What's wrong with my Intellivision?

[spacecadet]

I know there have been a couple of similar threads recently but I don't want to hijack them and I haven't seen an actual example of the same symptoms I have.

 

I finally got around to testing the Bandai Intellivision I bought last month, and this video shows what I got:

 

 

First game is Burgertime, second is Sea Battle. But this is consistent from game to game, and it seems very specific. Like it's character-based stuff that's corrupted. It seems like it's probably a specific chip that's bad, but I don't know enough to know which one.

 

I have no problem buying a donor system for another chip, I just don't know what to replace.

 

Thanks for any help!

 

[+Lathe26]

In a nutshell, it is having a problem with the GROM chip, the System RAM chip, the STIC chip, or the connections between them.

 

Namely, what is happening is that GROM characters are offset by having the their high bit set (i.e. like adding 0x80 or 128 to the GROM character's index). For example, at time marker 0:10 in your video, there is a narrow triangle that appears at the bottom of the screen that is from you selecting "1" as the number of players. From the chart below, a "1" is at index 0x11 while the narrow triangle is at index 0x91. Similarly, at the end of the top line is supposed to be a "2" (index 0x12) but is displayed as an upside-down narrow triangle (index 0x92).

 

On the upside, the GRAM chips seem ok, as does the CPU, main ROM, PSG, and other chips.

 

I'm not an expert at fixing Intellivisions, but I would start with the 4 chips mentioned at the top of this post first. Perhaps just check for any shorts first caused by dust.

 

Other folks will likely have more definitive suggestions.

 

 

[mr_me]

Its like bit 10 of system RAM data (8th bit of card number) is stuck on . If SNAFU works correctly or a game that uses foreground/background mode works correctly than the system RAM is fine. Like lathe26 said, open it up and give it a good visual inspection.

[spacecadet]

I will open it up and take a look tonight. Man, I hope it's just some dust shorting something out. It's in such great shape - the best looking Intellivision I've owned since my original - that I was honestly a little shocked when I fired it up and saw this. I've never had a problem with any of my Intellivisions, some of which were in much worse physical shape.

[HunterZero]

Indeed it looks like an address line between the STIC and GROM is stuck high (0x00 background card is displaying as 0x80). All the sprites and custom background cards look to be displaying fine.

 

Clean the cartridge slot and look for bent pins. On the motherboard, look for obvious shorts or visually defective components. Check the continuity between the STIC/GRAM chips and the RA-3-9503 GROM chip, reseat and clean the GROM, and if that fails the next step is to use a logic probe to check which data pin is stuck on U7/U8/STIC/GROM (not oscillating). Check that the 74LS86N logic gate at U17 is working correctly too.

 

- J

[spacecadet]

and if that fails the next step is to use a logic probe to check which data pin is stuck on U7/U8/STIC/GROM (not oscillating). Check that the 74LS86N logic gate at U17 is working correctly too.

 

Ok *this* I confess I don't really know how to do. I can check continuity, check for bent pins, etc. but I don't currently have a logic probe or know how to use one. I guess I'll have to decide if it's worth buying one just for this vs. buying another Intellivision and just trying to replace some chips.

 

How would I check if that logic gate is working, assuming I buy a logic probe?

 

I did open up the Intellivision but I didn't get to the point of removing the shields around the motherboard yet. Hoping to do that Tuesday or Wednesday. On initial inspection, everything is like freakin' new, though. I mean it looks like this thing has never even been taken out of the box. I can't even believe something's broken on it, but obviously something is. I really doubt it's dust causing a short, though. There's no dust whatsoever inside, and everything looks perfect so far. It's like it just came off the assembly line.

[HunterZero]

A logic probe connects to 5V and ground reference points, and you can then probe chip data legs/lines in 5V logic circuits for activity. The pen will then tell you if that line is held high, low or pulsing.

 

An oscilloscope can do the same thing, but can display the pulses on the data line.

[spacecadet]

An oscilloscope can do the same thing, but can display the pulses on the data line.

