Apple IIe - strange clock problem

25 posts / 0 new
Last post
Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
Apple IIe - strange clock problem

I've got a US Apple IIe Rev B. 64K MB that is DOA, was given to me.  Gradually I've been troubleshooting why it doesn't initialize.  The PSU is good, power light turns on, voltages are clean and proper values on all filter caps, but no beep on power-up, just a brief flash of about 2" of lines at bottom of Apple  Monitor /// then darkness.  I'm testing bare motherboard, nothing in slots.

1.  On power-up, pin 40 on CPU stays low, it does not toggle low from high and back to high.

2. All clock signals are present and correct frequency at clock timing circuits,  they swing ~5v p-p roughly, but they are 2.5v above and 2.5v below ground. (this has to be a major clue).

3. The 74LS125 tests good

4. I'm preparing to test the 74S02 next, which has no clock on pin 5 but a good clock on pin 6.

 

These off-centered clock signals do appear at pins 14, 15, and 16 of the HAL.

 

Anyone run into this before?

 

Thanks

 

Online
Last seen: 1 hour 12 min ago
Joined: Jun 18 2010 - 13:54
Posts: 586
Well, if pin 40 (RESET) stays

Well, if pin 40 (RESET) stays low, then it's never going to boot. Reset comes from the IOU but it also depends upon correct clocks. So you really need to start there.

Offline
Last seen: 8 hours 28 min ago
Joined: Apr 26 2016 - 08:36
Posts: 479
If you're not cycling the

If you're not cycling the reset signal there may be something wrong with the transistor or the 555 timer in the reset circuit.

Have you pressed ctrl-reset and watched pin 40 on the CPU while doing so?  Any changes?

Offline
Last seen: 8 hours 28 min ago
Joined: Apr 26 2016 - 08:36
Posts: 479
Come to think of it...that

Come to think of it...that circuit above is from the II+, not the IIe.

Int he IIe, the reset signal is generated by the IOU chip.  So something may be amiss there.

 

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
baldrick wrote:If you're not
baldrick wrote:

If you're not cycling the reset signal there may be something wrong with the transistor or the 555 timer in the reset circuit.

Have you pressed ctrl-reset and watched pin 40 on the CPU while doing so?  Any changes?

[[{"fid":"35605","view_mode":"default","fields":{"format":"default","alignment":"","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und]

@jeffmazur is exactly right.

 

Thank you Baldrick, I've been meaning to check Pin 40 while pressing ctrl-reset but that takes more hands than I have.  So my plan is to take a mini-clip and connect between pin 40 and my scope probe so my two hands for free for the ctrl-reset.

 

I'll post results.

 

I taken o'scope screenshots of all the good clocks of a working IIe with my Siglent.  I'll post if anyone is interested.

 

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
LaserMaster wrote:baldrick
LaserMaster wrote:
baldrick wrote:

If you're not cycling the reset signal there may be something wrong with the transistor or the 555 timer in the reset circuit.

Have you pressed ctrl-reset and watched pin 40 on the CPU while doing so?  Any changes?

[[{"fid":"35605","view_mode":"default","fields":{"format":"default","alignment":"","field_file_image_alt_text[und][0][val

Baldrick, this is a IIe and does not have the 555 timer circuit like the Apple II, but pressing the Ctrl+Reset button does cause the 6502 pin 40 line to drop slightly.  But since this line is already low with lots of noise the Ctrl=Reset drop is in the millivolts, so some 'bad actor' is holding this line low.

 

I tried uploading a small jpg scope image but it never appears in the "saved" reply.

Online
Last seen: 1 hour 12 min ago
Joined: Jun 18 2010 - 13:54
Posts: 586
You need to check the signals

You need to check the signals on pins 25, 26, and 27 of the IOU. They all should have roughly 5V pulses. If not, find out why.

Offline
Last seen: 5 hours 1 min ago
Joined: Jul 5 2018 - 09:44
Posts: 1558
Hope it isn't the IOU or MMU

Hope it isn't the IOU or MMU chip...  Those are basically unobtanium.  All the other chips on a //e motherboard are available.  The only way to get an IOU or MMU is to get it from a board being parted out.

