Jump to content
IGNORED

65xe repair and troubleshooting advice


shoestring

Recommended Posts

 

Hi all.

 

I have picked up a dead and battered 65xe that produces nothing but a black screen.

 

It has what appears to be 2 x 64Kx4bit DRAMs ( MT 4067-12 ). Are these just 4464s? The data sheet seems to confirm the same pinout. I'm still getting nothing after replacing them with known working DRAMs which are a little faster. FYI I'm using a known working power supply which I use for my other Atari's.

 

I tested the MT-4067 DRAMs out of circuit and they seem completely dead so I'm suspecting that a faulty power supply took these out and perhaps other circuitry as well. The 40 pin chip at the bottom left is stone cold compared to the other chips which run warm, looking at the circuit it appears that this chip is some kind of DRAM address decoder??

 

I'm just looking for some advice on probably causes before I start doing some troubleshooting of my own as I have no experience with the XE series and I'd like to get the memory in this upgraded eventually and the exterior cleaned up :)

 

Thanks in advance.

 

Some pics of the machine and my current progress.

 

f4323f72c4bf17509f8b39029c838202.jpg

 

0d0bba061e81754ae8b51d4fc173f347.jpg

 

523337d4065e251082367e6c3044262f.jpg

Link to comment
Share on other sites

They're just standard 4 bit 64k DRams. The MT brand are notorious as being unreliable to the point of some people replacing them even if they work OK.

 

Good practice is to put sockets in before the replacement chips. And use a known good power supply, there are some variants that will go overvolt when they get old and kill systems. Unwanted buzzing in the audio can be a symptom of a PS that's bad.

 

To get 128K you'd also need an EMMU chip which goes into that lone socket below/right of the Ram sockets. Not sure if that's the complete requirement, might be a few passive components needed as well.

Link to comment
Share on other sites

It has what appears to be 2 x 64Kx4bit DRAMs ( MT 4067-12 ). Are these just 4464s? The data sheet seems to confirm the same pinout. I'm still getting nothing after replacing them with known working DRAMs which are a little faster. FYI I'm using a known working power supply which I use for my other Atari's.

 

I tested the MT-4067 DRAMs out of circuit and they seem completely dead so I'm suspecting that a faulty power supply took these out and perhaps other circuitry as well. The 40 pin chip at the bottom left is stone cold compared to the other chips which run warm, looking at the circuit it appears that this chip is some kind of DRAM address decoder??

 

Yes, the MT-4067 are the same like 4464 or 41464. When you changed them already and system won´t start, the suggestion of using a wrong power supply by former user(s) is the most possible reason. I´ve had a lot of dead XL/XE on my desk which were try to power-up with a Commodore C64 power supply... instant dead will follow. :-o

 

The 40 pin IC down below the DRAMs is the Freddie. Freddie ist just a simple ASIC which containts some DRAM access logic and clock divider. In most cases the Freddie is also blown out when using wrong power supplies like the C64 one (this one has 9V AC... ). The reason here is, that Freddie is manufactured in CMOS, all the other custom chips from Atari are manufactured in NMOS. Most times all NMOS chips survice such a "Commodore attack", but DRAMs (also most CMOS in the XE series with two 41464) and Freddie are dead.

 

If you own an oscilloscope, check the following points...

 

- Pin 7 of the CPU (SYNC). Should be some transitions visible after RESET. Maybe flat line at 0V after some cycles (CPU hung up). If absolutely NO activity after RESET / Power-Up, no PHI0 clock is provided to CPU

- Check OSC output of Freddie (Pin 37). This is the "master clock" for ANTIC, this one will generate PHI0 for CPU. OSC should be around 3,5xx MHz. If no transistions visible, Freddie is dead

- Check pin 33 (RAS) and 35 (CAS). RAS must be constantly active, CAS also. Also no transistions = Freddie dead

 

Without oscilloscope my Sys-Check might be helpful, but of course not in all situations.

 

Good luck, Jurgen

  • Like 2
Link to comment
Share on other sites

They're just standard 4 bit 64k DRams. The MT brand are notorious as being unreliable to the point of some people replacing them even if they work OK.

 

Good practice is to put sockets in before the replacement chips. And use a known good power supply, there are some variants that will go overvolt when they get old and kill systems. Unwanted buzzing in the audio can be a symptom of a PS that's bad.

