Computer Saver from Ray Carlsen

[rber...@iglou.com]

Ray Carlsen, veteran Commodore repair technician, is now selling Computer Saver (an improved version of his original Computer Saver, the schematic of which had been on his website for years). Like the original, his new Computer Saver functions as a voltage limiter when plugged in-line between the C64 and the C64 power supply. Read what Ray says about his product --

---------------- Original Message -----------------
From: "Ray Carlsen" <rcarlsen(at)tds.net>
Date: Fri, June 21, 2013 11:53 pm
---------------------------------------------------

Hi,

Photos of the construction of those savers are on my site -- http://personalpages.tds.net/~rcarlsen/cbm/c64/SAVER/EXTERNAL/

AN EXTERNAL "COMPUTER SAVER" PROTECTION DEVICE FOR THE C64

This shows a small run of the Computer Saver I designed. It is made to be installed between any C64 and its power supply. The two LED's on the Saver case are indicators of the PS status. The LED on the left monitors the 9VAC from the supply and it should be on all the time the PS is plugged in to AC power, whether the computer is on or off. The other LED (marked "failsafe") is normally off. It only comes on if the PS fails due to a shorted internal regulator. That fault is what damages chips in the computer, most often the RAM. Since there is already an LED on the computer case which
monitors the regulated +5VDC, one on the Saver was considered unnecessary.
The cost of the Saver is $50 US. That price could be reduced somewhat with a larger production run, but unless there is greater interest in this device, I'm not ready to tool up for that. More than half that amount was spent for parts, including shipping. Any time something is hand-made and parts are individually purchased, the price of the end product will be higher than people expect. The construction of a stand-alone device is normally higher than the same circuit built into a computer such as the C64. That's due to the added expense of a case, cable wire and connectors as well as the added time it takes to assemble the device. There are few shortcuts, so the price is firm.
One version of this device has been seen for sale on the Internet, but the builder likely didn't consider one thing when constructing it. Most importantly, the "trip" point of the Saver is critical. That is the exact voltage level at which the device cuts off power to the computer. Because of variable tolerances of some of the components, the trip point must be manually set with an accurate voltmeter and variable bench power supply. This setting is done as a last step after the device is built. If it is set too high, the computer is still at risk, and if too low, the device may cut power with a normal power supply that has a slightly abnormal but acceptable output level during a "no load" condition such as when the supply is plugged in to AC power but the computer is turned off. Keeping those voltage limits in mind, I found the optimum trip point to be between 5.3 and 5.4 volts DC. RAM chips have an absolute maximum rating of 5.5 volts, so the protector -must- be set below that value.
Because the protectors LED's are on its case, I decided to make the cable between it and the computer rather short, about one foot long, so those LED's can be monitored. If desired, that cable can be made longer so the protector is off the desktop. However, very long cables will reduce the voltage at the computer since it draws nearly one Amp in normal operation. The short cable seemed like the best arrangement.

Ray

----------

See Computer Saver on exhibit at CommVEx,
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug
July 27-28 Commodore Vegas Expo v9 -
http://www.portcommodore.com/commvex

[Clocky]

Interesting device but how does it protect the C64 from voltage ripple
caused by bad PSU caps?

Of all repairs I've done of late, most are failed DRAMs, Mask ROM's and
other sensitive chips (in all kinds of retro systems) which I suspect
were due to voltage ripple caused by out-of-spec filter caps where the
voltage regulator has not failed and overvoltage hasn't been the cause.

Sometimes these capacitors reform in power supplies that have been idle
for a long time and once reformed are serviceable but in that period in
between damage can be done.

Adding a high quality filter capacitor may be all that is required to
smooth the voltage help protect the equipment in this computer saver
which may be something to consider, even if it increases the cost.

Correct me if I'm wrong or if that problem is already taken care of
using another method.

[Computer Nerd Kev]

On 23 Jun 2013, Clocky wrote:

>
> Interesting device but how does it protect the C64 from
> voltage ripple caused by bad PSU caps?
>
> Of all repairs I've done of late, most are failed DRAMs,
> Mask ROM's and other sensitive chips (in all kinds of retro
> systems) which I suspect were due to voltage ripple caused
> by out-of-spec filter caps where the voltage regulator has
> not failed and overvoltage hasn't been the cause.
>
> Sometimes these capacitors reform in power supplies that
> have been idle for a long time and once reformed are
> serviceable but in that period in between damage can be
> done.
>
> Adding a high quality filter capacitor may be all that is
> required to smooth the voltage help protect the equipment
> in this computer saver which may be something to consider,
> even if it increases the cost.

Eventually though the PSU capacitor would be likely to short
out, then taking out the fuse which I understand needs to be
unearthed from a brick of resin in order to be replaced (along
with the cap). It would be better to have an indication of
when the PSU cap was starting to go.