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
A sleuth's update-IIe Rev B.

Thanks jeffmazur and softwarejanitor for your suggestions and hopes.  I've found time to spend sleuthing and this is what I've done and found.

1. Pulled the 74LS125 and tested each tri-state buffer - no problem found

2. Pulled the 74S109 and recreated the IIe Dual JK Flip-Flop circuit fed by the 14Mhz oscillator output.  The 1st flip-flop (pins 1 thru 7) work just fine, the 2nd flip-flop fed by the 1st is bad.

3.  With the 74S109 socket in the IIe empty, I o'scoped the 14Mhz oscillator output at the top of R34 or pin 12 of the 74LS125.  This time the oscillator waveform showed to be uni-polar and the correct frequency, although its amplitude peak is slightly lower than the expected, it gets a boost by the tri-state buffer (it feeds into) to a more acceptable TTL level.  These two waveforms are attached.

For those of you unfamiliar, the lower left side of the scope screen has a yellow pointer with a number "1" in it.  This represents 0 volts or Ground for scope probe "1".

 

Pin 12 of 74LS125 or top of R34

 

Pin 11 of 74LS125

 

My conclusion here is that the 14Mhz oscillator circuit is capable of doing its job and the tri-state buffer gate boosts the signal to a good level, as seen on pin 11 of the 74LS125 above.  This output is feed directly into the first JK flip-flop of the 74LS109.  Together, both JK flip-flops make up TWO divide by 2 frequency dividers, yielding the 7Mhz and a 3.58Mhz clocks.  The Q output from the 1st flip-flop feeds the 7Mhz clock into pin 2 of the HAL/PAL chip and the K input of the 2nd flip-flop's pin 14 (as well as fedding a pin in the memory expansion slot 0) while the 2nd flip-flop Q output provides the 3.58Mhz clock for pin 3 of the HAL/PAL chip and pin 35 on Slot 7.  But since the 2nd Q output is always at a TTL High level, the HAL/PAL cannot do its job.

 

What are the chances that the only thing wrong with this motherboard is a bad 74LS109?  Like softwarejanitor said "Hope is isn't the IOU (HAL/PAL chip)..."  For you old timers out there "Eight Ball Says: Outlook Is Doubtful", but one can hope. 

We shall see.

 

 

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
correction:

In the last paragraph of my last post I quoted softwarejanitor with a typo.  It should have read "Hope it isn't the IOU..." or the HAL/PAL chip.

My apologies.

Online
Last seen: 1 hour 12 min ago
Joined: Jun 18 2010 - 13:54
Posts: 586
My money is on the 74S109.

My money is on the 74S109. Replacing that might just do the trick!

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
jeffmazur wrote:My money is
jeffmazur wrote:

My money is on the 74S109. Replacing that might just do the trick!

 

I hope you are right 'cause that's my hunch since everthing else looks good.  Here is how it came to me with its PSU out for testing and cleaning.  This came with the case but no lid and 3 or 4 of the air vent sections broken out of the top sides.

 

 

And here it is after a bit of cleanup.

 

 

I popped the motherboard off to photograph and visually inspect all the bottom traces last night and it looked perfect.

 

Will get an order in to Mouser for some LS109's (and to have some on hand) and some other incidental soon and once I have them in hand I'll report back to this post in a few weeks.

 

Online
Last seen: 1 hour 12 min ago
Joined: Jun 18 2010 - 13:54
Posts: 586
 LaserMaster wrote:Will get

 

LaserMaster wrote:
Will get an order in to Mouser for some LS109's

 

 

Just to clarify: the original is a 74S109. Not sure an LS109 would work reliably (or at all). You should replace with the original Schottky series part.

CVT
CVT's picture
Offline
Last seen: 1 hour 5 min ago
Joined: Aug 9 2022 - 00:48
Posts: 328
Yes, you don't want to

Yes, you don't want to replace it with 74LS109 - at around 30 nanoseconds it's way too slow. You can put either 74S109 or 74F109. My Apple IIe has the latter. They are both 5 nanoseconds or under.

