- voltages on both IC supply pin, is steady, fluctating...?

What dmm you have, can you measure whats going on gate of mosfet?

[PeteS in CA]

As dmill89 pointed out above, that is not a Delta Electronics//Delta Products power supply. Delta would not use wimpy heatsinks like that for a 500W power supply, and probably not even for a 250W unit. Delta doesn't install jumper wires where common mode inductors belong. Delta would NEVER use 1KV DC or 2KV DC ceramic caps instead of agency-approved Y-caps! Never!!! That output filter section is pretty pathetic. The input rectifiers are pretty small, and the output transformer core would not be good for much over 250W. I would not apply power to the thing without removing those unsafe caps in the Y-cap locations, but I think the 200-250 W estimate above is probably correct.

The 7500B is a clone of the venerable TL494. It's a decent enough voltage mode PWM, but Delta Products uses 384x series current mode PWMs or custom ASICs in its consumer-grade power supplies. The venerable LM339 is a very decent quad comparator, but Delta uses ASIC or dedicated supervisor ICs.

Do each of the input electrolytics have ~165V across them? Or about half that. If the latter is the case, then voltage doubler is not working or the jumper you think hard-wires it does something else.

[dragon3x]

Hi, thanks or replying
PeteS in CA : From my initial message:
"On the high voltage rectifier diodes I read +166 and -166 V. "
"On the high voltage heatsink measure is -166 V. "
harp : With power supply not connected to 120 V.:
There is no shorts on the 3 power transistors on the high voltage side.
There is no shorts on the LM339 and 7500B ICs

Yesterday while I was doing some measurements a small wire fell on the PCB.
Result: 2 resistors 1/8 watt are broken and the 5 V. standby is gone. To repair
I need a schematics. The included "DeltaPower_MAX-500ATX_Schematics.png"
in my initial message at first seemed OK but now I need precision and something
doesn't work in the schematics resistors around the 2 ICs compared with the
MAX-500ATX psu.
If KA7500B is a clone of the TL494 then I will search schematics with TL494 + LM339.

This is the situation for now and I continue my detective work on this.

[dragon3x]

Hi, this is an update about that Delta Power MAX 500ATX project.

After searching on internet I realised I could not find the exact schematics needed. Some "seem" to have
the same exact topology and same parts but, looking more attentively, many parts are not identified.
Others are well documented but the topology/circuit itself is not the same.

Then I tought why not remove/unsolder the resistors, capacitors etc. and after that build a new PSU from an
exact schematics that I have found. Anyway I cannot repair the actual Delta Power, what choice do I have ?
So I started that "revised" project. Suprisingly, removing the parts was easier than I tought in the beginning.
Once done, I began the reflexions about building the new schematics in the Delta Power box. Then I hit a wall:
where to begin with, how to adapt the PCB to the new circuit, are the Delta T1, T2 and T3 transformers
appropriate for the new circuit ......... Suddenly there was too many questions and not enough answers.

Then I tought (again) why not undertake a review of the recent/modern SMPS ICs that are available on the
market, the idea being they are surely better than the archaic chips in cheapo power supplies. The best of
them are protected against whatever can happen: overvoltage, undervoltage, short circuit, open circuit,
over temperature, overload, ...... Also, for a typical circuit, parts count is lower. On the negative side the
learning curve can be steep.
Example ICs: Infineon ICE5xSAG, ICE5ARxxxxBZS, NXP TEA1623P, TEA172x, Onsemi NCP1362, UC384xA,
------------------ST Microelectronics 6566B, Texas Instruments UC3842, UC3843 (not a complete list).

This is where I am now. I plan to start with Infineon, we shall see. This is not a short term project and I hope
it won't become a long term adventure. At least it is looking forward with new generation SMPS, not backward
to outdated electronics.
I will post a project update from time to time.