ewhac@well.UUCP (Leo 'Bols Ewhac' Schwab) (08/30/87)
[ This line intentionally left blank. ]
Well, I finally gave up on the Spirit Technologies battery-backed
clock I got. I was willing to hold on to it only if it kept good time.
It didn't. It gained about ten seconds a week. In my book, that's
awful. So, back it goes.
I called Spirit Tech and told them I wanted to send it back. The
person on the other end was very nice about it, and gave me a return
merchanise authorization number. I'll throw it at UPS next week. Exactly
when I'll see my money back is another issue, but I have no reason to expect
they'll be underhanded about it.
In the meantime, my little Amy has her RF shield back, and is happy
as a clam.
To be perfectly fair, Larry Rosenman (sp?) on PeopleNet recommended
the clock to me, saying that it kept wonderful time, and that he was elated
with it. So obviously, Spirit Tech does make stuff that makes people happy.
This particular people wasn't happy - what can I say?
--------
In the time that I had the clock, I came to some decisions about
what I'd like to see in a battery-backed clock, which I'd like to share with
people out there who might be making such beasts.
It has to be at least as accurate as my watch (for the record, my
watch gains about five seconds a month), preferably more so. I wince at
anything worse than 1 second/month. (Mebbe I should buy an atomic clock?)
It should auto-config. I want to be able to reset the machine, and
during the reboot process, have the system time be set. I would prefer that
I didn't have to clutter my Startup-Sequence with Yet Another Command. My C
development startup script is long enough, thank you. It seems to me that
it should be possible to have a true auto-configuring clock, possibly on a
Zorro board or something.
Failing that, then it would be nice to SetFunction() the DOS's time
of day calls to custom code to read the clock hardware (although I can think
of a couple of possible problems with this).
The battery should last "forever". In other words, I want to see
the battery trickle-charged by system power when the system is on.
Anyone working on such a beast? Perry, I remember you working on
some such thing....
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
Leo L. Schwab -- The Guy in The Cape ihnp4!ptsfa -\
\_ -_ Bike shrunk by popular demand, dual ---> !{well,unicom}!ewhac
O----^o But it's still the only way to fly. hplabs / (pronounced "AE-wack")
"Work FOR? I don't work FOR anybody! I'm just having fun." -- The Doctorbryce@COGSCI.BERKELEY.EDU.UUCP (08/30/87)
[ We interrupt this article to bring you a special broadcast: A while back I was typing in an acticle and wrote a sentence that I felt was awkward. The problem was "he did..." I made a flipant little remark that a neuter personal pronoun should be developed and popularized so that "he" will no longer need to take the ambiguous sexist double meaning of both "he" and "she". I'm very sory I started a flame war in the wrong place. Please stop commenting on this to this group. If you want I'll collect everybody's favorite and make a final posting. I've already had about 5 sets sent to me. They range from "yah" to "see" to "gree" etc., etc.] In article <> ewhac@well.UUCP (Leo 'Bols Ewhac' Schwab) typed: > > Well, I finally gave up on the Spirit Technologies battery-backed clock I > got. I was willing to hold on to it only if it kept good time. It didn't. > It gained about ten seconds a week. In my book, that's awful. So, back it > goes. There is a solution... that clock is _software_ adjustable. Simply pull out your handy-dandy Thomson Components-Mostek MK48T02/12B-20 data sheet and read page 5. What?? You don't have one?? Well then, I'd better commit a violation of federal law and type in the relavant information for you! Here goes: The calibration register occupies [bits zero to four] in the control register {$DC0FF1 on the Spirit board}, bit five of the same byte is a sign bit, "1" indicates positive calibration {ie. faster}, "0" indicates negative calibration {ie. slower}. {Set bits six and seven to 0, or you'll screw up the clock.