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
Thanks jeffmazur and CVT, I

Thanks jeffmazur and CVT, I was aware, just made a typo.  I ordered the 74F variant.

 

I'm not getting email notifications on updates to this thread.  I get notices of other Apple II posts and updates but not this one.  

I checked my profile and see no settings to change with regard to this.

Any ideas?

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
Results after replacing 74S109 with 74F109

So, its sort of good news...not what I was hoping...but good results.  With the new 74F109 the 2nd stage Q output is restored that drive the HAL pin 3.  The clock signal fon the HAL Pins 2 is also good.

The HAL clock outputs on Pins 14, 15 and 16 are also now spot on.   But there is still no Beep on power-up and no dispaly on the Apple /// Monitor.  O'scope screenshots are shown.

 

HAL-Pin3 clock signal

 

HAL-Pin2 clock signal

 

HAL Pin 14 clock signal

 

HAL-Pin15 clock signal

 

HAL Pin16 clock signal

 

Since the CPU Pin40 still doesn't behave properly, it is LO when powered on, barely bumps slightly above that and returns to LO it leads me to believe the IOU is bad.  At least now the CPU Pin40 is now a clean LO signal and the Apple clocks are now working properly.  I have two IIe Rev B's that work, but hesitant to risk using one for IOU chip substitution.

 

CVT
CVT's picture
Offline
Last seen: 1 hour 5 min ago
Joined: Aug 9 2022 - 00:48
Posts: 328
LaserMaster wrote:...I have
LaserMaster wrote:

...

I have two IIe Rev B's that work, but hesitant to risk using one for IOU chip substitution.

 

With two working fully socketed Apple IIe's what are you really risking?? You can use one of the working motherboards to test individual chips from the broken motherboard one at a time, which is the safer approach as opposed to putting good chips in the broken motherboard. If you are that worried, you can even do the suspect unobtainium last, or not do it at all and confirm it through elimination. Another low risk approach is to swap all unobtainium chips as a group.

 

After all, this is the huge advantage of having multiple fully socketed Apple IIe's.

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
CVT wrote:With two working
CVT wrote:

With two working fully socketed Apple IIe's what are you really risking?? You can use one of the working motherboards to test individual chips from the broken motherboard one at a time, which is the safer approach as

 

If it were only that simple, CVT.   I have given lots of thought in the recent past to what you suggest.  I just not there yet, since I don't really need this one working.  It could just be a parts IIe.

 

 

 

 

 

 

 

Offline
Last seen: 5 hours 1 min ago
Joined: Jul 5 2018 - 09:44
Posts: 1558
LaserMaster wrote:CVT wrote
LaserMaster wrote:
CVT wrote:

With two working fully socketed Apple IIe's what are you really risking?? You can use one of the working motherboards to test individual chips from the broken motherboard one at a time, which is the safer approach as

 

If it were only that simple, CVT.   I have given lots of thought in the re

 

 

I've got a couple of parts //e units...  given some of the unobtanium parts, it isn't a bad thing to have, especially spare IOU,MMU and AY-5-3600-PRO chips.

 

Offline
Last seen: 1 week 4 days ago
Joined: Jun 25 2020 - 17:00
Posts: 102
softwarejanitor wrote:I've
softwarejanitor wrote:
I've got a couple of parts //e units...  given some of the unobtanium parts, it isn't a bad thing to have, especially spare IOU,MMU and AY-5-3600-PRO chips.

Yessir, couldn't agree more provided  there is enough space.

CVT
CVT's picture
Offline
Last seen: 1 hour 5 min ago
Joined: Aug 9 2022 - 00:48
Posts: 328
softwarejanitor wrote:I've
softwarejanitor wrote:
I've got a couple of parts //e units...  given some of the unobtanium parts, it isn't a bad thing to have, especially spare IOU,MMU and AY-5-3600-PRO chips.

 

I got a perfect KR9600-PRO for $6 from AliExpress just last week, so the keyboard encoder definitely doesn’t belong to the unobtainium group.

Offline
Last seen: 5 hours 1 min ago
Joined: Jul 5 2018 - 09:44
Posts: 1558
CVT wrote:softwarejanitor
CVT wrote:
softwarejanitor wrote:
I've got a couple of parts //e units...  given some of the unobtanium parts, it isn't a bad thing to have, especially spare IOU,MMU and AY-5-3600-PRO chips.

