Indigo 1 Power Supply Basics

Elf

Storybook
Feb 4, 2019
478
129
43
We're here again for Power Supply Basics! This time, the Indigo (Indigo 1) is featured. The Indigo is the oldest machine I own, so I won't be covering any 68ks or IRIS 4Ds unless someone in the area brings one to examine.

The Indigo was a bit of a pain to figure out, only because I started with a dead supply! I spent a long time trying to figure out the secret to turning it on; but as it turns out, there isn't much of a secret to this at all. You just need to use a working PSU.

If referencing or duplicating this information, I ask that you maintain a prominent link to this original forum post for further updates, as well as a credit by name.

Just remember, the best SGI power supply information comes from the Silicon Graphics User Group ;)

Like in the O2, the Indigo power supply fits into a recessed cubby hole in the chassis. Here is the power supply pinout from the Indigo backplane perspective:
Indigo power inlet.png
Power is supplied over a 24 pin Molex Mini-Fit BMI connector. This is like the Mini-Fit Jr. commonly found on wire to board solutions, but has guide posts and no clip. A 24 pin Mini-Fit BMI "plug" (machine side) is pictured to the right, with the Mini-Fit Jr. to the left for pin number reference purposes.

The Indigo back plane (which connects the motherboard, power supply, graphics, and disks) has the male pinned connector, while the Indigo power supply has the female connector.

With a power supply disconnected from the motherboard / load, the main +5V rail is active by default. If you do not see +5V present on a powered but disconnected Indigo power supply, it is likely defective.

Wire colors are notated, however as you can see they are not distinct and do not follow any appreciable standard. Also as noted, the power supply contains a prominent internal speaker, like the Indy.

Pin #Wire ColorFunction
1Black+12V
2Black-12V
3Black+5V
4Black+5V
5Black+5V
6Black+5V
7Black+5V
8Black+5V
9Black+5V
10Black+5V
11BlackSoft start (ground to start)
12BlackActive (+5V logic high when PSU started /
0V logic low when PSU in standby)
13Black0V (return for +12V)
14Black0V (return for -12V)
15White0V
16White0V
17White0V
18White0V
19White0V
20White0V
21YellowSpeaker 1
22White0V
23RedSpeaker 2
24BlackShutdown (+5V logic high pulse to standby)
Pin numbers are as per the mechanical drawing. Wire colors are internal to the power supply, as only the connector is exposed when fully assembled.

Start-up and shutdown of the power supply is worth covering in a few bullet points:
  • The natural state of the power supply with no load attached is standby with the +5V rail active (the +/- 12V rails are disabled)
  • The natural state of the power supply with a load attached is running, where all rails (and the fan) activate; in other words, the power supply senses load on the +5V rail to enter a running state
  • A certain amount of load is required to enter a running state
    • At 0.25A on the 5V rail things start to enable; the fan is spinning at a low speed, and regulation is not stable
    • At 0.75A on the 5V rail, regulation stabilizes and the fan reaches its full speed
  • In the running state Pin 12 (Active) is logic high at +5V; in the standby state the Active line is logic low at 0V
  • The power supply can be soft shutdown by a brief +5V logic high pulse on Pin 24 (Shutdown)
  • In a soft shutdown state, all rails are disabled including the +5V rail, unlike the initial standby state
  • The power supply can be restarted from soft shutdown by a brief 0V logic low pulldown on Pin 11 (Soft start)
Pins 11 (Soft start), 12 (Active), and 24 (Shutdown) run from the backplane to the motherboard and so can be software controlled. Pin 11 (Soft start) is also grounded out to provide a soft start pulse when you push the front momentary Power Button on the PSU.

One thing also worth noting is the power pass-thru on the back; the IEC outlet most commonly used to pass through power to a monitor. This outlet is actually switched, with a relay that turns on only when the power supply is in the active state, as opposed to standby or soft shutdown. In other words, the Indigo power supply is very considerate, and will turn off your monitor for you when the machine is off!

The unit under test is an ITT PowerSystems model PEC4044, part # 6064431, with the following specifications:
RailRating
+5V34A
+12V4A
-12V1A

Test setup:
Indigo PSU wide.jpg Indigo PSU normal.jpg

The dead Indigo power supply (disassembled) is pictured on top of the working Indigo power supply, which incidentally came from the famous "Browndigo" prototype Indigo 1, and has dark purple/brown Personal IRIS colored plastic on the speaker housing to match the main skins.

SCR01.PNG

(C1 - Pin 12 / Active signal, C2 - +5V, C3 - +12V, C4 - -12V)

The above oscilloscope trace shows the power supply transitioning from standby to active, when load is enabled. The transition of C1 from low to high is the point at which the DC load was switched on. Notice that the Active signal does not provide a suitable "Power Good" signal. The +/- 12V rails continue to stabilize well after the Active line is asserted.

Unfortunately many Indigo power supplies seem to be dying or on their way out! At this point I would recommend pre-emptive replacement of capacitors, power diodes, and switching MOSFETs if possible.
 

Elf

Storybook
Feb 4, 2019
478
129
43
@Elf great info, just now seeing this! Did you try to fix your dead Indigo PSU?

