Intellivision graphics corruption

[kevincondrick@gmail.com]

Hay guys how are you all doing? Ok so I just got an intellivision and it will turn on and work fine for a few minutes then the graphics will get corrupted and eventually will freeze. If I try shutting it off and turn back on it just shows a black screen. Then if I wait awhile and try it again it will do the same thing. So I think it might be one of the ROM chips or something overheating. Although I tryed to feel around but I couldn't find anything hot. I also know it's not the cpu,ram and stic chips cuz I also have an INTV II swaped them out and they all work. So if anyone can help that would be great thanks ?

[+KylJoy]

I'd look for swollen/leaking capacitors.  I might just replace all the electrolytics.  I'd also check the power regulator or just replace it.  If it is dirty or if something was spilled in the system it may be overheating.  Is the power supply hot when this happens?  Could be a loose solder joint somewhere that is opening after system warms up.

[Relicgamer]

I 2nd that. Most of my old consoles fail from capacitors.  Sega master system gamegear 360s and fat ps3s

[kevincondrick@gmail.com]

I believe that these 2 transistors may have something to do with it. I also checked the power supply and everything on it seems to be working. I already ordered some up and ate planning to replace them I believe that's what it is.

[+-^CrossBow^-]

Funny you post this. I just had an Intellivision model 1 sent to me for AV upgrade. It wouldn't do anything at first when I powered it one. I would get the black screen that would flash if I pressed the reset. I noticed the original 3906s looked pretty toasty like yours and replaced them with some 4403s I had on hand. It fired right up. I then let it run for about 2 hours with a demo running. But, those 4403s were pretty blazing hot to the touch. It is funny because I've only see a few intellivisions that have this discoloration on the mainboard from these two transistors running so hot. And I can see why, they have a +5, +20 and +12 running into them. Or at least that is what I was reading when taking voltage readings off them. 

[kevincondrick@gmail.com]

2 hours ago, -^CrossBow^- said:

Funny you post this. I just had an Intellivision model 1 sent to me for AV upgrade. It wouldn't do anything at first when I powered it one. I would get the black screen that would flash if I pressed the reset. I noticed the original 3906s looked pretty toasty like yours and replaced them with some 4403s I had on hand. It fired right up. I then let it run for about 2 hours with a demo running. But, those 4403s were pretty blazing hot to the touch. It is funny because I've only see a few intellivisions that have this discoloration on the mainboard from these two transistors running so hot. And I can see why, they have a +5, +20 and +12 running into them. Or at least that is what I was reading when taking voltage readings off them. 

So is it normal for them to hot then?

[+-^CrossBow^-]

Apparently yes. Although after recapping this same system I'm now reading +5.5 and +17.2, +17.8 steady on Q1 and Q2. I've still got an issue somewhere else so the one I'm working on isn't 100% still.

 

[kevincondrick@gmail.com]

9 minutes ago, -^CrossBow^- said:

Apparently yes. Although after recapping this same system I'm now reading +5.5 and +17.2, +17.8 steady on Q1 and Q2. I've still got an issue somewhere else so the one I'm working on isn't 100% still.

 

Also your problem could be a bad chip.

I've noticed the transistors on then Atari 2600 wold also get hot like that. 

[+-^CrossBow^-]

9 minutes ago, kevincondrick@gmail.com said:

Also your problem could be a bad chip.

I've noticed the transistors on then Atari 2600 wold also get hot like that. 

Yes I do suspect this but I haven't any spare systems unless I want to start ripping chips out of my Sylvania model and I'm not really keen on doing that in case something else on the main board is causing the issue and could potentially cause an issue with the IC chips. I've started a new top to discuss it more and what I've checked so far.

 

[+-^CrossBow^-]

Looks like I got it figured out. Not sure why but it looks like the 4403s I used were the issue here. The original 3906s were definitely not working as I was back to the black screen issue when I put them back in. I'm guessing something in the different specs of the 4403s just didn't work very well. I installed 2 2709s instead which are a closer match to the original 3906s and had it running on a looping demo for over 2 hours without any issues. So I think I got the one I'm working on sorted for now.

 

[kevincondrick@gmail.com]

That's good man I'm still waiting on switching mine out as they are still coming in the mail 

[kevincondrick@gmail.com]

Well I changed both of the transistors but it still dose that same thing I also changed the one rom chip that dose the graphics ... So I really don't know ? idk if it's cuz it's still hot but I'll try it in a bit and see

[kevincondrick@gmail.com]

I suspect one of these chips mainly the on chip closest to the sound chip cuz when I tap on it the screen will fliker for a second then stop. Also I taped on the bord there on accidentally aright around there then it stopped working like instently. And now it wolnt come on again like it was.

[+KylJoy]

Resolder the legs and see if that fixes it.  Sometimes when things get warm a solder joint will open up due to expansion, etc.  Or, if you have access to desoldering equipment socket the ICs and test them in another unit.  And/or, Inspect the board closely for a broken trace.

[kevincondrick@gmail.com]

Well I replaced both of the transistors and all the main capacitors on this bord but the problem still present here a video of what it dose

[ChildOfCv]

On 4/22/2020 at 7:12 PM, -^CrossBow^- said:

Looks like I got it figured out. Not sure why but it looks like the 4403s I used were the issue here. The original 3906s were definitely not working as I was back to the black screen issue when I put them back in. I'm guessing something in the different specs of the 4403s just didn't work very well. I installed 2 2709s instead which are a closer match to the original 3906s and had it running on a looping demo for over 2 hours without any issues. So I think I got the one I'm working on sorted for now.

