LX200: DEC Motor Study
Howard C. Anderson
>drives uncontrollably in one direction.. then stops and starts shuddering..
>blew the main fuse into the scope..
>
Several months ago, I bought a second DEC motor to see if by "luck of the draw"
I could get one with less backlash. Let's call the original that came with
the scope "A" and the second one I bought "B". "A" has always worked fine and
still does but I had hoped to reduce the backlash which was approximately
6 seconds of delay.
DEC motor B arrived with 12 seconds backlash. Investigation of its internals
revealed that one of the plastic gears did not fit tightly on its shaft so
that was a major source of delay after motor reversal.
I was able to shim it with a tiny piece of box-cutter razor blade that I used
as a wedge. (You can break and shape such a razor blade with two pair
of needle-nose pliers.) I didn't figure out how to shim it properly until
after it failed however.
The failure of "B" occurred out in the desert at a dark site. I was having
a lot of trouble aligning. I use the Polar iterative alignment method where one
slews back and forth from polaris to the other alignment star adjusting
the wedge controls on Polaris and adjusting electronically and synching on
the other alignment star. The initial failure symptom was that the scope
was slewing past Polaris into a "stop" of some sort.
Later at home, after disassembly and multiple tests, the failure grew more
obvious, i.e., the motor would run uncontrollably. Or the motor would run
then gradually slow down then quiver, etc. I then switched back and forth
between motor "A" and motor "B". Motor "A" always worked fine and motor "B"
always failed. I assumed that this meant that the problem was confined
to the DEC motor and its tiny circuit board.
Investigation of the circuit board revealed two active semi-conductor components
which I believe I have identified as a three-pin LM78l05 5 Volt voltage
regulator
and a 14-pin DIP LM2901N low power low offset voltage quad comparator. I
ordered
three of each from Radio Shack and am awaiting delivery.
I was able to locate the National data sheets on both components. During
testing,
I noticed that the voltage regulator on both boards was running very hot. Could
almost burn your finger. The data sheet says that if the junction temperature
is excessive, the thermal shutdown circuit will take over, preventing the
IC from overheating. This seems consistent to me with the observed slow-down
during slewing, i.e., the voltage regulator heats up, begins to go into
thermal shutdown, the comparators (which are sensing the pulses from the
internal
LEDs) begin to fail and the motor control circuitry on the main board begins to
receive puzzling signals which in turn causes the motor to change speeds?
So I have tentatively identified the problem with DEC motor "B" as a problem
with
the LM78105 voltage regulator. There are versions of this device that are
in plastic
housings and other versions that have metal housings. The metal cans have
better
heat sink capabilities of course. The IC on the board is in a plastic housing.
I am wondering of course if the change from 12 Volts to 18 Volts might have
stressed
the voltage regulator.
This is all speculation of course. I will let you know if the problem is
resolved
by replacement of one or both of these ICs.
Thanks,
Howard
Thanks,
Howard
Robert Preston
Howard Anderson's analysis stands a good chance of putting us all ahead
of the DEC failure problem...I can hardly wait to see the effects of
his upgrading his LM78105 DEC miniboard regulator...(but mine says
"78L05" JRC not "78105", and it's clearly a capital "L"...is that a
misprint on the labeling for chrissake??? and the reason why I gave
up on reverse-engineering this bloody thing???)
Thank you, Howard, for the detective work and keen insights.
(Albert Schuitema said something about a regulator being extremely hot
on his scope...was that the same regulator IC, Albert?)
By the way, I've been operating my scope routinely with the max slew
speed set to "6" on the "telescope" menu, ever since I noticed how
much arcing there is in the DEC motor commutator at the full speed
slew. That may or may not help the 78105 regulator situation, but it
certainly would help with any possible heat problem in the L2724
motor drivers...those too might not appreciate the 12V to 18V switch
that was engineered by Meade.
Rob Preston
Pittsburgh PA
ra...@med.pitt.edu
Frank Loch
drive circuits yourself rather than use Mead replacements. Maybe you
would like to explain here.
Frank Loch -- fl...@prolog.net
Len Rose
The only thing unusual I have been doing lately is using a very slow slew speed
because a neighbor complained about the slewing noises at 3 am.. I was using
the standard power supply since I was at home. I am really sorry I didn't order
the spares kit just to keep on hand.