 

To get 128K you'd also need an EMMU chip which goes into that lone socket below/right of the Ram sockets. Not sure if that's the complete requirement, might be a few passive components needed as well.

 

The MT DRAMs were so far gone that my EPROM programmer ( Micromaster LV48 ) couldn't recognize them and was prompting to check device and position ( as though no device was installled in it ). Normally it will either pass tests or fail when it reads back the wrong value at a given address.

 

I put some sockets in this already and installed some TI branded DRAMs.

 

I'm more interested in upgrading it to 512kb which requires an SRAM package, it's requires more work but is more worth while than the 128kb mod.

Edited by shoestring
Link to comment
Share on other sites

 

Yes, the MT-4067 are the same like 4464 or 41464. When you changed them already and system won´t start, the suggestion of using a wrong power supply by former user(s) is the most possible reason. I´ve had a lot of dead XL/XE on my desk which were try to power-up with a Commodore C64 power supply... instant dead will follow. :-o

 

The 40 pin IC down below the DRAMs is the Freddie. Freddie ist just a simple ASIC which containts some DRAM access logic and clock divider. In most cases the Freddie is also blown out when using wrong power supplies like the C64 one (this one has 9V AC... ). The reason here is, that Freddie is manufactured in CMOS, all the other custom chips from Atari are manufactured in NMOS. Most times all NMOS chips survice such a "Commodore attack", but DRAMs (also most CMOS in the XE series with two 41464) and Freddie are dead.

 

If you own an oscilloscope, check the following points...

 

- Pin 7 of the CPU (SYNC). Should be some transitions visible after RESET. Maybe flat line at 0V after some cycles (CPU hung up). If absolutely NO activity after RESET / Power-Up, no PHI0 clock is provided to CPU

- Check OSC output of Freddie (Pin 37). This is the "master clock" for ANTIC, this one will generate PHI0 for CPU. OSC should be around 3,5xx MHz. If no transistions visible, Freddie is dead

- Check pin 33 (RAS) and 35 (CAS). RAS must be constantly active, CAS also. Also no transistions = Freddie dead

 

Without oscilloscope my Sys-Check might be helpful, but of course not in all situations.

 

Good luck, Jurgen

 

' Commodore attack" :D . I suspect a power supply was responsible for this.

 

A work colleague of mine picked up this 65xe and an 800xl off eBay. He described a buzzing sound and rolling picture ( sound familiar ?? ).I ended up repairing the 800xl which also had a single DRAM fault at U9, I replaced the whole stack of 8 with m3764s. The machines came with 3 power supplies, one of them was C70042-1 'The Logo' and the other two were identical but I had never seen those types before. All encased in epoxy and non repairable. Some vandal had also been poking around inside the 65xe and destroyed the variable resistor at R38, I suspect this is for the colour adjustment.

 

After testing the 800xl & repairing it, I salvaged the leads+7 pin din connectors for a switching adapter solution and tossed the power supplies in the bin. I was rewarded with the 65xe for help with the 800xl, so now it's mine to fix. I have a really nice Tektronix TDS3054 scope to get the job done.

 

Thanks for the troubleshooting points, I'll post some results once I make more progress.

Edited by shoestring
  • Like 3
Link to comment
Share on other sites

@shoestring:

 

It sounds like you had "The Ingot" supplies. The epoxy filled bricks should all be thrown in the bin before they do damage.

 

I highly recommend the U1M for a RAM upgrade. Just put 64K back in and get a U1M. You should also consider the SIDE2 cart because of how it works with the U1M. It becomes a bootable PBI device which can load disk images from a FAT32 CF card. In other words, you can download some Atari stuff on your Windows system and directly copy it to the CF card. Very easy to transfer files.

  • Like 1
Link to comment
Share on other sites

@shoestring:

 

It sounds like you had "The Ingot" supplies. The epoxy filled bricks should all be thrown in the bin before they do damage.

 

I highly recommend the U1M for a RAM upgrade. Just put 64K back in and get a U1M. You should also consider the SIDE2 cart because of how it works with the U1M. It becomes a bootable PBI device which can load disk images from a FAT32 CF card. In other words, you can download some Atari stuff on your Windows system and directly copy it to the CF card. Very easy to transfer files.