 

I was wondering about that - I have a scope, so I guess I can save a little money on buying a logic probe. I'm not that experienced with using the scope either, though - what am I looking for, something being held at 5V that should be pulsing?

 

I'm planning to try to get the shields off tomorrow. Hopefully they don't put up a fight.

[HunterZero]

 

I was wondering about that - I have a scope, so I guess I can save a little money on buying a logic probe. I'm not that experienced with using the scope either, though - what am I looking for, something being held at 5V that should be pulsing?

 

I'm planning to try to get the shields off tomorrow. Hopefully they don't put up a fight.

 

Yes, the oscilloscope will show you if the volts are held at 5V or 0V, or pulsing between. A digital 'scope with a memory can 'pause' the waveforms so you can read the data on the lines.

 

- J

[spacecadet]

Ok, this is an old topic now but I only just got to really trying to get it fixed, and so far I've been met with frustration. I need more help/advice.

 

First, it's not the GROM or STIC chip. I replaced both with known good chips and had the same problem.

 

So I decided I'd just replace *all* the socketed chips just to see if it's any of them, which seemed simpler than doing it one by one and then potentially finding out the chips aren't the problem only after a long process of testing and re-testing. But, disaster! The first chip I took out on the Japanese board (the sound chip) I had to really yank out, and now nothing's going back in that socket. Looking in the slots, it looks like there's some rust or dirt or something in a few of them. (See the photo of the socket; look at the 3 or 4 most top-right slots. Ignore that the socket looks warped; that's just my phone camera.)

 

Is there any way to dissolve this crud or dig it out without damaging the board? I tried Deoxit and that did nothing. Or is my only recourse to replace the socket?

 

I would really like to keep this board rather than replace the whole thing with an American board. (That's what I've done temporarily but I would like to fix the JP one and put it back.)

 

I'm attaching some pics of the board in case anyone can see anything obviously wrong with it - I don't see anything myself except for that one socket that's not flush against the board, but on the underside it looks like all the pins made it through and are soldered properly.

 

Thanks for any more help.

 

[Guest LiqMat]

I can't tell if its the lighting, but that area in the red circle almost looks like after factory soldering work was done on this board. May just be light reflecting, but it looks different in the photo. Do you know if this board was serviced at some point?

 

 

Edited December 29, 2017 by LiqMat

[spacecadet]

I can't tell if its the lighting, but that area in the red circle almost looks like after factory soldering work was done on this board. May just be light reflecting, but it looks different in the photo. Do you know if this board was serviced at some point?

 

That's actually where the shield was soldered to the board in that spot. The shield soldering on this board was really messy and *really* overdone, so I removed as much of it as I could from this area. But my American board had a big blob of solder in this same spot; it just wasn't quite as messy. I think this is probably the way it was originally; I don't think anything has been done to this system. It looks like it's barely even been used.

 

These systems were made in Hong Kong, whereas the Mattel systems I have were made in the USA. So it is a different factory and back then I don't think Asian factories were as highly regarded as they are now.

 

I've been trying to shoot a video of this repair for my YouTube channel so I happen to have a video clip specifically about this solder joint:

 

 

You can see the Bandai vs. Mattel solder joints there. But it's doing the same thing before and after taking the shield and that solder off, so I doubt it's related to the character graphics issue.

Edited December 29, 2017 by spacecadet

[HunterZero]

I am assuming this is your repair vid? There aren't many of these Bandai Intellivisions around after all! And you said you were working on a repair vid. Nice work!

 

Good to see it's working again! Looks like all the ICs were fine in the end, which is good as IC spares can be hard to come by. A couple the DIP40 IC sockets were toast. In particular, the RA-3-9600 RAM socket had a bad connection.

 

Looking at the soldering job on the shielding of the Bandai unit, the sodering does look like rework. I wonder if the unit used a motherboard that was taken from the normal 2609 USA production line, and swapped the RF modulator to turn it into an NTSC-J unit?

 

That huge chunk of solder on the cartridge slot is meant to be there, several of the central pins there tie to ground. The RF shielding is grounded to the ground plane that runs around the edge of the boards.

