• Please review our updated Terms and Rules here

Model 4 (serial 000018) scrambled screen, capacitor fail or something different?

whizzi

Experienced Member
Joined
Feb 13, 2016
Messages
113
Location
The Netherlands
A while ago I found in somebody's garagebox a Tandy TRS-80 Model 4. It had mold on the screen and was there for probably 10 years, buried under several boxes. I took it home and let it sit for a month in my room.

Out of curiosity and knowing the filter would break in worst case scenario, I decided to turn the machine on. It didn't start to smoke immediately and actually gave something on the screen. When I decided to get the camera closer to the computer, the capacitor went pffffft and the room started to smell ugly.

The video:

When I opened the machine, I noticed a lot of changes on the mainboard. Not sure who did it, it's been repaired by a Tandy repair shop and this is serial nr 000018 (the newest chip inside has datecode early 1983).

But I'm kinda thinking, is the weird screen a result of the capacitor being end of life or do I have another problem besides the now broken capacitor?

A picture of the mainboard:


Power supply with the main filter broken:


The bottom:


And the interesting sticker stating Tandy repaired it:


Does anyone ever see these modifications or are these modifications made by Tandy because it's probably a very early Model 4 board? The new capacitors are on it's way BTW :)
 
That's a Matra France production, which have their own serial numbers and catalog number (260-1069). I have had one here, S/N 000059. I don't know how many they made but probably only a couple of thousand.

They appear to be CKD version sent to France for assembly. What that achieved is anyone's guess but the European (and Australian) M4s this early on got the green CRT whereas the US and Canadian ones still got the B&W CRT.

That motherboard appears to have all the standard re-wiring all early M4 boards did.

since it has 128K memory, what it needs to max it out is a HRG board... :)

Cheers,

Ian.
 
Thanks Ian, that's actually a relief. My previous 'rescued' TRS-80 (The Model 1 with garbage on the screen) had so many mods done on the mainboard, that it was an impossible task to diagnose without in-depth knowledge of the mods. Lucky I found the TRS80 club in the Netherlands with 2 people who know everything about these machines and they got it working again.

When I saw this M4's mainboard, I thought oh no, more unknown mods, but it's a normal 'mod' as you say so that helps in diagnose.

First thing is to get the power supply back in action, I can only hope this also fixes the screen that is displaying weird characters on startup. The FDD is being read, so as far as I know, the basics are working (at least up until it trying to boot). I already found that a re-seat of all chips could fix the issue too.

I'll keep the progress up2date here. Still a nice item that I found in someone's garage right?
 
You want to replace (at least) the caps in the red boxes indicated in the picture here:
TRS80PowerSupplyCaps.jpg


For reference, here's a thread (and my post) with regard to some reasonable replacements that might work...
http://www.vcfed.org/forum/showthre...for-RIFA-ewplacement-caps&p=418649#post418649

Edit: It's quite likely the smoke came from the 0.1uF at the bottom of the picture. (I just watched your video and realized you already knew this.)
 
Last edited:
Yeah Whizzi those gold paper Rifa caps do like o go up in smoke and they smell really bad. So good idea to replace them all. And FYI they are also used in Apple II power supplies as well.
 
One is definately broken now, the other ones seems fine but I'm replacing them anyway.

Still no idea if the weird screen shown in the video is due the bad RIFA's or some other problem I can't tell anymore :/
 
I'd start with working on the power supply. These RIFA caps won't prevent it from booting, nor will they cause the garbage screen issue but it's still important to swap them out (or if nothing else, temporarily remove them from the board). That'll prevent further smoke at least.

Then start with the recommendations from TheCorfiot above. I'd additionally recommend reseating the ROM chips, too.

Since you mentioned in the video that the screen had some mold on it, it's possible that the unit wasn't stored in the dryest of conditions and you might be facing corroded sockets/chip legs. Reseating things will (temporarily) clean the contacts and sometimes give you just enough to tell you that you are on the right track. If this is the issue, sometimes reseating can make the unit work for quite a while. Other times it proves more temporary but at least will give you more information than you had previously.

If you do wind up having or suspecting corrosion and contact issues, DeOxIt can help and will be a longer-term solution than just reseating.

If that doesn't help, you could even go one step further and reseat anything socketed on the board. Reseat one chip at a time and power up between each one to see if you've made any progress/changes. Note which chips changed/helped the issue.

Be careful, work slowly and cautiously. You don't want to break or bend any pins on the chips. Because of this, I'd stick with the first recommendation of reseating just the 6845, Video SRAM and ROMs first. Sometimes the less you mess with on these old systems, the better. (Especially with all of those PCB fly wires.)

You can also try the "burnt finger test". Power up the machine and CAREFULLY see if anything on the board (RAM chips specifically) get hot to the touch. This can help isolate issues sometimes. Additionally, RAM chips should run a bit warm (not hot). If one is particularly cold compared to the others, that might lead you in the right direction as well. If you have a non-contact thermometer (laser ones work well) or even better a Thermal Imaging/FLIR camera you can sometimes use that to find defective chips.
 
