NeXT Cube - powers on briefly then off again

Toby Thain toby at
Sun Oct 25 21:46:26 CDT 2015

On 2015-10-25 8:56 PM, Toby Thain wrote:
> On 2015-10-25 8:42 PM, Ian Finder wrote:
>> No- if it works with the standard display, the supply is fine.
>> Now that that's clear, I recall some cube supplies would do this
>> without a load for a display- I used to test them by triggering the
>> power-on pin, and seem to remember this behavior occurring if I didn't
>> have a big-ass resistor attached across the pins that normally
>> supplied power to the CRT.
>> Try getting a dummy load on there, the circuitry you have may not be
>> putting enough load on the lines that usually run the CRT to keep the
>> supply in a steady state.
> Even though I'm an electronics noob, that seems pretty logical. Can you
> spell out what kind of resistor I'd need?
> Is it the 20 Ohm 20W between pin 12 (-12V) and GND that is mentioned here:
> Is 20W the right rating?
> Thanks!
> --Toby

I got a really helpful response from Rob Blessin, from whom I bought the 
splitter cable, along the same lines. My supply must be the 152 type 
that requires a load.

Boiling down all the info so far, it seems that a 20 Ohm resistor across 
-12V and GND (maybe 12V and GND would work equally well?) would 
dissipate 7.2W, which seems enough to keep the supply running (that 
other link talked about a 5W load, so this seems a good margin).

Now, 7.2W is more than one resistor in say a DB-19 shell could safely 
dissipate, so I'm maybe looking at some kind of ambiently cooled board. 
Rob provided these links:

   Here: and
   version 1.0 megaload here: 

So I'm wondering what kind of thermal design is both easy and safe. A 
single 10W resistor exposed to the air? Or should I spread it over a few 
resistors on a little board that might have a DB-19 male at the Cube end.

Just noob brainstorming here.


More information about the cctalk mailing list