 

Edited May 1, 2018 by HunterZero

[HunterZero]

Or was the original RA-3-9600 chip itself toast? If so, this is one of the most common (if not the most common) point of failure that I've come across for Intellivisions.

 

- J

[CharonPDX]

Okay, that was odd. I saw that this thread was updated, but didn't go in and read it.

Then I caught up with my YouTube subscriptions and saw one of the channels I watch had a video about repairing a Bandai Inty. Thought "huh, wonder if this video would be useful to the guy on AtariAge?" Turns out it *IS* that guy on AtariAge! Great vid, Modern Classic/ spacecadet!

[spacecadet]

Or was the original RA-3-9600 chip itself toast? If so, this is one of the most common (if not the most common) point of failure that I've come across for Intellivisions.

 

- J

 

Yeah that's my video and yeah, it was the chip itself. Though it did take a lot of careful finagling to get the replacement chip into that socket. But it's in now and it works. I didn't end up having to replace that socket, just the chip.

 

I think you're the guy who made the thread where I got the instructions on making the new power cable - I linked to that in the video description in case others need it, it was very helpful. I really didn't know what to do at that point otherwise, and was about to give up before finding that thread.

 

Oh, and just to clarify about the blob on the cartridge port - the video was unscripted and I don't always say everything I mean to say, but the reason I was concerned about that on the Japanese port was that it looked like there were more than 4 pins that solder had dripped on, and not having a pin mapping I was worried that there were one or more pins that had been inadvertently soldered to other pins that they shouldn't have been. Obviously it ended up not mattering and I almost took that section out, but I left it in the video because it led directly into the discussion about the quality and quantity of soldering throughout the entire assembly.

 

I also cleaned up that area when I put the assembly back together, so it looks more like the American one. (Not quite as clean, but better than it was before.)

Edited May 1, 2018 by spacecadet

[+cmart604]

Not that I'm ever going to do what you did but I thoroughly enjoyed your video. Hmmm, maybe that's why I enjoyed it. ?? The Bandai was probably my first Holy Grail which I genuinely never thought I would get my hands on. Thankfully I was able to check that box. ✅✅✅

[spacecadet]

Thanks! Yeah I had always wanted one too, and just about crapped myself when I saw this one sitting boxed in a store (I got it at the Super Potato in Ikebukuro, Tokyo... *not* the one in Akihabara. Just in case anyone's wondering.) It was labeled "untested" but I couldn't not buy it anyway. It looked great so I figured what could be wrong with it?

 

I'm really glad I got it working. Now that it is, the price I paid for it was a steal. (About $90.)

[+cmart604]

Thanks! Yeah I had always wanted one too, and just about crapped myself when I saw this one sitting boxed in a store (I got it at the Super Potato in Ikebukuro, Tokyo... *not* the one in Akihabara. Just in case anyone's wondering.) It was labeled "untested" but I couldn't not buy it anyway. It looked great so I figured what could be wrong with it?

 

I'm really glad I got it working. Now that it is, the price I paid for it was a steal. (About $90.)

Awesome price! I have purchased a couple at what I think are pretty good prices but most of the rest I paid quite handsomely for. Definitely a must have if one is going to start collecting hardware variations.

[HunterZero]

I think you're the guy who made the thread where I got the instructions on making the new power cable - I linked to that in the video description in case others need it, it was very helpful. I really didn't know what to do at that point otherwise, and was about to give up before finding that thread.

 

Yes, that was me - glad you found the info useful!

 

And I created the Intellivision PCB IC location graphic that you used too. Here's the latest copy...

 

 

- J

Edited May 2, 2018 by HunterZero

[HunterZero]

Yeah that's my video and yeah, it was the chip itself. Though it did take a lot of careful finagling to get the replacement chip into that socket. But it's in now and it works. I didn't end up having to replace that socket, just the chip.

 

Out of interest, did you try the RAM chip out of the Bandai unit in the USA motherboard?