I've seen ripple detection circuits when looking for design
ideas for the bench power supply I want to build some day.
I'll have to see if I can find them again.

--
__ __
#_ < |\| |< _#

[Clocky]

Computer Nerd Kev wrote:
> On 23 Jun 2013, Clocky wrote:
>
>>
>> Interesting device but how does it protect the C64 from
>> voltage ripple caused by bad PSU caps?
>>
>> Of all repairs I've done of late, most are failed DRAMs,
>> Mask ROM's and other sensitive chips (in all kinds of retro
>> systems) which I suspect were due to voltage ripple caused
>> by out-of-spec filter caps where the voltage regulator has
>> not failed and overvoltage hasn't been the cause.
>>
>> Sometimes these capacitors reform in power supplies that
>> have been idle for a long time and once reformed are
>> serviceable but in that period in between damage can be
>> done.
>>
>> Adding a high quality filter capacitor may be all that is
>> required to smooth the voltage help protect the equipment
>> in this computer saver which may be something to consider,
>> even if it increases the cost.
>
> Eventually though the PSU capacitor would be likely to short
> out, then taking out the fuse which I understand needs to be
> unearthed from a brick of resin in order to be replaced (along
> with the cap). It would be better to have an indication of
> when the PSU cap was starting to go.
>

Sure, but for around $25 you could get a new switchmode +5/+12V supply
module and 9VAC/2A transformer and case off Ebay and you could recycle
the cables from a C= brick to put it all together and you could place
the rest of the C= PSU where it belongs. Might be the better option long
term.

> I've seen ripple detection circuits when looking for design
> ideas for the bench power supply I want to build some day.
> I'll have to see if I can find them again.
>

There are bound to be schems online but if you find the one you have in
mind can you please put it up somewhere for download?

[Groepaz]

Clocky wrote:
> Sure, but for around $25 you could get a new switchmode +5/+12V supply
> module and 9VAC/2A transformer and case off Ebay and you could recycle
> the cables from a C= brick to put it all together and you could place
> the rest of the C= PSU where it belongs. Might be the better option long
> term.

i was thinking the same.... for $50 it should do a whole more than it
does... since you can easily get a (much) better psu for that money instead
of trying to make 25 year old powerbrick safer.

>> I've seen ripple detection circuits when looking for design
>> ideas for the bench power supply I want to build some day.
>> I'll have to see if I can find them again.
>>
>
> There are bound to be schems online but if you find the one you have in
> mind can you please put it up somewhere for download?

mmmh.... thinking about it... it should be as easy as a capacitor in series
(high pass, remove DC), then rectifier (rectify the remaining AC component),
schmidt trigger with threshold according to wanted max. ripple, and finally
a latch/flipflop that shows its status in a LED.

--

http://www.hitmen-console.org http://magicdisk.untergrund.net
http://www.pokefinder.org http://ftp.pokefinder.org

The weirder you are going to behave, the more normal you should look... When
I see a kid with three or four rings in his nose, I know there is absolutely
nothing extraordinary about that person.
<P.J. O'Rourke>

[Ray Carlsen]

> Interesting device but how does it protect the C64 from voltage
> ripple caused by bad PSU caps? Of all repairs I've done of late, most
> are failed DRAMs, Mask ROM's and other sensitive chips (in all kinds
> of retro systems) which I suspect were due to voltage ripple caused
> by out-of-spec filter caps where the voltage regulator has not failed
> and overvoltage hasn't been the cause. Sometimes these capacitors
> reform in power supplies that have been idle for a long time and once
> reformed are serviceable but in that period in between damage can be
> done. Adding a high quality filter capacitor may be all that is
> required to smooth the voltage help protect the equipment in this
> computer saver which may be something to consider, even if it
> increases the cost. Correct me if I'm wrong or if that problem is
> already taken care of using another method.

