Amiga battery replacement

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser. Please create an account and log in to see all content by default. This is a limitation of the forum software.

Also users that are not logged in can not create new threads. This is a, unfortunately needed, counter measure against spam. Please create an account and log in to start new threads.

Don't Panic. Please wash hands.
  • Amiga computers use a rechargeable battery. It very important to use a rechargeable battery when replacing a coin battery. Use a LIR2032 Li-Ion rechargeable cell which is a 3.6v cell, the same voltage as what is in the Amiga's. Do not use 2032 which are meant for one time use as they can cause damage to the board.

  • Almost :-)

    A Li-Ion cell needs an elaborate charge controller that starts with constant current, and when crossing a certain cell voltage, changes to constant voltage. This is the only safe way to charge a Li-Ion cell. Anything else (like the constant resistor in the Amiga) bears the danger of overcharging the cle,, leading to thermal runaway (aka fire that isn't easily extinguished). You don't want that in a classic computer, and you certainly don't want that in your house.

    The generally-accepted method is to use a standard coin cell and insert a diode that keeps it from charging.


    The buddha is running great! I did a 1gb file copy test from BDH1 to BDH1.1 hard drive #2 in right IDE connector, it took 17 minutes. I figured that's about 1mb/s transfer rate. Do you have fix for BuddhaSpeed that might improve this?

    I did a battery charging tests with voltages and temperatures. All temperatures are in C. Computer battery circuit diagram A3000D/T. All tests done after 2 hours power on computer running, motherboard fans off. Only fan running is the power supply. Side cover open. I have a fan board in the top zorro socket it is off power connector is unplugged. R 192 is next to the battery holder. All measured temperatures are with a non-contact digital thermometer with laser dot.

    Room temperature = 21.2c

    MB Top = 32.7c

    MB middle =46.7c

    MB Bottom = 37.4c

    Li-ion LiR2032 Battery = 31.6c

    68030 = 41.4c

    68060 = 45.8c

    68882 = 36.5c

    Top KS 3.2 Rom = 32.5c

    Bottom KS 3.2 Rom =33.3c

    R 192 = .778v / 470 ohm = 1.65 ma x .778 = 1.28 mw.

    Li-ion LiR2032 Battery = 4.49 v

    As you can see there is no chance of battery failure and setting your house on fire. Max current 12v to R190, D191 and R192 to ground is 10.8 ma = 130 mw that would light up a LED about 1/2 bright. Now a Tesla would burn down your house! I didn't try a CR2032 non rechargeable battery, but long-term use would lead to battery failure.

  • Do you have fix for BuddhaSpeed that might improve this?

    If you copy from Buddha to Buddha, you are cutting your performance in half. The average speed you get on IDE devices in aa Z2 bus is just over 2MByte, so your finding is that you're on par with the expected performance.

    All measured temperatures are with a non-contact digital thermometer with laser dot.

    I have one of those - and while it shows numbers with a dot, you assume that it's at least accurate to that one digit it's showing. It isn't. The laser is off by several centimeters, and temperature is off by several degrees, I so admit it was an el-cheapo thing for under 100,- EUR, but the readings I get from my IR camera (Flir, the hackable one) are substantially different. I tend not to trust these laser dot/IR thermometers, because there's so much Chinese junk out there.


    Max current 12v to R190, D191 and R192 to ground is 10.8 ma

    You need to look at that circuit at two parts: R190/D190/D191 form a clamp, and there is a constant 6.3V drop on R190, so that burns 84.42mW all the time, resulting in a 5.0V voltage at the top of R192.

    Voltage over R192 will not be the complete 5V, as the battery will be "full" at some point. So yes, the max. current is 10.8mA, but you continue charging a full Li-Ion battery, which is bad news no matter how low your current is. Even if the voltage drop over R192 will be 1V (=2mA charge current) because the cell is already full, you are still raising the cell voltage over recommended levels. This will reduce it's life considerably. I'd add another 1N4148 in series to drop the charge voltage to 4.3V, which is what most Li-Ion cell chemistries can accept. The "cleaner" approach would be to replace D191 with a 1.5V Zener type, which will limit the cell voltage to 4.2V, which is what I'd prefer for longer cell life.

    Those of you who want to use a non-rechargeable coin battery CR2032 just remove D191 then you can use the battery safely.