} $DC0FF1 - 0 0 S X X X X X That information really should have been in the user manual. :-( The main reason that this clock chip uses software calibration is that it emulates a 2K*8 bit static ram; there are no extra pins for crystals or trim pots or such. In fact, with your Spirit board you have 2040 bytes of completely useless non-volitile battery backed static ram. ---> Can *you* think of a use for 2040 bytes of battery backed ram on an Amiga? Remember the restriction that you have to load code, probably from disk, in order to read it! <--- The Spirit software has no support for setting the calibration so you'll need to yank out wack or some other memory editor. > In the meantime, my little Amy has her RF shield back, and is > happy as a clam. Having to modify/bend/mangle/remove the RF shield is reason enough to return that clock. Please tell them so. > It should auto-config. I want to be able to reset the machine, and during > the reboot process, have the system time be set. I would prefer that I > didn't have to clutter my Startup-Sequence with Yet Another Command. I agree 100%. Doing it on the other hand is another problem. It might make sense for clock reading software to be in the ROMs of the two machines with a standard hardware clock (The A500 and A2000). [Side note: Certainly the $C00000 RAM test should be able to tell the differnce between the clock and bad ram, and stop $C00000 sizing at that point.] In a more general case, you can't set the clock until the system is up, and Auto-Config happens kind of early in the boot sequence. What would be needed is a varriation of the init/diagnostic vector that would could copy from Auto-Config ROM into ram then execute *after* the timer.device, DOS, etc. is up. After all such code has been run, the initial startup-sequence would be hit. A very minor open question would be for machine-stealling programs that boot from code in the boot sector. Do you want your various configured devices to come up, or should some bit in the boot sector say "I'm going to steal the machine... don't bother configing for me." > The battery should last "forever". In other words, I want to see > the battery trickle-charged by system power when the system is on. The A501 board uses a ni-cad. Of course you'll need an Amiga 500 to use it :-). It features a much more primitive clock chip, the OKI MSM6242B. Adjustment is by trim capacitor. Works ok, however. To replace the battery on the Spirit chip would be an adventure. It's sealed as part of the chip's "Top Hat" along with the crystal. Even though it's backing up 2K of static ram in addition to the clock that battery should last somewhere in the range of 10-30 years This "top hat" is also 1/2 reason that the Amiga's RF shield won't fit back into the Amiga with this clock installed. :-( The other half is that the pins on the Spirit module are too long. :-( :-( Disclaimer: I have no future financial interest in Spirit Technologies, makers of this clock board. I did a few hours of work for them once, and was paid in full. End of relationship. |\ /| . Ack! (NAK, EOT, SOH) What's the difference between an Ack and {O o} . a Nak? Go ahead... try and say it. (="=) bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce U If "hoser" does not work, try my old address at "cogsci"
ler@pnet02.UUCP (09/02/87)
ewhac@well.UUCP (Leo 'Bols Ewhac' Schwab) writes: >[ This line intentionally left blank. ] > > Well, I finally gave up on the Spirit Technologies battery-backed >clock I got. I was willing to hold on to it only if it kept good time. > > It didn't. It gained about ten seconds a week. In my book, that's >awful. So, back it goes. > > I called Spirit Tech and told them I wanted to send it back. The >person on the other end was very nice about it, and gave me a return >merchanise authorization number. I'll throw it at UPS next week. Exactly >when I'll see my money back is another issue, but I have no reason to expect >they'll be underhanded about it. > > In the meantime, my little Amy has her RF shield back, and is happy >as a clam. > > To be perfectly fair, Larry Rosenman (sp?) on PeopleNet recommended >the clock to me, saying that it kept wonderful time, and that he was elated >with it. So obviously, Spirit Tech does make stuff that makes people happy. >This particular people wasn't happy - what can I say? > >_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_ >Leo L. Schwab -- The Guy in The Cape ihnp4!ptsfa -\ > \_ -_ Bike shrunk by popular demand, dual ---> !