I have a RAMWorks III card I am trying to get running in a Platinum IIe, and am running into stability problems. The board is a REV E.
When this card is inserted, there is a faint but noticeable high-pitched sqeal from the power supply. Despite the noise, the card itself works --all memory passes diags (using AE's RAMWorks tester) and is recognized in Appleworks, etc. The IIe itself will pass diags as well using Apple's IIe diagnotics, and the system will in fact run normally for about 15 minutes or so.
The first sign of something going wrong (other than the PS noise) is a bizarre "echo"/ringing sound from the power supply when the speaker beeps. This happens after the system has been running for several minutes. Very shortly thereafter the system will start crashing with symptoms that resemble bad RAM (random characters showing up on an otherwise normal text screen, hangs). There are no abnormally warm components on the RAMWorks board when this happens.
I have tried the following:
* Depopulating the board (1MB->256K) -- reduces, but does not eliminate, the power supply noise. The system will still eventually fail as noted above.
* Replacing the RAM chips on the card -- no effect.
* Swapping in a spare power supply -- no effect.
* Tried the card a completely different IIe -- same behavior.
Based on the symptoms and my testing it seems the card is overloading the power supply, but the issue is cearly on the the card itelf and not the power supply. The Platimum IIe I want to use this in is loaded down with cards (SSC, PCPI Applicard, Trawswarp, Mouse, slinky card, Disk 5.25 controller, CFFA), and it is quite stable with all of these cards and the normal Extended 80-column card. I tried removing all cards except the RamWorks and Disk 5.25 controller and it made no difference. The other system I tried is an enhanced //e with no cards other than the disk controller.
Any ideas of where to begin? I do have a DVM, but no scope.
Thanks
-Jason
I would try to replace those two electrolitic capacitors... it's the easyest thing to try
Boy you sure covered all the bases narrowing it down to the card. Maybe with the card removed, do an ohmmeter check across all the little caps and the filter caps mentioned. I'm thinking all those are from the 5V to ground just for noise suppression. Maybe a low ohm resistance will show up.
Larry G
Just to turn back to the beginning:
Bear in mind that also a PCPI card and a slinky card is inserted besides the 80 col card.
That´s a very high bunch of load to the +5 Volt rail !!!
Adding up the RAMworks III with full population you clearly run up beyond the limits of the PS !
This causes the PS to heat up rather fast and then the regulation inside the PS fails
resulting to the freezes and crashs.
That´s clearly all normal behaviour.
The question is:
by driving the PS repeatingly beyond it´s limits - did you allready damage parts inside the PS ?
And besides:
driving the PS beyond it´s limits till malfunction also bears risk of damaging other parts
in the computer and not in the PS.......
I´d recommend to stop overloading the PS untill you have purchased a "heavy duty" powersupply !
The difference ?
the normal powersupplies in the IIe give you at +5 Volt about 6 Ampere to 8 Ampere
"heavy duty" powersupplies offer at the +5 Volt rail 10 Ampere to 15 Ampere !
That´s exactly the missing portion of power requested to run a "heavy loaded" system !
just my 5 cents....
sincerely speedyG
SpeedyG: He said I tried removing all cards except the RamWorks and Disk 5.25 controller and it made no difference
BubbaShakers: remove ALL the cards from the Apple IIe, remove all memory from the RAMWorks,check the back side of the ramworks board for bent/shorting pins, install the RAMWorks and turn on the Apple IIe and tell us what happens.
With all cards removed except for the depopulated RWIII, I do not hear any PS noise. Of course, no 80 column, and not much else I can do except run the IIe's self test, which of course passes. I next tried the same setup with only the first bank of 256K populated. At this point I can hear the squeal, but it's extremely faint (have to basically put my ear close to the PS to hear it). I expect the noise would get louder as banks are populated.
I should point out that I have not had the RWIII and the slinky in the system at the same time. The ultimate goal of this exercise is to replace the slinky with the RWIII and free up a slot. I'll probably end up getting one of Reactive Micro's new upgraded PS internals once they are available. As it is I run the system with the cover removed (the monitor III stand precludes a Kensington-style fan, and I much prefer a noiseless rig anyway).
When I first brought this system out of storage last year, it had only the standard "mini" extended 80 column card and disk controller. It ran for about 5 minutes before a RIFA cap blew. I bought one of jmorris's recap kits for the PS, but initially only changed out the RIFA caps (there are four of those buggers in this PS). At this point I tried adding the RWIII and a mouse card, and this is when I noticed the issue with the RWIII. Thinking I had further power supply issues I went ahead and replaced the remainder of the electrolytic caps. but this had no effect on the RW issue. The other 2 power supplies I have tried with this card (one of them in the other //e) are all original.
I have been running the system wth the cards mentioned earlier for a few months now without issue, except for the Transwarp which I got last month. After the issues with the RWIII I set it aside.
Thanks,
-Jason
Just a small riddle of aftermath:
1 41256 chip average poweruse : 0,080A
Ramworks III = 32 chips
slinky card = 32 chips
_______________________
total of 64 chips x 0,080 A = ??? Ampere
and that´s a calc leaving the avarage use by the TTL logic and the other chips beside....
and just another explenation:
what you call "chirping" of the PS is a tone that is issued by the so called "shortcut protection circuit"
it´s designed with several components that measure the load under "normal conditions"
when that circuit starts limiting / switching the protection function it results to that chirping noise.
In general it´s designed to simply switch off the output of the powersupply if it detects "too heavy load"
at the output that might indicate a shortcut.
