iMAC G5 ALS PowerSupply API4PC46-290 REV.A R37 near 3845B

[Santaken]

I have iMAC G5 ALS 20' PowerSupply is API4PC26-290 REV.A
When I try to change all CAPs,I lost R37 with some pattern(pic 1)
Please show me pattern and register-R37.

[Toasty]

[Santaken]

Thabks!! very well.
Good skeleton!.
Register look like "621"(620 ohm)
I'll Try.

[Toasty]

A "regular" 620Ω, 1/8w, 2% resistor should do fine as the connection is direct to the (-) leg of C1. -BUT-, you have to establish the connection on the top of the board since the trace was damaged.

I find pre-tinning the capacitor leg with solder, leaving a -very- small "bump" or blob of solder on the leg near the cap body, usually works. When you solder the cap in, heat the cap leg enough to cause the solder "bump" to liquefy. Then, do a continuity check from the cap leg to the points indicated in the attached pictures.

Toast

[Santaken]

Thanks for reply. "C1"'s Through-hole seems good but "Q1" is full open.
I was looking for "Q1" another thread, why destroy "Q1"?
Over current TI3845B? or
Reverse schema a part of pattern possibly with some mistake.

[Toasty]

Nice schematic!

Q1 failure - My theory is that when C68 & C40 fail (the output filter caps on T2, the standby transformer {TR_SB}), the input drive from the TOPSWITCH (M6) is effectively "unloaded" and the voltage rises on the HOT side secondary*(?) winding of T2. Either it does not rise far enough to compromise C37's 50v rating, or C37 by its nature is adding some "limiting" to the voltage.

In many cases where Q1 has failed, M2 (3845) looks like it has been overheated. I have had a few where M2 has been "cratered" (big chunk missing).

Since Q1 is what feeds M2 power directly from Vcc (T2), and Q1 is "supposed" to be controlling the power to M2. The voltage must be in excess of the 34v internal Zener in M2, which in turn overloads Q1 eventually causing it to short and sometimes crater M2.


*T2 - has 4 windings.
- Primary from the TOPSWITCH
- What I call the "HOT side secondary" which feeds the circuit indicated here as indicated in your schematic. It is isolated from the primary, but on the same side of the core. Most likely "in phase" with the primary.
- 2 secondary outputs which supply the 5v and 24v standby


Toast

[torgeson5kqs]

I have a few power supplies with this exact model where the Q1 is blown or cratered. In some cases that chip is labeled 1P or W1P (is there a difference?). I've replaced the 3845B (M2) chip as well but still no luck. All caps are either replaced or tested with ESR meter. I've also done a diode check on M6 in circuit which appears good. Still I only get standby voltages. Do you find yourself just replacing Q1 and M2 and then get the system to power up?

[Toasty]

Must get standby power working first. Check outputs from TOPSwitch and T2.

[torgeson5kqs]

Testing from the bottom (underside) of the psu I get a massive 162.5 volts from one leg on TOPswitch (M6), two other legs read 5.7 volts, and the other is a shared ground. On the T2 two legs read 161.5 volts. I've included a picture to better illustrate what I'm talking about. On this particular PSU I get all the voltages except for pin 8 standby. The other pins on T2 I did not get any voltages at all.

[Toasty]

Perhaps something on the output side of T2 has shorted? Or is open?

Measure across the 2 caps right next to T2. In the OP's 1st post, 3rd pic they are just to the right of T2. The closest one is C68 for the 24vsb. The one next to it is C40 for the 5vsb.

Measure across C37, the 100µF/50v cap next to the TOPSwitch. It's the filter for the power to that standby/startup circuit that is fed from the "Hot side" secondary winding off T2. You should see ~21v. See schematic, upper left.

.

[torgeson5kqs]

C40= 5.1 volts
C68= 25.9 volts
C37= 12.1 volts

[Toasty]

All small caps in startup should be replaced with FM, FC, HE and the like.

C1 in schematic is actually C11. C1 connects to Vcc2 (upper right) and the 2 resistors there are actually R40 & R42 (2 - R40 shown).

As built values are:
C1 - 2.2µF/50v
C11 - 22µF/50v
C37 - 100µF/50v
C69 - 47µF/25v

Voltages I gave before are with PSU connected to mains and in standby, not "on".

Obviously something is dragging the C37 voltage down.

Across the 4 small caps the voltages when [standby] & <on>

C1 - [0] <5>
C11 - [3.2] <15.7>
C37 - [25.1] <21.3>
C69 - [5.8] <5.8>

.

[torgeson5kqs]

1st is standby, 2nd value is "On"

C1 - 0 and 4.9
C11 - 2.6 and 10.7
C37 - 15.4 and 12.2
C69 - 5.7 and 5.7

[Toasty]

Were the 4 caps replaced?

[torgeson5kqs]

I will have to order the caps from mouser. However, would I be able to pull them and test with an ESR meter or capacitance tester to see if they are still good?

[Toasty]

With the Blue ESR meter they can be tested in circuit. Capacitance is likely fine.

>>On this particular PSU I get all the voltages except for pin 8 standby.<<
What does that mean?

[torgeson5kqs]

I do have a Blue ESR meter, the readings are below. Regarding your question about pin 8, what I meant was I don't get a voltage on that pin. I believe pin 8 and 22 should be standby voltages. That voltage should be about 5-5.1 volts. However, I do get voltages on all the other pins (except for ground pins) when the unit is jumpered at 14 & 15. On the C1 reading below I also cross refrenced with another identical PSU that I have (also) bad and the reading for C1 on that was 7.5. C37 looks to be way off. Can you give me a ballpark of what the proper ESR values should be on all these?

ESR readings (in circuit)
C1 - -- (no reading)
C11 - 4.8
C37 - 40
C69 - 5.9

[Toasty]

http://www.badcaps.net/forum/showthread.php?p=78821

http://www.badcaps.net/forum/attachm...chmentid=10072

Power Good is a signal to the motherboard that the PSU is working properly.

C1 - bad
C11 - bad
C37 - waaaaay bad
C69 - bad

Acceptable:
C1 - < 3.0
C11 - < 1.5
C37 - < 0.5
C69 - < 0.75

Readings with FM/HE in place:
C1 - 1.5
C11 - 0.18
C37 - 0.04
C69 - 0.14

.

[torgeson5kqs]

I actually have both of those links you sent printed out on my workbench, just got pin 8 and 9 mixed up. I'll go ahead and order those caps from mouser, probably be about a week.

[torgeson5kqs]

Got the caps in and replaced all the ones you recommended and now that power supply works. Thank you so much. I do have another identical one though that just gets the standby voltages and nothing else. The Q1 is almost always blown on all these types of supplies so I've replaced it but still no other voltages show up when jumpered to "On". Any ideas?