Color Classic analog board trouble

[retr01]

Whoa. 8.5 to 9 Watts when off?! Is that bad news for the electricity bills? Hmmm.

Why would a compact Mac draw power even when it is off? Is it necessary so it can be ready the next time it powers on to boot up?

 

[JDW]

It is bad news for electricity bills, but the alternative of keeping that switch off is also bad because it will drain your PRAM battery dry pretty fast if you don't use that machine often. And there are big disadvantages to totally removing the PRAM battery.

I think it's bad design. The convenience of having the Power Key to power on the Mac isn't worth all that power draw from the wall socket. Maybe it didn't matter much in the early 90's when power was cheaper, but it matters more today.

 

[Crutch]

I’m sorry if this is in your video which I haven’t yet watched but - what is the extra work that the PRAM battery does if the Color Classic is unplugged?

 

[YMK]

Here's part of my video work on this topic, covering my Thermal Camera readings of the machine with the stock analog board components, with only the power switch at back on (the CC is shutdown).

My 120V CC pulls 10 watts according to the Kill-a-Watt.

I keep it switched off more for wear and tear reasons than power consumption.

@JDW, that is cool! So, what about the CPU and the Apple IIe Card?

Heatsinks are tricky to measure with the thermal cam since they act like mirrors.

Applying a strip of tape helps with getting an accurate reading.

 

[JDW]

@Crutch

Without a PRAM battery, you lose these settings...

• Mouse Tracking Speed (the biggest pain for me)
• Date & Time
• Startup Boot disk selection
• AppleTalk on/off status (unless "Open Transport" is installed)
• Disk Cache size setting
• Cursor Blink rate
• International Map setting
• Highlight colour setting

 

[retr01]

@Crutch

Without a PRAM battery, you lose these settings...

• Mouse Tracking Speed (the biggest pain for me)
• Date & Time
• Startup Boot disk selection
• AppleTalk on/off status (unless "Open Transport" is installed)
• Disk Cache size setting
• Cursor Blink rate
• International Map setting
• Highlight colour setting

• Sound volumes

and much more!

Keeping the PRAM energized is essential for 68k Macs. Not sure about Power Macs if fewer settings rely on PRAM.

There are a couple of extensions for Systems 6 and 7 that can save and restore if the battery goes south.

PRAM Auto-Restore
PRAM 5.0

 

[JDW]

After keeping my CC in the same state for 24 hours (switch at back of CC ON, but CC shutdown), my wattmeter tells me the following...

• 0.192kWh x 365 days = 70.08kWh / year
• ¥5.59 x 365 days = ¥2,040 / year (roughly US$20/yr)
• After 24 hours, power consumption stabilized to 8.3W (down from 9W initially)

But keep in mind, that's with the machine shutdown, CRT off, HDD not spinning, etc. It's mostly "waste current" flowing into these components, which in turn gets dissipated into the air as heat:

• RP12† @91°C: PTH451C, CRT degaussing thermistor
• RP23† @54°C: 22kΩ 3W resistor**
• RP8† @41°C: 15kΩ 5W resistor**
• DL21&DL22* @95°C: 18V Zener diode pair (1W?)
• RF11* @90°C: 390Ω, 1W
• RL62* @85°C: 47Ω, 1W
• DF2 @70°C: 1N5237B 8.2V Zener Diode (1/2W?)

† Located on the RIGHT side of the CC analog board, when you're facing CRT.​

* Located on the LEFT side of the CC analog board, when you're facing CRT.​

** These 2 resistors don't get hot enough to worry about. I just measured them FYI.​

NOTE: Temperatures above are approximations based on measurements with a 121GW Multimeter w/Type-K sensor and a UTi260B Thermal Imager.​

​

As you can see, RP12 and the DL21/22 zener diode pair are the hottest components on the analog board (machine in shutdown state).

​

There are also 5 capacitors in the area of DL21/22 which reach 50°C to 55°C.



That will be their constant temperature as long as the CC is plugged in and the switch on back flipped ON (machine shutdown but powered state). The constant heat will mean a shorter life for them versus other caps which are cooler or cold under that condition.

I will now press the keyboard power key and make further temperature measurements with the machine up and running and report on that later.

 

[JDW]

I have finished my PowerKey-ON temperature measurements (CRT ON, HDD spinning, booted to Finder doing nothing).

My watt meter shows 68W (0.94A@104VAC) with the CC in the booted state. I am using Japan's 100V 60Hz power. Also note that I have a spinning hard drive installed, a IIe Card installed, and a 50MHz full FPU 040 on the motherboard too. There is also a little 40mm fan on the motherboard, powered by 12V. On top of all that, I have the regular VGA mod installed. So my setup probably draws more power than the average Color Classic.