Clocky and to all concerned:
I think you are wrong in your assumption that failing filter caps
and the resulting ripple are responsible for failed chips in a computer.
When a filter cap dries out inside from repeated overheating, it becomes
less effective as a voltage smoother because its capacitance is reduced.
The resulting ripple, if not too great, is compensated for via the
regulator IC inside the supply. As the ripple increases, the voltage on
the input side of the regulator cannot be maintained and it's output
voltage starts to drop. At that point, there is ripple on the output
side which can appear as hum in the computer audio and/or a wavy video
image on the screen. If the regulator is good, the voltage at its output
will never exceed, even with ripple on it, the design limit of 5 volts.
At 60Hz, the regulator is fast enough to keep such voltage surges from
exceeding 5 volts. Therefore, that alone is not the cause of chip
damage. As old as this hardware is, chips fail for a variety of reasons
including excessive heating (the PLA & SID... add heat sinks!) and old
age for just about anything else.
With a lowered supply voltage and or excessive ripple on the supply
line, the computer may not even boot up (LED dim or off) or operation
may be erratic. One way to tell if the supply is the problem (inadequate
internal filtering) is to connect the PS to the computer and turn it on,
then plug the supply in to AC power. If the computer boots with that
alternate arrangement, the PS is probably failing from bad filters and
it's time to replace it.
By the way, adding extra filtering on the output side of the
regulator of a failing supply will not compensate for an open cap on its
input side because the regulator itself is "dropping out" under load. A
failed (for whatever reason) brick must be replaced. It cannot be
repaired since all components are embedded in epoxy, making repairs
uneconomical to say the least. If it's time to replace a failed PS, try
for a better one.
A failing (via internal shorts) regulator IC, on the other hand,
can and does cause chip failures. Inside the Commodore "brick", that IC
runs very hot since its heat sink is embedded in epoxy and gets no air
cooling but just a bit of conduction cooling through the epoxy from the
outer case. One typical failure mode for that IC is voltage "creep"
upwards as the supply reaches normal operating temperature. If you
measure the 5 volt line when the supply is cold, it may read normal. If
put on an artificial load (NEVER a computer for tests like this), the
voltage can read normal but then rise, sometimes abruptly, to nearly
eleven volts! The spec sheet for 4164 RAM chips states an "absolute
maximum rating" of 5.5 volts (typical for TTL) on its supply pin. When a
PS fails via overvoltage, the RAM chips (usually more than one) are the
first to suffer damage, usually by internal shorts. The damage is silent
and repeatable! That same supply may test OK on another computer after
it cools down but the same thing will happen to the next computer if
it's left on. It's a common enough problem that I always insist the user
send the PS along with a C64 when getting the computer repaired. I
always load test the supply to be sure it's OK.
The C64 brick has no internal fuse in the 5 volt line. Some
after-market supplies will have one or more however. The 5V regulator IC
inside a brick is designed with something called "foldback current
limiting". If an excessive load (a short in the computer, for example)
is put on that supply line, the voltage will drop down to limit the
current delivered to the load. With that feature, a fuse was likely
considered unnecessary. Of course if the regulator shorts, its output
voltage will rise... the worst failure mode of that device. That's the
reason for the Computer Saver if you have a brick as your only supply.
The Saver cuts power to the computer if the supply voltage exceeds a
safe level. Note that the Saver is only necessary if you're using the
original Commodore black or white bricks. After-market supplies, even
the analog ones, are built with larger heat sinks, are not filled with
epoxy and are therefore repairable, and are generally more reliable.
I've only seen one that needed repairs. The best supplies are those with
a switch-mode 5 volt PS installed and they are just starting to appear
on the market. Switchers run cooler and don't suffer the same problems
of the old analog power supplies and therefore don't need a Saver to
protect the computer. However, as with any new product, I've seen good
and bad (cheap) after-market switch-mode supplies. Buyer beware.

Ray

[Computer Nerd Kev]

On 24 Jun 2013, Groepaz wrote:

> Clocky wrote:
>>> I've seen ripple detection circuits when looking for
>>> design ideas for the bench power supply I want to build
>>> some day. I'll have to see if I can find them again.
>>>
>>
>> There are bound to be schems online but if you find the
>> one you have in mind can you please put it up somewhere
>> for download?
>
> mmmh.... thinking about it... it should be as easy as a
> capacitor in series (high pass, remove DC), then rectifier
> (rectify the remaining AC component), schmidt trigger with
> threshold according to wanted max. ripple, and finally a
> latch/flipflop that shows its status in a LED.

Found at least one of the circuits I was thinking of in the
"Electronics Test Bench" special edition of the Australian
Silicon Chip magazine from 2000, part of their project "A Dual
Tracking +/-18V Power Supply" (don't know which issue they
grabbed it from originally).

It works in the spirit of what you suggest, but uses a couple
of diodes after the cap to keep the voltage below .7V before
feeding it into a 741 Op-Amp that amplifies the ripple to
drive an LED directly through a bridge rectifier. The idea is
that the LED gets brighter as the ripple gets greater, or I
suppose an avid ripple-watcher could install a panel meter.

[Computer Nerd Kev]

On 24 Jun 2013, Ray Carlsen wrote:


> The C64 brick has no internal fuse in the 5 volt
> line. Some
> after-market supplies will have one or more however. The 5V
> regulator IC inside a brick is designed with something
> called "foldback current limiting".

Hmm, so if the filter capacitor does short, something melts.
I'm feeling more nervous about my Commodore PSUs every minute.

> I've only seen
> one that needed repairs. The best supplies are those with
> a switch-mode 5 volt PS installed and they are just
> starting to appear on the market. Switchers run cooler and
> don't suffer the same problems of the old analog power
> supplies and therefore don't need a Saver to protect the
> computer. However, as with any new product, I've seen good
> and bad (cheap) after-market switch-mode supplies. Buyer
> beware.

