Laserjet III DC power supply probs

[GanjaTron]

Hi folks,

has anyone out there ever managed to troubleshoot the DC power supply
in one of these dinosaurs? Mine's developed a fault in the 24V supply,
killing the fans and main motor < 1 minute after powerup. Input from
the AC module remains stable (220VAC) , as does the 5V output since the
panel still functions and gripes about a "SERVICE 55". According to the
service manual, 55 indicates a DC controller error, but I've ruled
that out since the 24V output from the DC PSU still drops after
detaching the DC controller.

I'd appreciate any comments on diagnosing this baby on the component
level. Sure, replacing the entire module is a piece of cake, but hardly
worth the cost when I can get an LJ4+ for the same price (which of
course isn't quite as rock solid)...

Thanks 'n' stuff!

--GanjaTron

 

[Ben Myers]

The LJ4, LJ4+ and LJ5 are just as rock-solid as the LJ3. They all have better
power-saving features. Retire the LJ3 unless you can get one for free to
scavenge for parts... Ben Myers

 

[Kevin]

Go to www.anderson-tech.com and look for the toll free number. Give Mr.
Anderson a call. He may be able to help you out.

 

[yourtown]

Let me know how you make out. I have several PS and remnants in the
shop which I can let go at modest price. Contact me
therepairman at optonline.net

 

[GanjaTron]

Let me know how you make out. I have several PS and remnants in the
shop which I can let go at modest price.

Assuming this is a thermal problem, I poked around with cold spray in
there -- to no avail. None of the suspect components (regulators, etc)
reacted to cooling. Without schematics, I'm pretty screwed anyway...

--GT

 

[ato_zee]

Assuming this is a thermal problem, I poked around with cold spray in
there -- to no avail. None of the suspect components (regulators, etc)
reacted to cooling. Without schematics, I'm pretty screwed anyway...

If you are happy to dig into component level, there are probably a
couple or so of 6-legged opto-couplers in the DC PSU. might be
worth monitoring voltages across the LED's and their o/p to see
if they are respondingt to drive. The LED's are known to die.
That was the problem with my HP (Canon engine) laser. Ended
up putting in a 6-pin skt, and a red LED in series with the opto
LED drive. Now red LED lights in sympathy with the fuser lamp,
if fuser doesn't light diagnosis is simple.
..
You will probably find a service manual on the internet, try
http://www.eserviceinfo.com/
for starters, though there are other sites. You might have to
do a bit of digging on the ESI site as it's indexing seems a
bit flakey, may be ok when you get the hang of it.

Other than that it's a case of following logic level signals
and strategic measuring of voltages. Like does the fuser
lamp come on at start up, if not why? Does it start to
cycle?

 

[Mike]

as old as this thing is, and the heat over the years, it wouldnt surprise me
that the radial electrolytics in the power supply is leaking or dried.

I know this is a far cry from a fax machine, but I worked on a fax machine
with similar problems, Recapped the power supply, and all went away.

 

[GanjaTron]

If you are happy to dig into component level,

Well... not exactly *happy*... :^)

there are probably a
couple or so of 6-legged opto-couplers in the DC PSU.

Optocouplers in a PSU? Didn't see those. I'll take a closer look.

That was the problem with my HP (Canon engine) laser.

All old(er) HPs use Canon engines, AFAIK.

You will probably find a service manual on the internet, try
http://www.eserviceinfo.com/
for starters, though there are other sites. You might have to
do a bit of digging on the ESI site as it's indexing seems a
bit flakey, may be ok when you get the hang of it.

Thanks for the link! Found some interesting stuff on there, though as
you point out, the index *is* pretty convoluted.

Other than that it's a case of following logic level signals
and strategic measuring of voltages. Like does the fuser
lamp come on at start up, if not why? Does it start to
cycle?

The fuser draws its power from the AC PSU, right? I doubt that's the
problem though, particularly when both fans are dead.

Thanks for the reply, mon!

 

[GanjaTron]

as old as this thing is, and the heat over the years, it wouldnt
surprise me

that the radial electrolytics in the power supply is leaking or

dried.