Also, was your PSU from a R4k machine? The post here claims the R3k machines had a 25A rating on the +5V rail. https://bukosek.si/hardware/collection/sgi-indigo.html
Thanks! The dead one was from an R4k I believe, wheras the live one was from a prototype R3k ( https://forums.sgi.sh/index.php?threads/what-is-the-rarest-piece-of-sgi-hardware-you-own.62/#post-239 ) :)

I have not tried to fix the dead one, mostly because I am interested in replacing the internals. More on that at a later date!
 

steeph

New member
Feb 21, 2021
3
1
3
Germany
Thank you very much for these information! I have an Indigo with a dead PSU. Especially if I won't be able to fix it, this will be of great help to me to replace it with my own. First I would like to have a go at fixing the original. But I need a bit of help.

When I try to turn it on, all pins (except those of the speaker of course) have about -1.8 V against the case. That's all. I had a look around and apart from dust I found a few scratches on the main PCB with bare copper (those shouldn't be a problem) and one part that looks burnt. It looks to me like a diode, but I can't see any markings. Especially on the bottom side it looks like it was hot for a long time. So maybe not related to the fault. But I would like to compare it to a known working PCB. Attached is a picture of the component in question.

Does this look good to you? I don't even know what I would replace this with.

Of course the capacitors are suspect as well. I measured a few of them without desoldering or cutting anything. But I don't know if that makes sense because wrong values could be cause by the rest of the circuit, right? There are no shorts, but some capacitors are way below what's printed on them. Would you just replace all of them while at it?
 

Attachments

Elf

Storybook
Feb 4, 2019
478
129
43
Thanks! :)

Looks like the refdes is "CR" which is a diode. The underside of the board looks charred and the top side looks like it has electrolyte goop on it. Perhaps the capacitors have leaked?

Accurate capacitance values can't really be measured in circuit, but Indigo 1 supplies are well past replacement age for their capacitors and people have been having success with doing so. What else might have broken (or not) is hard to say, though.
 

Elf

Storybook
Feb 4, 2019
478
129
43
Don't forget to carefully wash the board as well, the electrolyte is somewhat conductive.

Out of curiosity I took a look at some of the I1 supplies I have. While they appear to be different than yours (there were a few versions) they also have a diode in a similar position. However, I'm unable to discern any markings on it. On mine it's an uncharacteristically matte black package; sort of odd.
 

Archaic

New member
May 17, 2021
6
1
3
Does anybody happen to know the orientation of that diode? That 120uF cap right there had leaked all over everything and I forgot to check the diode before I took it out. Since it has no markings I am not sure which way to put it back in.

-edited because I cannot spell.
 
Last edited:

fraserN64

New member
Aug 20, 2019
18
9
3
sites.google.com
Well I figured this would be an easy thing to help with so I took out my Indigo Power Supply and then took the cover off and managed to find the diode in question but I’m not sure my photo will really help :(
25532B9A-FF27-4501-AE29-BDFFECB63978.jpeg
 

Archaic

New member
May 17, 2021
6
1
3
Thank you for the picture and the pointing out of the square vs round pad. With all of the corrosion the solder mask ended up coming off of the square side, so there is no square anymore. I will put the diode in with the anode to the square pad side and let you know how it goes.
 

Archaic

New member
May 17, 2021
6
1
3
DO NOT orient the diode this way!! It goes the other way. That is the loudest screeching noise that I have ever heard from a power supply! I am not sure if my supply is damaged in any other way now, but I have just finished reversing the diode, so we shall see.
 
  • Wow
Reactions: flexion

Archaic

New member
May 17, 2021
6
1
3
It appears that this is actually a 15V Zener diode. (1N5352)
 
Last edited:

Elf

Storybook
Feb 4, 2019
478
129
43
That is the trouble with repairing a lot of these power supplies; unfortunately there are a lot of parts that are difficult to characterize. Gets tough especially on the parts without markings. It might look like a regular diode one way, but without fully characterizing it you'd never know it was a higher voltage zener, for example. :(
 

Archaic

New member
May 17, 2021
6
1
3
That is the trouble with repairing a lot of these power supplies; unfortunately there are a lot of parts that are difficult to characterize. Gets tough especially on the parts without markings. It might look like a regular diode one way, but without fully characterizing it you'd never know it was a higher voltage zener, for example. :(
It is certainly not happy now. I have 1.2V on all of the 5V lines. The -12V is 8V and the +12V is 0V.
 

Elf

Storybook
Feb 4, 2019
478
129
43
Ouch, that is tough; sorry to hear that :(
My guess would be the zener was regulating something that got unhappy with either being pinned to 0.7V (instead of 15V) or went open loop, depending on which way the diode was supposed to be.
 

Archaic

New member
May 17, 2021
6
1
3
I guess that I am going to get some good use out of my desoldering gun this weekend. I will let you know what fried as well as what I did to fix it. (If I can fix it that is)

- edited because I still cannot spell
 
  • Like
Reactions: Elf

About us

  • Silicon Graphics User Group (SGUG) is a community for users, developers, and admirers of Silicon Graphics (SGI) products. We aim to be a friendly hobbyist community for discussing all aspects of SGIs, including use, software development, the IRIX Operating System, and troubleshooting, as well as facilitating hardware exchange.

User Menu