Interestingly, DL21 & DL22, which are the hottest components in the shutdown state, dropped 12°C in the booted state to about 83°C, down from 95°C.

I did not find substantial temperature differences on the other components mentioned in my previous post, except for the capacitors, which get hotter. Specifically, the caps which are 50°C to 55°C in the shutdown state, rise to 70°C or slightly above in the booted state. There's nothing that can be done about that. It's just something to note because the hotter electrolytic caps, the shorter their life.

In the shutdown state RP12 (thermistor/degausser) is the second hottest component at 91°C. While that temperature remains the same in the booted state, RP1 (round thermistor marked "PH NTC 100"), which is off in the shutdown state, heats to 100°C in the booted state. Not something we need to worry about, but just FYI.

　

As @YMK discovered earlier in this thread, RL22 which isn't powered in the Shutdown state, becomes the single hottest component on the Analog Board with the machine booted. My measurement of 170°C is pretty much what YMK measured with his thermal camera. Note that I set the Gain to LOW because my UTi260B has two ranges. Anything being measured which is hotter than 150°C should be measured with the 260B in LOW GAIN, and that is what was used when I made the capture below.



RL22 is one of the parts I purchased, and I will swap it out with a higher wattage part, while also stand it up on its legs for better cooling.

Bear in mind that all my measurements to date have been done with the case back removed, which means there is no air flow from the fan. In the shutdown state, the component temperatures are likely a bit higher that what I measured because the case back is normally on, and of course the machine is shutdown so there is not case fan running. Not sure how much difference there is in the booted state with fan running versus the case back being off. Maybe its a wash, or maybe the board is slightly cooler with the case off.

I will need to continue that project next week.

By the way, I found an excellent PARTS LIST for the Color Classic ANALOG BOARD. It's not comprehensive, but it certainly helps when trying to figure out component markings or sizes or even wattage ratings!

Thanks again to YMK for having alerted everyone to this. Just as the Macintosh 128/512/Plus Analog Board was enhanced with better parts recommended by Larry Pina, we now have a list of parts to slightly improve the Analog Board of the Color Classic.

 

[Crutch]

@Crutch

Without a PRAM battery, you lose these settings...

• Mouse Tracking Speed (the biggest pain for me)
• Date & Time
• Startup Boot disk selection
• AppleTalk on/off status (unless "Open Transport" is installed)
• Disk Cache size setting
• Cursor Blink rate
• International Map setting
• Highlight colour setting

Sorry yes I am very familiar with PRAM and its function - but I read you to say that a CC battery drains faster when the CC is unplugged: “… the alternative of keeping that switch off is also bad because it will drain your PRAM battery dry pretty fast if you don't use that machine often”.

This implies to me that either the PRAM battery draws wall current to keep itself charged (does it? I don’t have a CC but I don’t think this is true) thereby losing charge faster if not plugged in, or that the PRAM battery somehow does some extra work when the CC is unplugged vs plugged in. Is one of these things true?

 

[JDW]

...either the PRAM battery draws wall current to keep itself charged (does it? I don’t have a CC but I don’t think this is true) thereby losing charge faster if not plugged in, or that the PRAM battery somehow does some extra work when the CC is unplugged vs plugged in. Is one of these things true?

I have no proof, but I think that as long as you have a cord in the wall socket and have the switch on back flipped ON, your PRAM battery won't drain at all. If there is drain, it would be the normal losses as if the battery were sitting on a shelf. But the moment you yank the power cord or flip the switch on back, the circuit suddenly starts drawing power from the PRAM battery. If you later flip on power again (not booting, just the switch on back), then the PRAM battery is basically disconnected again.

I speculate this is true because when I had my CC in storage and a fresh PRAM battery installed (3pcs of LR44), when I later pulled in the machine about 3 months later, the PRAM battery was dead. But I later decided to keep the machine plugged in, and after a few months of that, I would kill power and then apply power and boot, and the settings were still my custom settings, proving the PRAM battery was still alive.

This is why I think that when you have a power cord connected to the wall socket AND have the switch on back set to ON, it pulls PRAM settings power from the wall socket power, drawing nothing at all from the PRAM battery. So it's not a matter of it trying to "charge" the PRAM battery. It's a matter of the PRAM battery being drained or not. I don't think it is drained when ON and attached to the wall socket.

 

[Crutch]

Interesting. I didn’t know that and I’d like to research it someday. Thanks.

 

[YMK]

I have no proof, but I think that as long as you have a cord in the wall socket and have the switch on back flipped ON, your PRAM battery won't drain at all.

I made some quick measurements.

For the LC575 LB at least, you are correct. There's no drain on the battery while the switch is on.

When it's off, the drain is 42.5uA.

For a typical 0.7Ah lithium battery, this yields a lifespan of 16,470 hours or 1.88 years.