Len
W. Miller
Howard, here are some of my notes on the LX200 drive circuits.
Hope they are of help.
As to why some of us do this type of reverse engineering and not just buy
replacement parts from Meade, I can only speak for myself:
1. I have the time and enjoy learning what makes things work in my hobbies.
2. I like to save money (so did my ancestors), and my warranty has expired.
3. I dream of improving on the robustness or performance of our favorite scope.
Has anyone observed waveforms coming out of pins 1 & 2 of the L2901N? It would
be useful to know how the dual photodiode sensor circuits are utilized and what
is an optimum duty cycle and phase coming out of their respective comparator
(on small board, measure from top of each 27Kohm resistor to ground gold pin).
These are easy to access test points. The variables seem to be the position of
the LED mask in the gearbox, and setting of the blue 20Kohm potentiometers.
I used a volt ohm meter to fine tune the pots on my new RA drive assembly to
the following: 3.1v DC, 2.3v AC output from the comparators; 2.14v DC, 1.04v AC
output from the IR detectors; and 325 ohms on the pots. I did not have an
oscilloscope to view the waveforms. Just a 1/10th turn on the pots will make
a big change. Measure the resistance before you turn the pot. It ranged from
274 to 865 ohms on 3 drive units.
--------------------------------------------------------------------------------
This is the drive unit board which detects flashes from the Light Emitting
Diode shining through a 90 slot rotating strobe disk mounted on motor shaft:
New Japan Radio Co, 2901N, 14 pin dual inline package
Quad single supply comparator (2 comparators are used) pin3=18v, pin12=GND.
The circuit function is two inverting comparators with hysteresis.
When input pins 4 or 6 go >2.50v, output pins 2 or 1 respectively go to 5.0v.
When input pins 4 or 6 go <2.11v, output pins 2 or 1 respectively go to 0v.
Photo detector bias current is adjusted by turning the 20Kohm potentiometers.
The RA unit 2901N output pins 1 & 2 go to input pins 5 & 9 respectively of U13
a SGS-Thomson T74LS14B1 hex Schmitt trigger (hysterisis trips at 0.8v & 1.6v)
located on the main board, then out of U13 pins 6 & 8, over to U12 pins 6 & 7
a XYLINX XC3030-70PC68C field programmable gate array. The RA and Dec boards
have identical circuitry, except the RA also contains a switch located on the
worm that sends 18v/GND signal to pin 9 of U11 another T74LS14B1 located on
the main board.
5v >--------o-------------------o-----o-----| |<--- 2901N ---->|
regulator | | | |
78L05 | | | 10Kohm
56ohm 20Kpot<--| | |-------o-----27Kohm-----|
| | | | 10Kohm | 5 |
| Strobe Disk | | | | |----| + 2 | Black
| | | | | GND | Comp >----o-----
| V |-------o-----------------------------| -
LED ~~ DET | | 4
| |--GND | 10Kohm
LED ~~ DET | |-------------o----27Kohm------|
| | 20Kpot<--| 10Kohm | 7 |
| | | | |----| + 1 | Blue
| | | GND | Comp >----o-----
GND |-------o-----------------------------| -
6
--------------------------------------------------------------------------------
I have not yet determined the purpose of the duality of this control circuit.
The 2 groups of slots in the detectors' mask do line up together with the
slots in the strobe disk. Perhaps it is for redundancy and reliability.
Bill Miller, Los Gatos, CA
Robert Preston
>The 2 groups of slots in the detectors' mask do line up together with the
>slots in the strobe disk. Perhaps it is for redundancy and reliability.
Thanks for the analysis, Bill. I was thinking the dual LED control had to
do with forward vs. reverse motor rotation, assuming some sort of Moire
effect between the two mask slot groups would give the direction.
There would almost have to be some such Moire effect since the mask would have
to be inserted at very precisely the right depth in order to *avoid* it (since
these slots are on a very small radius, a very slight radial displacement gives
a different slot spacing for mask vs. wheel). This is speculation. I have not
systematically tested the effect of the mask position on any voltage or on
the scope performance (I just stick the thing in there and cross my fingers:
so far that has worked). On the other hand, it could also be a redundancy
thing, or something else, for all I know.