You'd have a great spiel if you decided to make new C64 power
supplies (a lot more work than the Saver, I know). I'm already
contemplating building one myself.

[Clocky]

Computer Nerd Kev wrote:
> On 24 Jun 2013, Ray Carlsen wrote:
>
>
>> The C64 brick has no internal fuse in the 5 volt
>> line. Some
>> after-market supplies will have one or more however. The 5V
>> regulator IC inside a brick is designed with something
>> called "foldback current limiting".
>
> Hmm, so if the filter capacitor does short, something melts.
> I'm feeling more nervous about my Commodore PSUs every minute.
>

They don't short that often. When sitting idle for a long time they
reform but it takes time and in my opinion that is when the ripple can
do damage to sensitive chips where the voltage output is still within spec.

>> I've only seen
>> one that needed repairs. The best supplies are those with
>> a switch-mode 5 volt PS installed and they are just
>> starting to appear on the market. Switchers run cooler and
>> don't suffer the same problems of the old analog power
>> supplies and therefore don't need a Saver to protect the
>> computer. However, as with any new product, I've seen good
>> and bad (cheap) after-market switch-mode supplies. Buyer
>> beware.
>
> You'd have a great spiel if you decided to make new C64 power
> supplies (a lot more work than the Saver, I know). I'm already
> contemplating building one myself.
>

I use an ATX supply with a 9VAC housed internally (found a spacious one
that had enough room).

I avoid the C= bricks wherever possible, they simply can't be trusted IMO.

[Computer Nerd Kev]

On 24 Jun 2013, Clocky wrote:

> I use an ATX supply with a 9VAC housed internally (found a
> spacious one that had enough room).

Is the 5V from an ATX PSU any more reliable than a Commodore
one in this set-up (no load on 12V)? I've stayed away from the
things in these cases after reading this website:
http://www.playtool.com/pages/psunonpc/nonpc.html

Never got around to measuring things myself though...

My approach would likely involve the merciless slaughter of a
couple of plugpacks and the addition of a 7805 Voltage
Regulator.

[Martijn van Buul]

* Computer Nerd Kev:

> On 24 Jun 2013, Clocky wrote:
>
>> I use an ATX supply with a 9VAC housed internally (found a
>> spacious one that had enough room).
>
> Is the 5V from an ATX PSU any more reliable than a Commodore
> one in this set-up (no load on 12V)?

I've found the behaviour of ATX PSU's to be rather unpredictable; some
will just fine without any loading on (both!) 12V rails, some won't
start, some will make a nasty, unhealthy whine.

In my experience, older ATX-1.3 PSUs (without the extra 12V rail for the
mobo) tend to be less finicky, and often require only a minimal load
on 12V. More often than not, the built-in fan will suffice.

That said, YMMV, and I would personally test things thoroughly before
connecting equipment that is slowly getting frail and hard(er) to repair :)

--
Martijn van Buul - pi...@dohd.org

[Clocky]

It' actually an AT power supply, I use an ATX one to power external
drives hence the confusion.

No problems with the supplies, the ATX one works needs very little load
to start working.

[rber...@iglou.com]

Ray Carlsen now offers three versions of the Computer Saver. There is the $25 Computer Saver module that a user can install in their own computer, there is the $35 Computer Saver cable (without LED indicators) version, and there is the $50, original, case-mounted Computer Saver (with LED indicators). To read more about Computer Saver and to see pictures, go to

http://personalpages.tds.net/~rcarlsen/cbm/c64/SAVER/

Truly,

[Andrew Wiskow]

I purchased two of the Computer Saver devices from Ray a couple months ago (I got the $50 model), and they've already saved one of my C64s from certain death due to a faulty power supply. They're definitely worth it!

I decided to get these mainly because I have a bunch of the old "brick" power supplies. They test good now, but who knows when they'll die and try taking a C64 along with them? With Ray's device, I can use the old "brick" power supplies without worrying about them. :)

-Andrew

[Clocky]

The saver does nothing to prevent voltage ripple damaging components,
which in my experience is a much more insidious killer of IC's then a
regulator dropping it's bundle which is an obvious fault with the supply.

Don't get me wrong, if people are going to use the junk that C= passed
off as power supplies the computer saver is a must have device.

Best advice would be to use a good quality modern supply, but ofcourse
the 9VAC component is what trips most people up there as that isn't
available with standard ATX supplies.

[rber...@iglou.com]

Ray Carlsen now has one webpage that lists all his newly-improved Computer Savers and adapter cables for Commodores. Go to

http://personalpages.tds.net/~rcarlsen/cables.html

The current prices are listed there. Check out the new mid-line Computer Saver for the Commodore Plus/4.

Merry Christmas,

Robert Bernardo
Fresno Commodore User Group

http://www.dickestel.com/fcug.htm