Re: TECH:Rottendog Data East Power Supply DPS004 + older DMD

On Mar 14, 12:06 pm, Elvis <jc...@xxxxxxxxx> wrote:
On Mar 14, 9:50 am, "GPE" <GPE_NoS...@xxxxxxx> wrote:





"Elvis" <jc...@xxxxxxxxx> wrote in message

news:6e3e14ae-b8e2-4a65-8611-495757417c65@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

On Mar 13, 7:31 pm, donw <dweingar...@xxxxxxxxxxxxxxxx> wrote:
On Mar 13, 4:30 pm, Elvis <jc...@xxxxxxxxx> wrote:

On Mar 13, 1:38 pm, donw <dweingar...@xxxxxxxxxxxxxxxx> wrote:
On Mar 12, 9:36 pm, "GPE" <GPE_NoS...@xxxxxxx> wrote:
"chuck" <ch...@xxxxxxxxxxx> wrote in message
news:b5382c36-ae0b-4be7-a2f0-44c2d6d59fa4@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

I picked up a GnR pin with a rottendog power supply board and a
non-
workingdmd.  The power supply board had a blown zener diode and
lm317
in the 68 volt section.  I replaced the diode with one Jim sent
me.  I
put theDMDin another machine and it worked fine.  I tested the
power
output and while not spot on it was close. Fuse blows, diode
blows,
lm317 blows.  Replaced everything again.  Double check
everything, the
smell of burning electronics again.  Ok, I'm not going to blow a
third
set of components.  Check everything again.  All seems to be
working.
Decided to consult with Don (pb2k team) and he suggested I check
the
ohms on the display.  Sure enough 68v to ground was 38 ohms which
drew
too much current and blew the section.

It seems the original wms board is current limted and Rottendog's
board is not.  Where this really comes into play is with older
displays that may be on their last leg.  Be careful when using
this
power suppy with any older display.  A new display should work
just
fine with it.

LM317AHV is rated 1.5amp. Your display was trying to draw (68/38)
or nearly
1.8 amps.  Yep, I can see why the LM317AHV was having problems.
But, the
LM317AHV *IS* internally current limited and will shut itself down
with
excessive current draw.  There is another problem with the LM317AHV
design
that has been mentioned previously.  What you are seeing is EXACTLY
the type
of failure that was mentioned in prior posts  ... but to stay out
of an
arguement, I won't get into this.

- Ed- Hide quoted text -

- Show quoted text -

For whatever reason, the display was causing a load of 5 ohms on the
output of the LM317.  ON power up, the LM317 would be in current
limit.  The voltage on the output of the LM317 will be 1.5 volts.
The
voltage across the LM317 will be 100*1.414 - 44 - .7 -1.5 = 95volts.
The datasheets max voltage is 60 volts.  Goodbye LM317!

Replacing one of the zener diodes with a current limiting power
resistor would protect this circuit in cases where the display is
dying.- Hide quoted text -

- Show quoted text -

I have a similar issue with aR&Bbut with a brand newDMD.
I'm on my second set of diodes and resistors, but the LM317 seems to
be ok.
It starts off withflickeringDMDduring high use modes like tri-ball,
then eventually just stops working.

Quote: "Sure enough 68v to ground was 38 ohms which drew too much
current and blew the section."
How can I check this for myDMD?

We disconnected theDMDand measured from the 68v pin to the ground
pin on thedmd.  Then compared the measurement to a known good one.
Not sure which was damaged first, thedmdor the regulatore but the
result is the same.- Hide quoted text -

- Show quoted text -

Just wondering, since I don't know, would the resistance with theDMD
unplugged (inactive) be different from an active display.
In other words, can this type of resistance measurement from supply
pin to GND be directly translated to the actual operating current?

Have you had any issues with the -100V or -110V supplies?
I was checking the data*** for LM337 and it says the output voltage
is adjustable between -1.2V and -37V.
Are these being driven out of spec to provide the -100V and -110V?

If so, there must not be another appropriate TO-220 form factor
regulator available or it probably would have been the first choice.

For the LM337 -- this is actually more likely to cause part failure in case
of power supply short.
The absolute worst case voltage differential for the standard LM337 is 40
volts whereas the LM317AHV was 60 volts.  A short on the negative supply
would put greater than a 100V differential across the regulator.

I believe a better choice for the HV positive supply would have been the
Texas Instruments TL783CKCSE3.  This part can withstand a differential of
125 volts.  Not a drop in replacement - I believe you need to revise your
voltage divider resistors - but a very close replacement.  Problem with this
part is that nobody stocks it.  You need to buy these 500 at a time.
Unfortunately, I don't see a product like this for the negative side.

Best ways around this but aren't practical for install on this board -- 1:
use series pass transistor type regulator.  Very tough regulators and have
been around for decades.  2: Use a switching stepup regulator.  Nice supply
and can run from your 12V supply.  Pascal uses one on his 60V supply for his
all-in-one Gottlieb board.  3: Use a switching stepdown regulator.  Again,
hard to find a fault with these other than cost.
When I do HV supplies - I tend to use the series pass transistor type
regulators.  If you beef them up enough, their bullet proof.
The LM317AHV and LM337's are actually series pass transistor type regulators
internally... they just don't use high voltage components.

-- Ed- Hide quoted text -

- Show quoted text -

So are you saying that when data*** says output voltage adjustable
from -1.2V to -37V, this is really base on a voltage differential from
the input voltage and not an absolute output max? So driving to -110V
could be ok if the input is close enough to output?

I did see a LM337HV, perhaps that would be a little better choice.- Hide quoted text -

- Show quoted text -

How did you solve this, or did you? I'm having the exact same problem
now. Flickering.
.