 

Thanks Kyle, I'll keep that in mind as well :)

 

Pictures of the two power supplies from the auction.

post-45355-0-97829900-1478150514_thumb.jpg

Link to comment
Share on other sites

@shoestring:

 

It sounds like you had "The Ingot" supplies. The epoxy filled bricks should all be thrown in the bin before they do damage.

 

Just a note... The "ingot" ones are only availible in the U.S. (or in other words, for usage in 110/115 volt areas). I didn´t know, if the european/australia models (220/230 volts) also were produced by "ingot", but I have had dozens of them, and never one with such a behavior like often read here. The 220/230 volt models have the typical issues like all other XL/XE power supplies:

 

- They work :)

- They didn´t work - no output

- They have defective electrolytic capacitors side. Output voltage running idle is fine, but when a computer connected and switched on, it breaks down to something less then 4 volts with visible voltage dropdown on the scope

 

Lower voltages didn´t kill something on the Atari, mostly such defective power supplies results in non-booting computers with red/brown screen. ANTIC, GTIA and CPU (clocks) may be able to generate a video signal until the effective voltages drops less then 4,2V.

 

Jurgen

Edited by tf_hh
Link to comment
Share on other sites

Interesting weekend so far.

 

So I socketed all the chips in the 65XE and did a little troubleshooting. ( By the way this is a PAL System )

 

Here's what I found.

 

Custom chips & roms tested by substitution in known working machine. ( A 600xl )

 

7400/4000 series chips all tested good.

Basic and OS rom tested good

 

ANTIC: CO21698 Tested good

GTIA: CO14889 Tested good

POKEY: Good too

PIA/6520: Also good

The new 2x4464s: No problems

 

6502 tested bad in the 600xl, brain farts an executes random addresses

MMU tested bad ( red screen ) ( I have some spare GAL16V8s to fix this )

 

With the good MMU & CPU installed in the 65xe from the 600xl still doesn't fix it.

 

So that obviously isn't the only thing wrong with it

 

I haven't trouble shooted FREDDY with the scope just yet, but I suspect it's dead.

 

We will see.

 

My 600xl has developed a problem whilst using it as a donor.

Keys 6,7,Y,U & N aren't working now.

 

Edit: KB fixed. Bent pin on the 4051 mux will do that :)

Edited by shoestring
  • Like 1
Link to comment
Share on other sites

 

If you own an oscilloscope, check the following points...

 

- Pin 7 of the CPU (SYNC). Should be some transitions visible after RESET. Maybe flat line at 0V after some cycles (CPU hung up). If absolutely NO activity after RESET / Power-Up, no PHI0 clock is provided to CPU

- Check OSC output of Freddie (Pin 37). This is the "master clock" for ANTIC, this one will generate PHI0 for CPU. OSC should be around 3,5xx MHz. If no transistions visible, Freddie is dead

- Check pin 33 (RAS) and 35 (CAS). RAS must be constantly active, CAS also. Also no transistions = Freddie dead

 

Without oscilloscope my Sys-Check might be helpful, but of course not in all situations.

 

Good luck, Jurgen

So I followed your tips and it seems I'm getting the right signals with my scope. See attached photos.

 

I also followed the troubleshooting guide from the 130xe service manuals. Microprocessor operation and Clock and dividers. I couldn't find any issues.

 

I'm going to remove these sockets again to see if I've gauged any traces or damaged any plate through-holes.

 

CPU clock Phi0 - pin 37

4ca4cbb965197939969afab99694fb06.jpg

 

CPU sync - pin 7

3b9898e991c88e1a77f6613f526f03cc.jpg

 

Pin 37 - FREDDIE

ee1a2518e0fed6b99c2fb28a87875628.jpg

 

Pin 35 - CAS & RAS

 

3065172bd97d9774a9446edea49c02b7.jpg

 

55a29734f383f8d8996e94bc9583567a.jpg

Edited by shoestring
Link to comment
Share on other sites

So I followed your tips and it seems I'm getting the right signals with my scope. See attached photos.

 

I also followed the troubleshooting guide from the 130xe service manuals. Microprocessor operation and Clock and dividers. I couldn't find any issues.

 

I'm going to remove these sockets again to see if I've gauged any traces or damaged any plate through-holes.