    You should also remove D190 in that case, so R190 won't burn energy constantly.

  • Jens.

    I agree with you on there could be improvements on the Amiga charging circuit to use the Li coin battery. I still stand by my testing that the is no chance of any damage to the Li battery with the current Amiga circuit also the Amiga does not charge the battery when turn off. I tested a discharged Li rechargeable 3.6v coin battery measuring 3.1v and charged it at a constant current of 20ma and set constant voltage to 4.7v the max battery temperature measured was 27.2c it was cold to the touch. Most Amiga owners don't have the proper tools to work on static sensitive circuit boards, grounded wrist strap, anti-static mat, vacuum de-soldering station and battery powered soldering iron to mention a few. Removing the mother board and changing out the battery with an original nicad battery could be dangerous to the MB but the battery needs to be removed cleaning the MB and replaced. Changing out other components is taking another chance. You are dealing with Amiga electronics that are over 35 years old. Just remember the "KISS...." KEEP IT SIMPLE STUPID", if it isn't broke, don't fix it! When taking your Amiga for repair ask them for the above-mentioned tools.

    I have a 9 x 15-meter shop but in the winter, I can't work there as it is at times -20c in the winter and +60c in the summer. In my house I have a 3 x 3-meter shop, my wife hobby is sewing she has two rooms 3.6 x 3.6 meter and a 3.6 x 4.6-meter, she also uses the kitchen and living room at times so my shop is crowded with all my test equipment and electronics that can't be in the large shop.

    Here are some of my test equipment, I had to crop them to get under the 1mb size.

  • I got the battery holder from old windowzzz MB, find one that has the right pin spacing. It may have clearance issue with R192, just file off a little from the holder to get clearance. Make sure you have the correct polarity. Here are the 3 holes.

    + +



    I'm not complaining about the buddha transfer speed but if you have made improvement on the buddhaspeed program.

  • but if you have made improvement on the buddhaspeed program.

    The Buddhaspeed tool only does one thing: It writes a value to a register that only the very old (non-flash from 1997 and early flash models from 1998) have. These registers do not speed up, but they slow down for compatibility reasons. Back when I developed the first Buddha, the two resellers who pushed this were keen on making old drives compatible, so I added that feature.

    I removed the slow-down register for two reasons:

    1) Z2 is already pretty slow, and the only drives that really required it were 20MByte AT drives that pre-date the IDE standard.

    2) it did not work properly with some (questionable) Z2 implementations for the A1200

    So there is no improvement to make to the Buddhaspeed tool: There is just no control over speed any more, as the Buddha will always go as fast as the Z2 bus will allow it to.

  • Jens

    So... because buddhaspeed keeps popping up during each bootup can I delete buddspeed in the install in the InstallCD directory script on the DOM? On HDToolBox I made a copy of HDTB and renamed it to BuddhaHDToolBox, the file by right click it and changed scsi.device to buddhascsi.device it now detects the buddhascsi.devices hard drives which allows adding additional drives and make additional partitions. I have made backups of my boot HD in case something goes wrong but so far it is working great.

  • Jens

    With power off I removed D191, and the battery there is 0v after the capacitors discharged on pin 18 of the clock chip. With the power on the voltage went to 4.8v on pin 18 and 4.7 at the empty battery holder. Is there an updated schematic diagram? There must be another path feeding the clock chip pin 18 Vcc. With the battery (3.92v) installed the voltage went to 4.43v with power on and charging, Your thoughts.

  • Jens

    I removed the battery and clock chip RP5C01, with Amiga powered on I measured 0v on pin 18Vcc on the socket. It seems that the clock chip is getting Vcc from itself, the only source is from the IO buss, WR, RD, CS1 and from the alarm output which has a 1k pullup resistor to +5v RP180 pin 15. I am going to put the Li-ion battery and clock chip back in and just live with.

  • I was too quick following your suggestion of removing the charge diode. The suggestion of removing D191 will certainly keep the battery from being charged, but it will also cut off the suppoy of the RTC chip, requiring it to be powered from the battery all the time. That's of course not what you want.

    So D191 should go back in, but R192 should be replaced with a diode that makes sure current only flows from battery to the clock chip. Since the battery voltage is lower than that of the powered Amiga, it will not be drained while the computer is on.