Very useful stuff here. Another reason why I love Tandy computers so much. So far, all my Tandy's gave me new friends and contacts and the new knowledge on so many levels. I love it, I absolutely love it.

Maybe a stupid question, but I can just desolder the RIFA's and it should work? My knowledge of electronics is rusty but is being refreshed since I started with collecting and maintaining these computers. I have to wait until the 20th of August before I get 3 new RIFA's, but if I can make it run without these RIFA's, that'll be very nice.
 
You can run the power supply without them temporarily I have done it before but replace as soon as you can.

As for the corrupt screen my 1st instinct is RAM as was mentioned before.

So maybe pull all the RAM from the right side and swap with the RAM on the left side see if that helps any. Basically swap banks right to left, left to right.

If not then also reseat the 68045 CRT Controller the large chip above the Z80 CPU and video RAM as thecorfiot mentioned at U21.

FYI all those little patch wires are unaffectionately known as the "Tandy Tangle" not a flattering nickname by any means.
 
Yep, the RIFA caps are completely optional on a temporary basis. They are there to prevent interference from exiting the computer and affecting neighboring electronics. Essentially filter caps, but to filter the electrical line, not the computer itself.

Also: LOL @ "Tandy Tangle".
 
And FYI they are also used in Apple II power supplies as well.

Thanks for the warning. I just opened up my Apple IIe power supply (lots of screws!), found one and replaced it. I think my IIe is a later production machine, and the Rifa cap I replaced wasn't as badly cracked as the ones in older TRS-80 gear, but it still exhibited some case cracking that I would expect to lead to eventual failure. Only one of the AC line EMI caps in my Astec-made Apple IIe supply was a Rifa paper dielectric; the other two are Y-rated disc ceramic caps.
 
I changed ram from left to right and reseated the chip above the z80. Now it doesn't read the fdd anymore and the screen shows lines.

Reseated the z80 to be sure, no change. Reseated videochip again, no change. I guess I have to swap back the memory?
 
Reseated all chips (with tests in between) and put the ram chips back in it's original position.. No luck.

I think something else is broken at some point, so I have to start with a logic analyser or something. None of the chips are getting warm nor does it respond to reset or something else. It powers on, drives are powered on but not being read, screen shows this (the black bar is refresh rate)



That's it, I have to say this is still with the broken RIFA in place. It doesn't smell or smoke at this point. Any ideas what to test more?
 
Ok I see 3 things that don't look right. 1: Most obvious is the EPROM or EEPROM in position U70. That just happens to be the startup ROM if I am not mistaken.

2nd & 3rd item. Above U71 there is a circuit trace that looks like maybe it is cut? Also same thing near U72 near C62. If you can check to see if those traces are connected
to either side of the trace it would eliminate those as a possible problem. If they are open than you may need to jumper them to reconnect the circuit.

Also that display could mean a clocking problem those 2 PAL chips in the upper left area of the board control the timing signals.

Just some ideas. But try what Audronic says 1st. Hopefully will get the "Cass?" display.

And if you don't have the Model 4 Technical manual then you need to find it someplace.
 
I've tried to do the break, no change. The reset doesn't seem to do anything, so I've also tried to power on and press break key, nothing.

I'll check the traces for connectivity, I don't think they are broken but that's easy to measure. The EEPROM I also found very weird and this could be as well the cause of everything.

What I've done so far:
1) Started up -> System did read drive but garbage on screen (as seen on video)
2) Put camera close and retried start up, screen didn't show anything (but did have power), fdd was being read and short after the RIFA died.
3) Opened up the case and swapped memory banks from left to right and reseated video chip (big chip above Z80)
4) Powered on (with broken RIFA in place), monitor shows output as shown in picture and FDD isn't being read anymore on startup.
5) Reseated all other chips, retried power on in between, no change.

The only regret I have is not turning the system on after the RIFA died and before the memory swap + video reseat. Now I can't tell for sure if the memory swap caused the non-booting system. By non-booting I'm talking about reading the FDD drive which it doesn't do at this point.

I basically can come with 2 conclusions:
The RIFA dying broke something (a power spike on the mainboard breaking the boot ROM?)
The memory swap / videochip reset broke something (which would be the first time I'm breaking a chip, I'm really careful getting these chips out of a sockets and back in).

Will the Tandy boot up with half the RAM inserted so I can test if one of the IC's died or can I just try to get it booted with only 1 RAM IC in place?

Can the broken RIFA cause the system to not power up at all? To me it looks like the mainboard doesn't get power at all, only the monitor. I guess I have to measure that somehow.
 
Ok

Please try this, Turn the 2 video controls to about the halfway position and try the reset / break. As the brightness may be masking the "Cass"
Do you have a logic probe ? and a Multimeter ??
Check Voltages on main board against the Circuit +5 / +12 etc , Check signals on z80, Clock, and all lines.
Good Luck Ray
 
Last edited:
Back
Top