turning on old ibm power supply api-6043

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
it has no green or black-brown-blue-white wires?

Re: turning on old ibm power supply api-6043

mynick wrote:
Quoted text here. Click to load it
Because it just had a physical on/off switch in those days?
Just -12/12/5/ground wires.

Re: turning on old ibm power supply api-6043

Quoted text here. Click to load it

yes but with at you had switch gong to/from mentioned mains wires but
this ibm pc300gl psu
has some different system(I doubt ac is going up the connector to mbd
to switch and back)

Re: turning on old ibm power supply api-6043

On 1/26/2011 8:17 PM, mynick wrote:
Quoted text here. Click to load it

I thought you were talking about some super ancient computer but that model
was in this century at least. Does this mean anything?:

Table 25. Power supply connector pin assignments
Pin Signal Function Pin Signal Function
1 3.3 V dc +3.3 V dc     11 3.3 V dc +3.3 V dc
2 3.3 V dc +3.3 V dc     12 -12 V dc -12 V dc
3 COM Ground             13 COM Ground
4 5 V dc +5 V dc         14 PS-ON DC Remote Enable
5 COM Ground             15 COM Ground
6 5 V dc +5 V dc         16 COM Ground
7 COM Ground             17 COM Ground
8 POK PWR GOOD           18 Reserved Reserved
9 5 VSB Standby Voltage  19 5 V dc +5 V dc
10 12 V dc +12 V dc      20 5 V dc +5 V dc

I'll leave it to you to straighten out the columns which show the pin
functions for all 20 of them. Or you could just go to the Lenovo site and
look up the manual for that model as I did


but on first glance the pinout that the Lenovo manual shows seems
suspiciously like a regular old AT 20-pin layout. Could it be that the
color codes on the conductors is simply different while their functions
remain the same?

Re: turning on old ibm power supply api-6043

Quoted text here. Click to load it

Well I suppose it is from pc300
however it has 4 mdb 6 pin connectors and one of them has 2 extra
insulated wires orange and black(other are red,yellow,blue and black)

Re: turning on old ibm power supply api-6043

Quoted text here. Click to load it
thanks for link
this is the one
Connector Location Pin 1 Pin 2 Pin 3 Pin 4 Pin 5 Pin 6
P1 System Board PWRGOOD +5 V +12 V -12 V GND GND
P2 System Board GND GND -5 V +5 V +5 V +5 V
P3 3.5" Diskette Drive +5 V GND GND +12 V
P4 DASD +12 V GND GND +5 V
P5 1 DASD +12 V GND GND +5 V
P9 2 System Board +5 V CONTROL GND
P10 Riser 3 V +3.3 V +3.3 V +3.3 V GND GND GND
P11 System Board 3 V +3.3 V +3.3 V +3.3 V GND GND GND
so only P1/pin1-pwrgood needs to be connected to some black wire-gnd?
btw what is P9/pin2-control for

Re: turning on old ibm power supply api-6043

Quoted text here. Click to load it

conecting pin1 to pin5 or earth did nothing!

Re: turning on old ibm power supply api-6043

mynick wrote:
Quoted text here. Click to load it

PWRGOOD is an output. It could be open collector, so grounding
it wouldn't hurt it. But it isn't "going to make anything go".

(PDF page 55 has Figure 48, with your above quoted table)

That leaves two CONTROL pins as potential candidates. The CONTROL signals
could be wired together. They could be pulled up to +5V with a pullup
resistor. And then, the CONTROL could be active low. Such active low
schemes are good, for "Wired OR" logic, where you want multiple inputs
to affect the same function.

To ascertain what the CONTROL is doing, you'd need a multimeter, and a
minimum of one resistor for testing. The fact there is +5V, CONTROL, GND
on the same cable, is telling you the signal is TTL compatible, with a
five volt range.

First, measure the voltage on CONTROL with the multimeter. If it
reads 5 volts or close enough, the signal could be active low,
and the 5V reading implies that is the off state.

If CONTROL reads zero with respect to GND, that would be more unusual.

Next, you'll need a 1K ohm resistor. You may find 1/4 watt resistors at
Radio Shack. If you were to put the resistor right between +5V and ground,
the power dissipated is V**2/R = 25 milliwatts, so not a lot of power
will be dissipated while you're using it. It'll be safe to use, to probe
the CONTROL signal.

(Example of a 1K ohm. Power rating isn't important here, as we're expecting
no more than 0.025 watts of power dissipation. The larger resistors, will
have stiffer leads on them, and may fit the connector a bit better.)


Now, use the 1K ohm resistor, to pull the CONTROL signal in the opposite
direction the initial meter reading gave you. If the CONTROL signal read
+5V in the first step, connect CONTROL to GND on the connector, using the
1K ohm resistor. Now, measure the voltage between CONTROL and GND. If you
saw no change at all, in the measured value, then CONTROL must be an output.
If you see a "voltage divider action", then you know it's an input.
For example, if the original reading was 5V, you connected the 1K ohm to
ground, and the voltage between CONTROL and GND was 2.5V, that means the
internal pullup is also a 1K ohm resistor. If you measured the voltage, and
the value was 0.45 volts, then the pullup is 10K internally, and you're
seeing 5V * 1K/(1K+10K) or 0.45 volts as a result. Now, if CONTROL was
at 0.45V, that would also be a low enough value to cause the power supply
to start. Which might scare you a bit.

Probing with the 1K ohm resistor, is intended to reassure you whether
CONTROL is an input or an output. And do it, with less chance of damaging

Again, with the multimeter, switch to OHMS range, and with the power off,
measure from CONTROL on P9 to CONTROL on P12. They could be the same
signal, and the intent *might* be to support Wake On LAN as well as
soft power via the system board. This isn't too important, except to
understand why there are two CONTROL signals. If it is the same signal
on both connectors (as proved by a zero ohm meter reading), then it
makes the function of CONTROL a bit easier to understand.

After your initial "research steps" are completed, you've assured
yourself it's an input, *then* you can try pulling the signal
a bit harder in the opposite direction. On an ATX supply, for example,
PS_ON# is active low, and you ground it to enable "soft power" and
start the ATX supply. I don't know if that 145W supply works that
way or not, but I'd take my time probing the thing, to get a better
handle on how it works. Bare minimum, I'd work with a 1K ohm test
resistor, and a few ranges on my multimeter, to decide whether
CONTROL is an input or output. PWRGOOD will be an output. There is
no point grounding that one.

Does anyone have a clue, what power supply standard that is ? It's
pretty weird. If I check this table, the "P8/P9" on an AT supply,
seem to match the P1/P2 in the IBM/Lenovo document. But I thought
AT didn't have soft power, and came on when you hit the switch ???


    P8 AT   PG +5V +12V -12V GND GND
    P9 AT   GND GND -5V +5V +5V +5V


    "Soft Power (Power On and 5V Standby Signals)

     Early PCs using the PC/XT, AT, Baby AT and LPX form factors all
     use a mechanical switch to turn the computer on and off. Newer
     form factors, starting with the ATX/NLX, and including the SFX
     and WTX, have changed the way the power supply is turned on and off.
     Instead of using a physical switch, these systems are turned on
     by a signal from the motherboard telling the power supply what to do."


Site Timeline