 

[JDW]

For the LC575 LB at least, you are correct. There's no drain on the battery while the switch is on.
When it's off, the drain is 42.5uA.
For a typical 0.7Ah lithium battery, this yields a lifespan of 16,470 hours or 1.88 years.

Yes, my motherboard is an LC575, not a stock CC or CCII motherboard.

The stock PRAM battery is a rather large 4.5V RAYOVAC (even for the stock CC motherboard), and while there are replacements found on EBAY these days, I refuse to pay upwards of $30 (not including shipping) for one, even though they would last longer than any other replacement PRAM battery solution.

What I ultimately decided to go with was recommended by @Kay K.M.Mods . I used a cheap triple LR44 battery case sold on Amazon (affiliate link). You can see me demonstrate that LR44 battery case at the 9:57 timestamp in my video here:

I later moved it to a new location in order to make room for my Apple IIe Card, which you can see here:

It is that same triple LR44 replacement PRAM battery solution I spoke of in my earlier post when I said that it was drained dry in only a few short months when I had the CC in storage and not connected to wall socket AC power.

LR44 batteries have a rather pathetic capacity of between 105mAh and 150mAh. Even if you can buy them at the upper end 150mAh capacitor, using your measurement of 42.5µA current draw as a guide, that calculates into 4 months and 3 weeks of life when your CC is placed in storage. And if you happen to get the lower end 105mAh LR44s, 3 pieces in series of that yields only 3 months and 1 week of battery life. I used this battery life calculator.

Keep in mind that merely connecting a power cord to the wall socket is not enough. You must then flip the power switch on back to ON for the AC power to keep the PRAM memory intact, which then will sever the connection to your PRAM battery, thereby prolonging its life and making it available whenever you either flip the switch on back OFF, or yank the power cord.

 

[YMK]

The stock PRAM battery is a rather large 4.5V RAYOVAC (even for the stock CC motherboard), and while there are replacements found on EBAY these days, I refuse to pay upwards of $30 (not including shipping) for one, even though they would last longer than any other replacement PRAM battery solution.

For now, I soldered in a leaded 3.7V lithium battery, like some SEs have.

I will probably go with an Eneloop AA/AAA pack with a charging circuit like I have in my Gateway 486. It's maintenance-free as long as the machine runs once in a while.

 

[JDW]

Regarding current draw on the PRAM battery, I found this old post of mine. As you can read there, I put a bench top power supply on the disconnected LC575 motherboard and used an ammeter to measure a 63µA current draw (constant). That is a bit higher than your 42.5µA measurement, although both numbers are in the same ballpark and shows just why LR44's drain dry in only a few short months.

Your 3.7V Lithium battery solution, despite being lower than 4.5V, will still probably last quite a bit longer than my triple (in series) LR44 solution. And your rechargeable Eneloop pack will likely get the longest life of all.

Honestly, it would be nice to have 1 year of life. That way you can put a machine in storage and reasonably pull it out once a year to power it on. We ought to do it more than once a year, otherwise our vintage computer collection is rather wasted, but life often gets in the way, and we may lack time to power on all our individual machines but once a year.

 

[retr01]

How about...RECHARGEABLE battery?

It may be expensive, but it can have a USB cable to the battery terminals. Then that USB A cable is plugged into a USB power source, such as a battery bank.

Another option is a rechargeable battery with a USB placed in the battery socket with USB cabling to the lock slot. A flashing light can indicate a low battery, prompting you to plug into a USB power source.

There are many ways we can devise modern solutions to this problem impacting all compact and modular Macs.

 

[YMK]

How about...RECHARGEABLE battery?

The Eneloops I mentioned are rechargeable. In the picture of the 486 above, the two AAA batteries are charged through the motherboard's fan header and the yellow LED indicates charging current.

A triple AA/AAA arrangement for the Mystic is certainly possible, but the charging circuit will be different due to the higher voltage.

 

[retr01]

The Eneloops I mentioned are rechargeable. In the picture of the 486 above, the two AAA batteries are charged through the motherboard's fan header and the yellow LED indicates charging current.

A triple AA/AAA arrangement for the Mystic is certainly possible, but the charging circuit will be different due to the higher voltage.

YAASSSSSSS!!! That! Make the fan (or a squirrel) do the work whenever the Mac is used. Mounting options are flexible. I'd love that in my Macs, especially the SE/30 and P476.

Would this work? And the terminals with light?

 

[YMK]

Would this work? And the terminals with light?

Yes, those are the right batteries. The holder is from Mouser, and the blue part is a model I printed. The charging circuit is embedded within it.

 

[retr01]

Yes, those are the right batteries. The holder is from Mouser, and the blue part is a model I printed. The charging circuit is embedded within it.

Can you share the mouser PN and that 3D print model?