W. Miller
Reply to alb...@pi.net 21APR96
>and the other comparator IC is a LM339N (not LM2901N)!!
Albert,
These are all essentially the same functionally
LM139/LM239/LM339/LM2901/LM3302
Low Power Low Offset Voltage Quad Comparator
as per a National Semiconductor Datasheet.
Then there is the 2901 on my board, same thing but
built by New Japan Radio Co, Ltd (JRC).
Your LM339 just made the better grade on input offset voltage 5mv
vs 7mv or voltage gain 50 vs 25 when it was tested by the supplier.
Bill Miller, Los Gatos CA
Brandon Jones
> I have not yet determined the purpose of the duality of this control circuit.
> The 2 groups of slots in the detectors' mask do line up together with the
> slots in the strobe disk. Perhaps it is for redundancy and reliability.
>
Even though they use two sets of photo diodes I beleive they do not use
them for what they are usually used for ie to determine motor
direction.This is usually by putting the squared up out of phase sine waves into
a bistable the output of which would change high or low depending
on whether the motor was going forwards or backwards.When I did the
same as you to fix a similar fault on my Lx I only found one TTL
level square wave going back to the scope which is all you would need
to determine relative position. Perhaps I missed out on something.
Brandon S.Jones
Howard C. Anderson
previous
transmission. Further study leads me to believe that the voltage regulator is
used to current-limit the LEDs and does not regulate the LM2901N. The LM2901N
seems to run directly from the 18 volt line.
I have concluded that the traces on the tiny board are extremely small and
desoldering cannot be achieved with any hope of success - even with a
desoldering
iron with bulb to suck the solder from the connection. Anyone attempting to
replace parts should be prepared to rewire the board with external wires because
the traces have a high probability of breaking. Overall, I believe it to be
impractical. I guess the only thing I am relatively sure of is that the failure
is on that tiny circuit board but doubt that I will be able to proceed further
with verifying precisely which component is failing.
Thanks,
Howard
Thanks,
Howard
Pedro Breda Santos
I have been very busy lately and so have no the time to follow closely
this group. Sometimes like today I give a spy, and today I saw this Dec
motor discussion. The problems I think are being mentioned are
progressive loss of control of the Dec motor, strting to move alone,
etc. This was the problem I had some months ago. At that time I wrote
a post telling the whole story (you can check the archives).
Anyway I can tell you that the problem is in the semi-variable resistor
in the pseudo-controller board. this resitor deines the comparation
level to activate the comparator you mention. Try to adjust this
resitor. Or do beetr and read my previous post.
About the existence of 2 photodiodes I think it is ment to make sure of
the movement direction. For example, with only one detector, if the
wheel was in a intermitent position it coulb be read as a continuous
move.
Best
Pedro Breda Hi people.
I have been very busy lately and so have no the time to follow closely
this group. Sometimes like today I give a spy, and today I saw this Dec
motor discussion. The problems I think are being mentioned are
progressive loss of control of the Dec motor, strting to move alone,
etc. This was the problem I had some months ago. At that time I wrote
a post telling the whole story (you can check the archives).
Anyway I can tell you that the problem is in the semi-variable resistor
Best
Pedro Breda Santos
Martin Tom Brown
> This is the drive unit board which detects flashes from the Light Emitting
> Diode shining through a 90 slot rotating strobe disk mounted on motor shaft:
> I have not yet determined the purpose of the duality of this control circuit.
> The 2 groups of slots in the detectors' mask do line up together with the
> slots in the strobe disk. Perhaps it is for redundancy and reliability.
I would hazard a guess that it has the two diode/detector pairs at
a half pitch difference relative to the rotating mask. This enables
the feedback to provide 2x resolution of the mask and direction info.
So I'd expect the two waveforms to look something like:
________ ________
Channel A ____| |_______| |_____
________ ________
Channel B ________| |_______|
Rotating forwards gives the sequence 0,0 0,1 1,1 1,0
backwards 0,0 1,0 1,1 0,1
This is of course pure speculation on my part,
I have no intention of measuring it on my LX200
Regards,
--
Martin Brown <mar...@nezumi.demon.co.uk> __ CIS: 71651,470
Scientific Software Consultancy /^,,)__/
Martin Tom Brown
> Investigation of the circuit board revealed two active semi-conductor components
> which I believe I have identified as a three-pin LM78l05 5 Volt voltage
> regulator
78L05 regulator seems plausible rated to 100mA and 360mW nominal power.