The caps are definitely suspect, although they don't appear to be
leaky. I guess I'd have to get one of those fancy ESR meters. I've seen
a tiny cap develop a thermal fault in my Grundig VCR without leaking.
Took me ages to figure that one out. On the other hand, I've had 1/2
dozen leak all over an HP9000 apollo PSU, destroying the labelling on
board in the process! That was a no brainer (duh). In short -- caps
SUCK!

 

[ato_zee]

The fuser draws its power from the AC PSU, right? I doubt that's the
problem though, particularly when both fans are dead.

Yes, fuser (in mine) draws power from AC PSU. There is a power
board and a logic board in the AC PSU (Canon SX), and the DC PSU
feeds it with assorted signals. The logic board has a fattish resistor
of 1K2 driving the power boards opto-coupler, the one I added an
extra LED in series. Opto drives triac, with the extra red LED,
LED goes on/off, triac goes on/off, fuser lamp goes on/off. Now
it's easier to trouble shoot. AC PSU also has a relay in series
with mains supply to triac. I found I could short it out for testing,
so I put a domestic lamp in place of the fuser lamp, set it up
on a table, and drove it with a 9V battery and resistor, for testing.
Saved installing and uninstalling the PSU all the time.

DC PSU has opto couplers, there is an isolated switch mode,
raw AC in. It's txfmr drives all the rectifiers, logic etc, there
are output V level sensors, and their o/p feeds back to the
isolated switch mode. So we have a mains side, and
an isolated logic side.

 

[GanjaTron]

DC PSU has opto couplers, there is an isolated switch mode,
raw AC in. It's txfmr drives all the rectifiers, logic etc, there
are output V level sensors, and their o/p feeds back to the
isolated switch mode. So we have a mains side, and
an isolated logic side.

I can't for the life of me find any optocouplers on that board! Are we
talking about the same module? Mine's an RG1-1395, manufactured by
Sanken (presumably Japanese for "it sucks"). Sits in the front right
corner, right behind the control panel (which, as I said, is alive and
well).

Fuser glows alright, but not long enuff to cycle -- power now dies
after 15 sec or so. Anyway, I doubt a fuser prob would kill all the
fans and the main motor (which are driven by the DC PSU). I've already
had the "50 SERVICE" thing, and even then the fans still hummed away,
as this was a fault in the AC module.

Considering the page counter on this printer just reads a pathetically
low 23000 pages or so, this thing must've been a complete dud from day
one! No wonder it was sitting in the lab's basement for so long before
getting tossed out! :^)

--GanjaTron

 

[ato_zee]

My DC PSU is Sanken, tracing through it I found the raw
mains AC (from the AC PSU) was rectified, smoothed, and
used to drive a switchmode, with multiple output windings
on its switchmode txfmr. Same principle as late TV's with
switchmode supplies. The rectified output from the DC side
is fed back through opto-couplers to alter the switchmode
duty cycle, and hence regulate the DC logic side. I found
2 opto-couplers in mine. Possibly mine isn't the same
model as yours, can't get my PSU number without
pulling printer apart, but I would expect, as its from the
same era and company, to be similar.

From what you say, control panel still stays working, but
not the motors, suggests that the switchmode is still
running (otherwise how is the contol panel being supplied?)
yet the motors DC supply isn't there, since they stop.

Fuser seems ok, it lights up and goes out, so its supply,
opto-coupler, triac probably ok.

Start up (in mine) seems to inolve the fuser thermistor,
if it doesn't see the fuser getting hot it shuts down.
There are also several protective paper path sensors,
any one of which can create a fault condition.

You really need a schematic (or at least a block
schematic) so you can find and monitor the DC levels
at where they plug into the chassis.

 

[GanjaTron]

I found
2 opto-couplers in mine. Possibly mine isn't the same
model as yours, can't get my PSU number without
pulling printer apart, but I would expect, as its from the
same era and company, to be similar.