 

I got a perfect KR9600-PRO for $6 from AliExpress just last wee

 

Could have been a one in a million thing though.  I suspect you got one salvaged from e-waste.  I'd be really surprised if you were to order, say 10 of them if you'd get 10 working ones.  I can't imagine where any such stash of properly programmed //e chips would come from.

 

Didn't you order some others from eBay or somewhere and get bogus ones?

 

 

 

CVT
CVT's picture
Offline
Last seen: 1 hour 5 min ago
Joined: Aug 9 2022 - 00:48
Posts: 328
The change of something being

That is not the case. The chance of something being one in a million is... one in a million.

 

I got a couple of unprogrammed AY-5-3600-PROs from eBay and then a bad AY-5-3600-PRO from AliExpress. But then I decided to go with KR9600-PRO, since I noticed that they don't seem to come unprogrammed and people that get them leave good reviews. And the first one I got worked. Also I found many places that sell the KR9600-PROs and have them in stock. Even if half of them are bad, they would not be considered unobtainium, since at this price you can get 2 or 3 and have a very good chance that at least one works.

 

Not sure where the seller on AliExpress gets them from, but he has listen more than 27,000 in stock. Also I tried adding 100 to my cart and it allowed me, which means I can place an order for at least that many. If I do and the seller doesn't have them, it's a problem for him as he has to cancel the order and incur a penalty form AliExpress. This leads me to believe that he's got quite a few. The one I got was SMC manufactured in 1989 and from what I've read the Platinum came with this chip. It is most likely from e-waste, but usually 80% of e-waste chips are good based on my own experience on a wide variety of different chips.

Offline
Last seen: 5 hours 1 min ago
Joined: Jul 5 2018 - 09:44
Posts: 1558
CVT wrote:That is not the
CVT wrote:

That is not the case. The chance of something being one in a million is... one in a million.

 

I got a couple of unprogrammed AY-5-3600-PROs from eBay and then a bad AY-5-3600-PRO from

 

Well, it's certainly good news that this part is not unobtanium.  The question is that if these are coming from e-waste from scrapped //e motherboards, then there should also be a similar amount of IOU and MMU chips.  And that's not the case as far as I can tell.  I haven't seen anyone selling them, pretty much at all.  That either means that there was some kind of surplus overrun of KR9600-PRO chips that were programmed for a //e or there is coincidentally some other application which was compatible.  Just doesn't make sense that there would be either...  but if you can indeed order 27k of them or even 100...  It's interesting.  If there was a source for IOU and MMU chips or FPGA/CPLD compatible ones then someone could make a batch of brand new //e motherboards...  Maybe not economically feasible but interesting.

 

CVT
CVT's picture
Offline
Last seen: 1 hour 5 min ago
Joined: Aug 9 2022 - 00:48
Posts: 328
Unlike the keyboard encoders

Unlike the keyboard encoders used in the Apple II+ and Apple III , the 9600-PRO chips can either be empty or programmed in only one particular way, which is described in the datasheet. This makes them more generic and allows the keyboard mapping to be in a separate EPROM, instead of inside the chip itself.

 

Why we cannot get the IOU or MMU through e-scrap and why any search for them on findchips.com, eBay or AliExpress returns absolutely nothing is a really good question. I would speculate that this happens due to the following contributing factors:

 

  • The keyboard encoder seems to be always socketed, while the IOU and MMU are mostly soldered. E-scrappers usually ignore soldered chips – they are not worth the trouble.
  • E-scrappers ignore chips with the copyright symbol, as they are usually one-time-write PROMs or custom chips that are too specific to be used anywhere else.
  • E-scrappers don’t have any knowledge of the millions of different possible devices a particular board can come from. All they see is generic chips that can be sold and custom chips that cannot.
  • The keyboard encoder has all the telltale signs of a generic chip. It has a public datasheet that fully describe its function, it has no copyright symbol and a quick Google search shows that at least one other production board uses it: click!

 

 

Log in or register to post comments