{well,unicom}!ewhac >O----^o But it's still the only way to fly. hplabs / (pronounced "AE-wack") >"Work FOR? I don't work FOR anybody! I'm just having fun." -- The Doctor The Spirit Clock that I have in My amiga *DOES* keep wonderful time, compared to my quartz watch. It has been in for 2 months now, and hasn't lost a second compared to my watch, which I check regularly to the Phone Company. Apparently Leo got a bum one. Sorry to here it. I have not put my RF shield back on my Amiga, and don't seem to have a problem. I guess that is One problem with the clock board. Seriously, I would STILL recommend the board. BTW, You DID get the spelling right (Thank You) -- Larry Rosenman INTERNET: ler@pnet02.CTS.COM UUCP: {hplabs!hp-sdd, cbosgd, ihnp4 }!crash!gryphon!pnet02!ler UUCP: {philabs, seismo!scgvaxd}!cadovax!gryphon!pnet02!ler
charles@hpcvcd.HP (Charles Brown) (09/03/87)
>> The battery should last "forever". In other words, I want to see >> the battery trickle-charged by system power when the system is on. >The A501 board uses a ni-cad. Of course you'll need an Amiga 500 to use it Ni-Cads will not last "forever" anyway. Just sitting there being trickle charged wears them out. It would be nice to know how long these will last. > Even though >it's backing up 2K of static ram in addition to the clock that battery >should last somewhere in the range of 10-30 years >(="=) bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce I find this hard to believe. Shelf life for Lithium batterys is about three years, and those are long life batteries. Charles Brown hplabs!hp-pcd!charles
dillon@CORY.BERKELEY.EDU (Matt Dillon) (09/03/87)
:>> The battery should last "forever". In other words, I want to see :>> the battery trickle-charged by system power when the system is on. :>The A501 board uses a ni-cad. Of course you'll need an Amiga 500 to use it : :Ni-Cads will not last "forever" anyway. Just sitting there being :trickle charged wears them out. It would be nice to know how long :these will last. : :> Even though :>it's backing up 2K of static ram in addition to the clock that battery :>should last somewhere in the range of 10-30 years :>(="=) bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce : :I find this hard to believe. Shelf life for Lithium batterys is about :three years, and those are long life batteries. : Charles Brown hplabs!hp-pcd!charles Me to. 10-30 years is wayyyy to long. Internal eddie currents inside batteries tend to put an upper limit on shelf life (3-8 years for lithium batteries depending on the type of battery and average temperature over the years). Usually a lithium battery backing up a CMOS static ram chip will last its shelf life as the static ram draws microamps when not selected. -Matt
bryce@COGSCI.BERKELEY.EDU (Bryce Nesbitt) (09/03/87)
[If you reply, be sure to strip off comp.sys.amiga from the distibution!] > In article <4410010@hpcvcd.HP> charles@hpcvcd.HP (Charles Brown) writes: >> >>...that battery should last somewhere in the range of 10-30 years... >>(="=) bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce > >I find this hard to believe. Shelf life for Lithium batterys is about >three years, and those are long life batteries. Ok! There has been widespread disbelief in my statement that this battery could last 10-30 years. I can't argue the technical points, so I will just list my source; the data sheet for the chip. Thomson Components-Mostek claims to have some special longer life batteries. In the data sheet they have a formula for discovering battery life. It depends primarily on temperature. I can't reproduce the charts here, so if you want to see them get data for the MK48T02. Assuming a powered system, with no drain on the battery, wear out is almost exclusively a function of temperature. The chart goes from slightly over 2 years at 75 C to to over 100 years at 5 C. The three year point is at 70 C. When the battery is actually backing up the ram, consumption is a factor. Again the colder it is, the longer things will last. In a simple case with a constant temperature of 25 C and a 65 percent duty cycle the chart shows 10 years. At a zero percent duty cycle (always backing up) the life at 25 C is shown as three years. For more exact figures, see the data sheet. oo bryce@hoser.berkeley.edu -- ucbvax!hoser!bryce if "hoser" does not work, try "cogsci" If this article is mangled, let me know. This rn was finished compiling 10 minutes ago and may not be fully installed yet.