But the most resistors in the powersupply are made of carbon and carbon might age very fast when exposed to heat.
Then that resistors may drift off from the determined value.
Another part within the powersupply is the transmission transformer with a copperwire isolated by a
protection paint and that paint gets damaged by high temperature and gets liquid.
When exposed to high heat that paint flows away and causes small shortcuts between the windings of copperwire
and thereby reduces the amount of isolated windings and changing value of the transformer.
Not to forget the common known issue with the electrolytic capacitors that get dry when exposed to heat or
if the case gets damaged by heat they start leaking and the bor-acid flows out of the cancontainer also changing the determined value of the capacitor.
So if the powersupply is forced to deliver far more current than it was designed for it heats very fast and
you start frying the resitors, toasting the transformer and cooking the capacitors. Then the "shortcut protection circuit" tries to defend the powersupply from damage.... but it´s only designed to give a short time protection
and demands to reduce load. If that does not happen the heat starts working and damages the components and they start
drifting of the desired values changing the electronic design. This causes changes that remain and there is no kind
of "self repair" ..... that drives the shortcut protection to even start it´s action more early because the
values of the design have changed by the change of the values of the components involved. The normal design of the shortcut-protection circuit is to protect and not to regulate ! The feedback line is designed for the regulation.
And at this explenation i didn´t even mention what happens to the rectifying diodes ( those mounted to the cooling sinks ) if they are forced to deliver more current than they have been designed for and get started toasting too......
maybe this explenation helps to understand what happened with your powersupply....
speedyG
Speedy:
I do have a question though regarding the shortcut protection circuit. If it is being tripped, would the machine continue to operate? I was under the impression the system would not operate at all if this circuit is tripped...
Thanks,
-Jason
Hello BubbaShakers,
well the note about the RAM chips was just a short point spotted....
at the PCPI card there is also RAM on board ( of course less chips ) but ading up
all usage together ( mainboard, 80 col card diskdrives ) and then adding the RAMwoorks III
( even if replacing the slinky ) it´s quite a load and the RAMworks itself - if you also add up the other chips -
well then that are 3 Ampere which bring the system close or a bit beyond it´s limits...
Ramworks III only with diskdrive should remain within normal usage.... but remember the PCPI also demands
a bunch of current...
specially remembering that Apple had "bad behaviour" and equipped it´s systems with "rather weak" powersupplies.....
up to the moment i´d guess that even a chip is that far damaged that it pulls more current than it should be expected...
maybe you can identify that source of trouble by a chip or component getting rather hot ?
And related to the question about the powersupply:
the moment it issues that chirping noise - it´s time to immediatly identify the trouble for that behaviour and
it´s a bad idea running the system and watching the PS getting hotter and hotter....
Under normal conditions the short cut protection circuit behaves like a switch that turns of the entire
output of + 5 Volt and + 12 Volt and thereby preventing system from starting up / or shutting the system down.
But if the PS has been running quite a time "at the limit" close before shutoff
it might get altered that way that it behaves like a fast switching trigger (making the chirping noise ).....
In that case the entire powersupply must be checked and several parts beyond "recapping" must be checked for correct function and demanded values and if needed they must be replaced....
fried resistors for example are in general indicated by the fact that below them the PCB colow turns slowly it´s
color to brown shading... and at the transformer or the capacitors you might spot small leakage or "shades of brown color" at the PCB.
I´ve seen this kind of behaviour only few times - but it happens...
sincerely
speedyG
One of the first things I tried was checking for overheating components on the RWII, but found none that were noticeably warm. I did try replacing the one socketed non-RAM chip on the board (a 74LS273), but that had no effect.
FWIW, the power supply is a Dynacomp and has a 4 amp rating on the +5V output (I've seen others with a 2.5 amp). When I recapped it, the only signs of charring or other heat damage was where the RIFA cap burned out. A couple of the original electrolytic caps near the front (output) section of the PS did show signs signs of leakage, but the board itself was not discolored.
While researching this issue I also ran across the following on a another forum, which may or may not be relevant:
On one hand my card is a REV E, but I am assuming there are folks using Rev E's without issue in their systems, else these would almost certainly have been recalled...
-Jason
Bubbashakers,
You already said that You tried with different power supplies and that it did not change anything and I take it for granted so the following is just to continue talking about the problem:
when You recapped the power supply, did You replace the caps following the polarity of the old caps or did You follow the silk screen "+" sign on the PCB? I ask because I did recap mine and I found 3 caps whose polarity was opposite to the one marked on the pcb; obviously I followed the polarity marked on the capacitors.
Anyway, if You accidentally reverted some capacitor polarity, I don't think the power supply would ever work.
Luca
IIRC, most were marked on the board, but a few were not, so I had to examine the originals. I am fairly confident this is not an issue in this case, as reversing the polarity on an electrolytic will cause an immediate (and usually audible) failure when the system is powered up. Ask me how I know
-Jason
Just to push things, I stuck a Rev E 1mb RamWorks in a IIe with the original power supply and a Transwarp, mouse card, a 5.25” FD card. I gradually added two 1mb RamFactors and two 1mb Apple slinky cards. Ran it for a couple of hours, no chirping but I must be pretty darn close.
I’m assuming you’ve cleaned the edge connector and examined the card for physical damage so:
Since my Rev E doesn’t have an issue then I’d discount the board layout. Since you’ve tried the card in 2 IIes and replaced all the socketed components then I’d say that it has to be a faulty soldered on part. Caps would be the easiest to replace.