A quick back of the envelope calculation suggests that if this circuit
draws more than 30mA at 18v it may cause the regulator to overheat.
AFAIR the output voltage will drop rapidly if this happens.
How much current does the circuit actually draw?
20mA would appear to go down the LED+LED+56R chain for starters...
> and a 14-pin DIP LM2901N low power low offset voltage quad comparator. I
> ordered three of each from Radio Shack and am awaiting delivery.
Seems excessive to me - it's virtually impossible to damage a 78 series
voltage regulator without doing something really horrible to it.
If they do fail permanently they go open or short circuit raw supply
to ground or to load (the latter being potentially more destructive).
Both are easy to check with a DVM.
> I noticed that the voltage regulator on both boards was running very hot. Could
> almost burn your finger. The data sheet says that if the junction temperature
> is excessive, the thermal shutdown circuit will take over, preventing the
> IC from overheating.
You could always check to see if it's output voltage stays at 5 volts
during the fault condition. The other quick and dirty test would be
to give the regulator a bit of help, 0.5mm copper foil strip bent round
a pencil leaving a short tab stuck out makes a fair clip on heatsink.
(You can also buy proper clip on heatsinks for the purpose)
Be careful not to short any other components or leads!
Howard C. Anderson
Just simple curiosity. The Meade replacements are the only practical
way to go. I thought that if the problem could be identified and
corrected with heftier components perhaps, then no future failures
would ever occur. By the way, I love my Meade telescope. I think
Meade did a superb job overall.
Thanks,
Howard
Thanks,
Howard
Gerry Reynolds
>I have concluded that the traces on the tiny board are extremely small and
>desoldering cannot be achieved with any hope of success - even with a
>desoldering
>iron with bulb to suck the solder from the connection. Anyone attempting to
>replace parts should be prepared to rewire the board with external wires because
>the traces have a high probability of breaking. Overall, I believe it to be
>impractical. I guess the only thing I am relatively sure of is that the failure
>is on that tiny circuit board but doubt that I will be able to proceed further
>with verifying precisely which component is failing.
>
>
Why don't you try solder wick instead of a solder sucker. You will be less prone
to damage the board. Also too hot of an iron will damage the traces. If this is
a thru hole component, you may want to wiggle each component leg individually to
break loose (after wicking up excess) before trying to remove component from
board.
Gerry Reynolds
ge...@fc.hp.com
Robert Preston
>78L05 regulator seems plausible rated to 100mA and 360mW nominal power.
>A quick back of the envelope calculation suggests that if this circuit
>draws more than 30mA at 18v it may cause the regulator to overheat.
>AFAIR the output voltage will drop rapidly if this happens.
>How much current does the circuit actually draw?
I measured the drop across the 56 ohm resistor (in series with the LEDs)
at 2.6 volts. Thus, the current from the 78L05 is 2.6V/56ohm=0.046amp, since
the other loads on the 78L05 are negligible compared to the LED branch.
That's 50% "more than 30mA", so it looks like this thing is overloaded.
The output from my 78L05 is 5.3 volts, the input is 17.9 volts (I'm
running the scope on the Meade 120VAC/18VDC power supply) so the
78L05 drops 17.9-5.3=12.6 volts. Thus the power expended in the 78L05 is
12.6V x 0.046amp = 0.58 watt=580mW. If Martin's "360mW nominal power"
means 360mW maximum power, then the 78L05 is surviving by virtue of its
internal protective circuitry, which may put the dec. circuits at risk.
Despite having had two dec. motor replacements since 1994, this
scope has been operating well for months, despite the hot 78L05. However,
the number of dec. failures on this list suggests that there's something
marginal about the system, and it seems to me that these hot ICs are a clue.
It seems to me that either the scope should be operated at something less
than 18 volts, or these regulators should have heatsinks added to them (or
both). The lower voltage seems preferable, since that would also allow the
L2724 motor driver ICs (and the other 5V regulators on the motherboard) to
operate with more conservative power expenditures.