It's a Sanken PS10059K. I now remember reading something about
optocouplers for regulation in Sam Goldwasser's SMPS FAQ. I think the
two optocouplers you mention are 2 TLP634 6-pin DIP ICs sitting side by
side between the primary and secondary sides. But how do I test them???

From what you say, control panel still stays working, but
not the motors, suggests that the switchmode is still
running (otherwise how is the contol panel being supplied?)
yet the motors DC supply isn't there, since they stop.

Actually, the PSU emits a very faint high pitched whine when the fans &
motor stop, so apparently it shuts down partially.

You really need a schematic (or at least a block
schematic) so you can find and monitor the DC levels
at where they plug into the chassis.

All I have is the block diagram in the service manual -- no schematics.
I've monitored the voltages, and it's definitely the DC PSU, but I
can't figure out the exact component.

--GanjaTron

 

[ato_zee]

It's a Sanken PS10059K. I now remember reading something about
optocouplers for regulation in Sam Goldwasser's SMPS FAQ. I think the
two optocouplers you mention are 2 TLP634 6-pin DIP ICs sitting side by
side between the primary and secondary sides. But how do I test them???

From what you say, I suspect that the high pitched whine, mentioned
below suggests a DC PSU overload. Shorted or partially shorted
DC rail. So I'd keep the opto's as suspects, but look elsewhere first.
It's hard to test opto's in situ, took mine out, and put in 6-pin sockets.
Then you can bench test them, energise with 9V battery, 1k in series,
on the LED side. Try another LED, same 9V with another 1k on the
other side. Apply 9mA to opto side should turn output side LED on,
and it should stay latched on, until you disconnect the battery, and
start again. If you desolder carefully you can probably put it back
(which is why I used a skt) afterwards.

Actually, the PSU emits a very faint high pitched whine when the fans &
motor stop, so apparently it shuts down partially.

Several approaches to testing switchmodes. One is a wind up Variac.
The other that has worked many times for me is to have a battery
of domestic lamp skts, in parallel, in series with the mains.
Then plugging in lamps lets increasing amounts of current through.
With luck you can get the PSU to stay running, at reduced
input voltage, while you poke around with a meter. I've got
the holders on a board, mains lead in and out.

Sometimes in the absence of a full schematic I have to trace
PCB tracks/components, and draw it. Of course it may not
be in the motor supply line, but an overload somewhere else.
With switchmodes, with shorted rails, I often break/disconnect
one rail at a time to see if it stays up. Also do rail to ground
resistance checks. You can also, for low voltage rails, eg 12V,
wire auto bulbs in series, you can find 5W, 8W, 10W, up to
about 50W at 12V as appropriate, they are ok for 5V rails
as well. The one in series with the shorted rail comes pretty
well full on, but the switchmode should stay running. Once you
stop the switchmode going into spasm and shutting down
you stand a chance of diagnosing which rail is causing the
problem, then by splitting the circuit down (by lifting one leg
of components) just where the problem is. Diagnosing
switchmodes has a learning curve, then you get to do it
with some confidence.

 

[Dani]

I think everyone got distracted and ignored the "55 error" in the post.
In an HPIII a 55 error is a communications error between the DC
controller and the Formatter board. Both need to be replaced. If you go
to fixyourownprinter.com and type HPIII 55 error in the search box,
you'll see the info. There is also a service manual available for free
download. If you so decide, you can get replacement parts as well.

 

[Ben Myers]

Yes, rather than taking the time to troubleshoot and repair components, REPLACE
them !!!! Even when this printer was HP's top-of-the-line, HP techs never
repaired individual components.

But if you like challenges, have at it. I have enough challenges in my life
without attempting board repairs... Ben Myers

 

[GanjaTron]

Well, I've recapped the PSU... no dice. :^(

--GanjaTron

 

[CJT]

Well, I've recapped the PSU... no dice. :^(

--GanjaTron

The simplest solution would probably be to go to your favorite computer
graveyard and pick up a "parts" machine (or two), then replace a board
at a time.

 

[Ben Myers]

So stated by other postings in this thread... Ben Myers

 

[CJT]

So stated by other postings in this thread... Ben Myers

Sorry ... guess I came in late.