 

That´s a good idea, too. Ok, looking at the scope pictures the signals you show are good. CPU doesn´t deadlock, so CPU, major chips and O.S. ROM seems to be fine. Makes it not easier...

 

In the attachment you will find the XE System schematics drawn by Jerzy (don´t know if you have them already).

 

Examine every signal at freddie related to the schematics. Also MMU. When you´re lucky, you wil find a missing or bad (high level below 2,5 volts or something like that) signal.

 

Good luck, Jurgen

xesys.zip

Link to comment
Share on other sites

Thanks for that. I'll take a look.

 

FREDDIE definitely has an issue though, I forgot to mention that at times activity was nil on the said lines, nil video sync and measuring the Y1 crystal directly revealed a very weak sine signal ( 2.5v max ) in that instance.

 

Now FREDDIE is completely dead and I just managed to accidentally snap a leg off the Y1. So I'm up for another ( c070034 ) to go with the new FREDDIE chip..

 

I have a spare from a PAL Apple //e but I think the frequency is 14.25mhz and is a little off.

Edited by shoestring
Link to comment
Share on other sites

FREDDIE definitely has an issue though, I forgot to mention that at times activity was nil on the said lines and measuring the Y1 crystal directly revealed a very weak sine signal ( 2.5v max ) in that instance.

Now FREDDIE is completely dead and I just managed to accidentally snap a leg off the Y1. So I'm up for another ( c070034 ) to go with the new FREDDIE chip..

I have a spare from a PAL Apple //e but I think the frequency is 14.25mhz and is a little off.

 

The computer will start with the wrong crystal, but you won´t get any colour or wrong colours. But for testing it´s ok, also you can use a 14,318 MHz crystal, which is common.

 

If you doesn´t need anything from Bradley (BEST Electronics) sent me a PN, I will put a crystal in a letter and sent it to you. Have some dozens of them ;-)

 

Jurgen

Link to comment
Share on other sites

Yes, it´s one of the typical NTSC base frequencies, but the Atari will still work with that. Only the GTIA won´t produce correct PAL color clocks, but nothing can be damaged, you get just a b/w picture.

 

The voltage level directly measured at the crystal is very low, mostly between 0.8 and 1.2 volts. This is normal and no indicator for a defect.

 

FREDDIE or MMU are the first things I would try also. The MMU can be exchanged by a 16V8 GAL. The GAL code incl. JEDEC-file in the attachment ;-)

 

 

XEGS_MMU_GAL_Replacement.zip

Link to comment
Share on other sites

Thanks for letting me know, would have never guessed.

 

That might explain not being able to get the computer to boot when I had half a chance.

 

With the PLD reprogrammed and installed nothing happens as I suspected. No consistent video sync or timing signals. Sometimes it will splutter video as I see it on the screen for second or two then it will stop.

 

There are a couple FREDDIE chips floating around, I'll order one and let you know how I go. I looked at Best Electronics but he has a min order of $50 and I don't need that much stuff.

Link to comment
Share on other sites

Thanks for letting me know, would have never guessed.

 

That might explain not being able to get the computer to boot when I had half a chance.

 

With the PLD reprogrammed and installed nothing happens as I suspected. No consistent video sync or timing signals. Sometimes it will splutter video as I see it on the screen for second or two then it will stop.

Just in case, here is the source of the XL's c061618 MMU equations (PAL & GAL). In the zip you'll also find an image with the MMU's pin layout and a textfile describing the function of the pins. This info was uploaded by Bob Woolley a few years back.

I have included similar files about the XE's c025953 EMMU as well, sadly can't remember where I got them.

 

Hope this helps!

 

re-atari

 

Atari MMU (C061618).zip

 

Atari EMMU (CO25953).zip

Edited by re-atari
  • Like 3
Link to comment
Share on other sites

  • 3 weeks later...

Thanks to @tf_hh for the crystals.

 

So I tracked down the intermittent video to a dry joint(s) in the actual RF modulator, I re-flowed everything there with new solder so now the composite signal is completely stable and I no longer have to tap the PCB to get a solid picture :)

 

A NOS Freddie chip arrived, installed that but still get a solid red screen except now the picture doesn't roll after a few seconds like it did with the old Freddie chip.

 

I'm suspecting the gal16v8D device I'm using for the MMU might not be compatible in some way...

 

The quest continues :)

Edited by shoestring
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...