I thought that limiting the slew speed to "6" in the telescope menu would
be a conservative move. However, on further thought, while that may be true
for the motors themselves (and their commutators), it may mean that the heat
dissipation in the L2724s would actually increase because they would sustain
a larger voltage drop during slews at speed 6 than at speed 8.
Comments on these thoughts would be appreciated.
Robert Preston
dec. axis miniboard showed it was dissipating 580mW when the scope was run
on the Meade 120VAC/18VDC supply. Based on Martin Brown's message which
*suggested* a 360mW limit for the 78L05, I thought that the 580mW might be
excessive. However, the specs for this IC in Horowitz and Hill state that
600mW is the maximum power allowed. The National Semiconductor datasheet
for the IC does not give a fixed value for the max. power but says that
internal protection limits the power, and then it goes into two pages of
very nice discussion about the fine points of heat-sinking.
Bottom line: I think the lx200 circuit design around the 78L05 is probably
adequate to the task. The 78L05 runs very hot to the touch, but that in
itself is not necessarily a problem, since it can handle a lot of heat.
Personally, though, I'm going to add a regulator to the Meade supply to run
my scope on 15 volts rather than 18, as an added safety factor, in case
overheated ICs might be *contributing* to some complicated destructive oscil-
lation.
As to the actual culprit(s) in dec. axis failures, I'm still mystified.
As ric suggested in a recent message, these circuits are quite a bit more
complicated than a model T Ford, so the cause(s) are very difficult to
isolate, probably for Meade engineers as well as for us amateurs. Meanwhile
I need to say that my lx200 has been operating flawlessly (although not
frequently, during the winter months) and it remains the best investment
I ever made, bar none.
Rob Preston
ra...@med.pitt.edu
Martin Tom Brown
> Martin Tom Brown (Mar...@nezumi.demon.co.uk) wrote:
> >78L05 regulator seems plausible rated to 100mA and 360mW nominal power.
> >A quick back of the envelope calculation suggests that if this circuit
> >draws more than 30mA at 18v it may cause the regulator to overheat.
360 mW is from memory for a TO92 package without heatsink, and
probably includes a design safety margin. Don't have refs to hand.
Spec is strictly <100mA and temperature <125 C operating range.
Note anything >70C will be too hot to touch.
> I measured the drop across the 56 ohm resistor (in series with the LEDs)
> at 2.6 volts. Thus, the current from the 78L05 is 2.6V/56ohm=0.046amp, since
> the other loads on the 78L05 are negligible compared to the LED branch.
According to ????'s sketch circuit there are *two* LED's in the chain
so I'm surprised how large the voltage drop is across the 56R.
> That's 50% "more than 30mA", so it looks like this thing is overloaded.
> The output from my 78L05 is 5.3 volts,
56R
5.3v ---/\/\/\-----o------>|-----o----->|-------- OV
^ 2.7v ^ W
This leaves me puzzled about the voltage drops across the LEDs.
Even if they are IR LEDs I'd have expected a forward voltage of
at least 1.6v each - I wonder if one of the LED's is dodgy?
Can you easily measure the point W to see if the LED's are
taking equal shares of the voltage drop?
> means 360mW maximum power, then the 78L05 is surviving by virtue of its
> internal protective circuitry, which may put the dec. circuits at risk.
100 mA or overheating are the hard limits. You can run it at higher power
simply by placing the case in contact with a lump of metal or heatsink.
Hence the copper foil clip suggestion.
> Despite having had two dec. motor replacements since 1994, this
> scope has been operating well for months, despite the hot 78L05.
If it gives 5 ish volts output then its working OK.
Can you reproduce the motor fault on the bench where you can measure
voltages for good and bad motor specimens - that might shed some light.
Good luck,
Robert Preston
>This leaves me puzzled about the voltage drops across the LEDs.
>Even if they are IR LEDs I'd have expected a forward voltage of
>at least 1.6v each - I wonder if one of the LED's is dodgy?
>Can you easily measure the point W to see if the LED's are
>taking equal shares of the voltage drop?
Well, tonight the 78L05 is putting out 5.0 volts, not 5.3, which
seems weird; anyway the voltages above the diode 1, diode 2, and
resistor are, in order, 1.2, 2.4, and 5.0 volts. Guess these are
low drop LEDs or something...I don't know much about this subject.
