How to test a 301-07855 AMI-ROWE Memory Unit?
CoindoorDave
know - that usually means non-working...but I'll play the optimist for
now). Is there any way I can test this in my R-81? Is it "pin
compatible" with the 601-07666 I have in there now? How does the
"autoplay" function? Does it play random selections, or the most
played?
The battery (fuzzy) was disconnected some time ago, and there's no
visible damage to the PCB. If this unit is working, any idea what I
should look for in a replacement battery?
Thanks.
KR
There is an excellent chance that it wont work. These units have a
high failure rate in my experience with them.
IIRC - they used 3 x 1.2v AA ni-cad cells in series. You probably
will have to use the AA cells and not lithium coin cell etc, as this
circuit used a couple of stages of voltage doubing to get the voltage
up near the rail voltage of the custom chip - which is typically -17 v
DC. Voltage doublers are not very efficient usually.
Note that if you dont need the selection memory function, then the
unit will work perfectly without batteries. I personally would leave
them out as they are just a constant leakage risk.
If you do put batteries in, make sure you put them the right way
round, according to the markings on the original plug and connector -
or the manual, and don't necessarily rely by trying to trace the
circuit, as the whole machine runs on negative voltages in the logic
section, which can cause confusion as to polarity.
The 2 types of memory unit are fully interchangeable, one has a
selection memory, the other one doesn't.
The autoplay function simply selects a random B (right) side record,
at the number of minutes you set, after the last played record is
finished.
20 odd years back I did manage to make one of these play the A side by
inserting a diode somewhere in the circuit (as if it played random B
sides, it would manage to find one with a scratch on it that made it
stick, annoying the location owner - and you could never find the bad
record) - note that this doesn't matter in paid use, as no one usually
plays B sides.
CoindoorDave
in plugging it in (i.e. can I test for a short that could take out
some other component? or is the worst case simply that it won't
work?) Should all three LEDs light up when working? Thanks again
Dave
KR
> Thanks for the quick reply! Aside from not working, is there any risk
> in plugging it in (i.e. can I test for a short that could take out
> some other component? or is the worst case simply that it won't
> work?) Should all three LEDs light up when working? Thanks again
>
it just won't work. Very unlikely to be a short on this board, but if
there is a short, all that will happen is the power supply will
overload and safely switch itself off.
(general note: if there is a short on the 28vac, a circuit breaker
will trip, and will need to be manually reset by pushing its button
back in, but 28vac doesn't go to the memory unit)
You will know by looking at the 4 LEDS on the power supply if this
happens, then the faulty one will be not lit.
if it shorts, remove the short and the power supply should start
again. turn mains on/off if needed.
IIRC - the memory unit DATA and CLOCK leds will only light when you
finish punching in a valid 3 digit selection number, and then both
will stay lit until all records selected have finished playing and the
mechanism has fully stopped again, waiting for the next selection to
be punched in. (actually they are pulsing on and off quickly, but to
the naked eye they appear to be fully lit).
IIRC the other LED on the memory unit is a -17v led (it will be
marked on it) and this should be on at all times that the machine is
switched on (and the service switch at the left of cabinet is turned
on too). if its not on, the fault will be in the memory unit as it has
an internal regulator that drops the -27v down to -17v for the logic.
This -17v is then supplied to the selector logic (and possibly credit
computer) through the wiring harness from the memory unit.
CoindoorDave
it plugged in without over stressing any wires...and it's playing!
The -17VDC LED came on, all four the LEDs on the PS are lit, the data
and clock LEDs on the memory unit are lit (not as brightly as the
17VDV light though)...the data light looked like it flickered as I was
punching in the numbers. The second record is coming up now....<drum
roll>.........and the Red Hot Chili Peppers are-a playin "Higher
Ground" as expected (or hoped, rather). After the next record, I
guess I'll let it sit (I'll set the auto play for 20 minutes...will
shut off power before I change it tho...just in case...
CoindoorDave
a few seconds it spun and is now playing a B-side (like you
said)...Stevie Wonder's "He's Mister Know-It-All"...perhaps it the
juke's way of honoring you for yer help! :-)
Dave
CoindoorDave
two are out...waiting patiently for the 20 minute mark
and....success!....Dave waits another 24 (approx...it starts counting
at the end of the last record, I take it)....and 10cc's "Hot to
Trot" (never even heard this one before)....awesome!
KR
it sounds like it is working as it should, and you should go out and
buy a lotto ticket - with luck like you have
just had ;). These boards have been notoriously unreliable in the past
in my experience.
Note - the LED's look dim as they are actually flickering on and off
very fast. (faster than you can see to detect any flicker) The dimness
is normal and is due to the % ratio of on-off time of the led, and is
directly related to what data is being transmitted.
For example selection 111 will probably have the least number of "on"
pulses (assuming a logic 1 = off) and something like 277 will have the
most number of "on" pulses and therefore will show maximum brightness.
in theory - in practice it might not be that noticeable, as there is
substantial "dead time" between data "byte" transmissions.
the clock led should always be the same brightness when its on, as the
clock is constant on/off signal.
If the LED on indicates a logic "0" then the opposite will apply.