lishka@uwslh.UUCP (Christopher Lishka) (09/04/87)
Now, for my minor $0.02 worth:
Mr. Brown doubted that a Lithium battery could last 10-30 years,
stating that the life would be more like three years...well, I have
one of these buggers in a digital watch, and I've had the watch for
six years without having to replace the battery. Who knows how long
the watch was sitting in the store before I bought it. I have used
all the functions (stopwatch, timer, alarm, beeper) throughout the
years, at times very heavily and at other times not so heavily. The
rated performance is five years...like I said above, I have been going
for six. That may not be quite ten years (or 30), but when my battery
lasts that long, I'll post a message ;-)
Say, wasn't Apple having problems with their Lithium batteries in
their Mac II's? I heard (from a friend) that the batteries are
*soldered* onto a major board. Is this true, and if so, why couldn't
they spend just a bit more money to have a replaceable battery
(especially with the hefty price tag on the Mac II)? Do they also
trust Lithiums to last that long? Enquiring minds want to know...;-)
-Chris
--
Chris Lishka /lishka@uwslh.uucp
Wisconsin State Lab of Hygiene <-lishka%uwslh.uucp@rsch.wisc.edu
\{seismo, harvard,topaz,...}!uwvax!uwslh!lishkafnf@mcdsun.UUCP (Fred Fish) (09/04/87)
In article <4410010@hpcvcd.HP> charles@hpcvcd.HP (Charles Brown) writes: >Ni-Cads will not last "forever" anyway. Just sitting there being >trickle charged wears them out. It would be nice to know how long >these will last. I don't know if it is a record or not, but I have a G.E. electric toothbrush, presumably containing a ni-cad, that still works quite well after at least 10 years of daily use (the motor part, not the brush :-). -Fred -- = Drug tests; just say *NO*! = Fred Fish Motorola Computer Division, 3013 S 52nd St, Tempe, Az 85282 USA = seismo!noao!mcdsun!fnf (602) 438-3614
denbeste@cc5.bbn.com.BBN.COM (Steven Den Beste) (09/04/87)
Chris Lishka writes: > > Mr. Brown doubted that a Lithium battery could last 10-30 years, > stating that the life would be more like three years...well, I have > one of these buggers in a digital watch, and I've had the watch for > six years without having to replace the battery. Sorry, that's a SILVER battery, not a LITHIUM battery. Nonetheless, A project I was on at Tektronix used a large lithium battery (two inches diameter by an eighth of an inch thick) to back 4K of CMOS RAM. Our spec was guaranteed 7 years. The only kind of battery I know of that can stand up to continuous charge and uncharge is a lead-acid (that's SULPHURIC acid) battery. They are BIG and CLUNKY and DANGEROUS and let's not and say we did, OK? There are some newer sealed lead-acid batteries intended for this kind of application, but they are still substantially larger than a lithium battery (20-fold?). About the upper limit on NiCads is 100 charge cycles - less if the battery doesn't totally discharge between charges. Of course, I could be wrong about this stuff... I just twiddle bits, I don't grease the electrons... > > -- > Chris Lishka /lishka@uwslh.uucp > Wisconsin State Lab of Hygiene <-lishka%uwslh.uucp@rsch.wisc.edu > \{seismo, harvard,topaz,...}!uwvax!uwslh!lishka -- Steven C. Den Beste Bolt Beranek & Newman, Cambridge MA denbeste@bbn.com (ARPA or CSNET or UUCP) harvard!bbn.com!denbeste (UUCP) I don't think BBN cares what I think about this stuff.
grr@cbmvax.UUCP (George Robbins) (09/05/87)
In article <1@cc5.bbn.com.BBN.COM> denbeste@cc5.bbn.com.BBN.COM (Steven Den Beste) writes: > Chris Lishka writes: > > > > Mr. Brown doubted that a Lithium battery could last 10-30 years, > > stating that the life would be more like three years...well, I have > > one of these buggers in a digital watch, and I've had the watch for > > six years without having to replace the battery. > > Sorry, that's a SILVER battery, not a LITHIUM battery. > > Nonetheless, A project I was on at Tektronix used a large lithium battery > (two inches diameter by an eighth of an inch thick) to back 4K of CMOS RAM. > Our spec was guaranteed 7 years. > > About the upper limit on NiCads is 100 charge cycles - less if the battery > doesn't totally discharge between charges. Look, batteries are a complicated subject. If there was one kind that was truly magic, than we wouldn't have to deal with so many varieties. Lithium batteries are probably OK for most ultra-low power applications. They should last for the time most users are likely to have there equipment, however some will fail early and some users will expect their equipment to last for like to forever. On the A500 I selected a Ni-Cd, because even though the performance might deteriorate over the years, the cheapest (which is not the smallest) Ni-Cd provides so much more capacity than the clock chip needs, that it should work fine a long time from now. Next time, perhaps we'll use a lithium battery, it really doesn't make that much of a difference.... -- George Robbins - now working for, uucp: {ihnp4|seismo|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@seismo.css.GOV Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
bilbo@pnet02.UUCP (09/07/87)
So lets mount the battery off the board in the frig...
Bill :-)
UUCP: {hplabs!hp-sdd!crash, seismo!scgvaxd!cadovax}!gryphon!pnet02!bilbo
INET: bilbo@pnet02.CTS.COM
*Bilbo*
(Recombinant Hobbit)