Usage: averages around 7 hours per week, with wide variations.
Pgo...@aol.com
>> >78L05 regulator seems plausible rated to 100mA and 360mW nominal power.
>> >A quick back of the envelope calculation suggests that if this circuit
>> >draws more than 30mA at 18v it may cause the regulator to overheat.
>
>360 mW is from memory for a TO92 package without heatsink, and
>probably includes a design safety margin. Don't have refs to hand.
>Spec is strictly <100mA and temperature <125 C operating range.
>Note anything >70C will be too hot to touch.
>
According to the National Voltage Regulator Handbook, the 78L05 will hit
thermal shutdown at about 800 mW in free air at 20 degrees C ambient.
>> I measured the drop across the 56 ohm resistor (in series with the LEDs)
>> at 2.6 volts. Thus, the current from the 78L05 is 2.6V/56ohm=0.046amp,
If this represents the output load of the 78L05, that would be 624 mW at an
input of 18V. Close? Maybe at elevated ambients.
Paul Goelz
Evanston, IL
Gerry Reynolds
>According to the National Voltage Regulator Handbook, the 78L05 will hit
>thermal shutdown at about 800 mW in free air at 20 degrees C ambient.
>
>>> I measured the drop across the 56 ohm resistor (in series with the LEDs)
>>> at 2.6 volts. Thus, the current from the 78L05 is 2.6V/56ohm=0.046amp,
>
>If this represents the output load of the 78L05, that would be 624 mW at an
>input of 18V. Close? Maybe at elevated ambients.
>
>Paul Goelz
>Evanston, IL
>
Was this a typical shutdown or guaranteed spec. If typical, some parts could
shutdown before 800mw is reached. If guaranteed, was it guaranteed to shutdown
by 800mw (meaning that it could shutdown before 800mw) or was it guaranteed to
shutdown at not less than 800mw.
Also, one needs to consider the variation in the load and thus the actual power
dissipated. Sounds scary to me, if your numbers are correct.
Gerry Reynolds
ge...@fc.hp.com
Pgo...@aol.com
>>If this represents the output load of the 78L05, that would be 624 mW at an
>>input of 18V. Close? Maybe at elevated ambients.
>>
>>Paul Goelz
>>Evanston, IL
>>
>
>Was this a typical shutdown or guaranteed spec. If typical, some parts could
>shutdown before 800mw is reached. If guaranteed, was it guaranteed to
>shutdown
>by 800mw (meaning that it could shutdown before 800mw) or was it guaranteed
>to
>shutdown at not less than 800mw.
>
The figure was taken from the graph representing the maximum dissipation. I
assume they set the maximum dissipation to be the point where thermal
shutdown occurs. Maybe a bad assumption? I dunno. In any case, if these
are correct current levels, it sounds like a more marginal design than I
would be comfortable with. I don't know the ramifications of the regulator
dropping out, but I would imagine it would cause the circuit to overcorrect
as it missed pulses?
Paul Goelz
Robert Preston
>....I don't know the ramifications of the regulator
>dropping out, but I would imagine it would cause the circuit to overcorrect
>as it missed pulses?
>Paul Goelz
I don't know either, but one possible scenario (I haven't studied the
circuit closely enough to be sure of this) would be failure to detect
ANY encoder pulses, so the motors would "overcorrect" to the max by running at
full tilt, "looking for" the next pulse (and that will never arrive).
Drives turned uncontrollably ON is a symptom that someone posted here, isn't it?
Rob Preston
ra...@med.pitt.edu
Pgo...@aol.com
>I don't know either, but one possible scenario (I haven't studied the
>circuit closely enough to be sure of this) would be failure to detect
>ANY encoder pulses, so the motors would "overcorrect" to the max by running
>at
>full tilt, "looking for" the next pulse (and that will never arrive).
>Drives turned uncontrollably ON is a symptom that someone posted here, isn't
>it?
>
>Rob Preston
>ra...@med.pitt.edu
>
>
>
>
If I were designing it, I would add a routine to compare the time the motor
drive had been on to the arrival of a pulse. Too long on without a pulse
would tell me to shut it down and sulk. I hope Meade did something like
that.
Paul Goelz
Evanston, IL