how to test a megaST

[Bikerbob]

Trying to repair a Mega.

 

Using a diag cart.. I get a T4 memory error.. removed all socketed chips and reseated. Same issue. These plcc chips do not play well with a std PLCC chip puller.. not liking it. So would prefer to not be swapping chips in an out all after noon.

 

If anyone has suggestions as to what to test next.. very open to ideas.

[+tf_hh]

On 11/16/2019 at 5:34 AM, Bikerbob said:

Using a diag cart.. I get a T4 memory error.. removed all socketed chips and reseated. Same issue. These plcc chips do not play well with a std PLCC chip puller.. not liking it. So would prefer to not be swapping chips in an out all after noon.

 

T4 is not a memory issue. Memory problems are shown with Rxxx, Txxx means Timing. According to the manual T4 means: "

 

 

Myself didn´t have had this error-code yet. I would suggest replacing the following parts for a test:

 

- MMU

- Shifter

 

If this doesn´t help, I would start to change the memory databus line transceivers. Two 74LS244 and two 74LS373 are found left hand where the RAM is placed. Most problems with interaction between RAM and Shifter are located there. You can use HCT types instead of LS for replacement (mostly easier to get these days).

 

[Bikerbob]

Thanks tf_hh  I have this thread going on Exxos board - I have replaced both the Shifter and the MMU - both those chips work fine in KNOWN working Mega.. so thats not it. ( I in fact 1 at a time replace all socketed chips)

 

BUT - after the suggestion of corrosion or bad solder joints. I removed the backplane and the lower plate and looked at the bottom. Although very clean.. well maintained.. I can see joints that have been reflowed? maybe a part or two replaced? also arrow marks on solder points that I guess were suspected to poor.  My point.. this board has been through this before.

 

On examination I do not see any BAD joints.. but I guess I could reflow the joints of the MMU and the SHifter??  would that be a suggestion before going at the 74LS?    The 373s are F types.. which I hear are better. Can they be tested? or do you just RE-RE and thats the best way?

 

James 

[guus.assmann]

Hello Bikerbob,

 

It is very possible that the revision of the test cartridge is the problem and not the hardware itself.

I've had this issue, until I used another (multi) testcartridge. It might need a test cable or another fixture to pass this timing test.

Does the testcartridge start the menu? Maybe after hitting the space key?

You should be able to do all other tests.

However, I cannot tell you what revision to use. (I'm not home at the moment)

 

BR/

Guus

[+tf_hh]

18 hours ago, Bikerbob said:

BUT - after the suggestion of corrosion or bad solder joints. I removed the backplane and the lower plate and looked at the bottom. Although very clean.. well maintained.. I can see joints that have been reflowed? maybe a part or two replaced? also arrow marks on solder points that I guess were suspected to poor.  My point.. this board has been through this before.

 

On examination I do not see any BAD joints.. but I guess I could reflow the joints of the MMU and the SHifter??  would that be a suggestion before going at the 74LS?    The 373s are F types.. which I hear are better. Can they be tested? or do you just RE-RE and thats the best way?

 

Hard to say. The arrow marks and visible traces of a try to repair this machine showed me, that this is a hard-to-master patient. The already done change to F types is typically for repair guides in the 90s (where HCT were not so common, most guys use the F types then).

 

Truely this is one of the defects you can only check every connection with a scope or use brute force: desolder every part and change it for a test.

 

This is not a valid probe for all defects like this, but I have to exchange the MFP68901 two times in the last 6 months when doing repairs on STs. In both cases the diagnostic cards starts with Txxx errors. The MFP is used for the timers, too. So maybe it´s worth a try to replace the MFP.

 

[Bikerbob]

23 hours ago, guus.assmann said:

Hello Bikerbob,

 

It is very possible that the revision of the test cartridge is the problem and not the hardware itself.

I've had this issue, until I used another (multi) testcartridge. It might need a test cable or another fixture to pass this timing test.

Does the testcartridge start the menu? Maybe after hitting the space key?

You should be able to do all other tests.

However, I cannot tell you what revision to use. (I'm not home at the moment)

 

BR/

Guus

I am using the exxos multi test cart. It is on his store webpage. There are jumpers for the various types of ST.. but otherwise they are using the Atari corp ROMS.. as far as I understand it.

 

James

[+tf_hh]

On 11/18/2019 at 3:50 PM, guus.assmann said:

It is very possible that the revision of the test cartridge is the problem and not the hardware itself.

I've had this issue, until I used another (multi) testcartridge. It might need a test cable or another fixture to pass this timing test.

 

The Timing tests are independent from the external test fixture, so this can´t be the point. When a wrong diagnostic cart is selected, the system will come up with a red screen and "SYSTEM ERROR" or some Ixx codes (equivalent to the bombs).

 

[Bikerbob]

problem solved. I was able to contact wizztonics and they did have some info and DOCs..  I was missing a jumper under from the cpu to the megabus. This helped with blitter issue.. all better now.

 

Thanks for the suggestions all.

 

James