That part is such a weird circuit.

The transistors are two simple voltage regulators that should subtract about 3V from the 16V input.  The 7407 there has inputs from the STIC whose outputs pull each voltage low, apparently for a phased clock.  Finally, there's a diode to leech voltage from each output, regulated by another zener of unspecified voltage.

 

If the transistors are getting hot, that does lead me to wonder if any of those diodes, or the 7407, went bad.  Because those are the two drain sources for the transistors' output.

[kevincondrick@gmail.com]

What's the 7407? Cuz there's a big white thing that looked like it had boils on it. Also I checked the diodes with my mm and I think they looked ok 

45 minutes ago, ChildOfCv said:

That part is such a weird circuit.

The transistors are two simple voltage regulators that should subtract about 3V from the 16V input.  The 7407 there has inputs from the STIC whose outputs pull each voltage low, apparently for a phased clock.  Finally, there's a diode to leech voltage from each output, regulated by another zener of unspecified voltage.

 

If the transistors are getting hot, that does lead me to wonder if any of those diodes, or the 7407, went bad.  Because those are the two drain sources for the transistors' output.

 

[ChildOfCv]

The IC near the transistors is labeled HD7407P.  That's the 7407.

 

I don't know what that encased part is.  They don't seem to label anything on the board and I can't see the bottom traces to match it up on the schematic.  Well they show a jumper wire somewhere.  Maybe that's it?  See if its resistance is near 0.

[kevincondrick@gmail.com]

Yah I did check that it was like .01 ohm. The only thing other then the chip. Or maybe the residters. 

 

You know we're I can get one of those chips?  

 

Also is there any way to check it in circuit?

Edited April 26, 2020 by kevincondrick@gmail.com

[kevincondrick@gmail.com]

3 hours ago, kevincondrick@gmail.com said:

Well I replaced both of the transistors and all the main capacitors on this bord but the problem still present here a video of what it dose

 

[ChildOfCv]

For whatever reason, the videos don't show up for me.  Anyway, the resistance seems to mean that that's a jumper wire.

 

I'll note a couple of things:

 

1) 7407 is a standard chip.  https://www.mouser.com/Semiconductors/Logic-ICs/Buffers-Line-Drivers/_/N-50nah?P=1z0y3wa&Keyword=7407&FS=True

    Either of those would work.  Not even the datasheet shows a difference between the two.

 

2) It may help to calm down the transistors, but probably won't help with the graphics corruption.  If this circuit quits working, I'd expect the whole thing to freeze up.  However, make sure, as Crossbow mentioned, that the transistors are 2N3906.

 

3) Reflow the solder around the ICs that tapping affected.

[ChildOfCv]

14 hours ago, ChildOfCv said:

That part is such a weird circuit.

The transistors are two simple voltage regulators that should subtract about 3V from the 16V input.  The 7407 there has inputs from the STIC whose outputs pull each voltage low, apparently for a phased clock.  Finally, there's a diode to leech voltage from each output, regulated by another zener of unspecified voltage.

Looking closer and doing some research, those are current limiters.  Q2 is limited to around 40mA while Q1 is limited to about 27mA.  Then on each current limiter there's a voltage regulator (CR3/CR5 or CR4/CR5).  That limits the clock pulses to 12V.

Why the difference in amps?  <shrug>.  Maybe the CPU has different requirements.  At least it better have some.  A 7407 rates an output's absolute maximum current as 40mA, which is what Q2 outputs.
 

The loading effects are not so easy to surmise though.  With no load, the transistor has to dump all of its current through the base.  While base current will be limited by the 1K resistor, that's still a lot.  With full load, the transistor is passing all of its current, but absorbing very little power.  So at which condition does it heat up more?  Might just have to breadboard an equivalent circuit and try it out.

[kevincondrick@gmail.com]

I know the transistors are both 2n3906 and there both brand new so thos arnt the problem. Also it seams like the top one closer to the CP chip is the one get hot. The bottom one get warm but not as hot as the other. Also I may just get one of those 7407 chips and try replacing it.

[ChildOfCv]

23 hours ago, kevincondrick@gmail.com said:

I know the transistors are both 2n3906 and there both brand new so thos arnt the problem. Also it seams like the top one closer to the CP chip is the one get hot. The bottom one get warm but not as hot as the other. Also I may just get one of those 7407 chips and try replacing it.

The hotter one may be Q2.  Looks like it has to provide clock power for both the CPU and a RAM chip.  That probably explains why it outputs more current.  BTW, according to the data sheet for the CPU, it needs 15mA for the clock inputs.  The RAM needs 10uA, which is chump change.  Seems like it would be safe to replace that 68-ohm resistor with another 100-ohm.  Maybe that would also cool down the transistor.  The RAM is U2 and may be marked as "RA-3-9600"  Is that the one you were tapping on?

[kevincondrick@gmail.com]

Well no it wasn't the ram chip I know it's not that. I swapped that out with my intv ii and it turns out thats what I need for it cuz it works on it. Those two smallr chips down by the sound chip was what I tapped on but I tryed it again and nothing happened. So I really don't know any I'm trying to replace the HD7407P chip hex buffer. 

21 minutes ago, ChildOfCv said:

The hotter one may be Q2.  Looks like it has to provide clock power for both the CPU and a RAM chip.  That probably explains why it outputs more current.  BTW, according to the data sheet for the CPU, it needs 15mA for the clock inputs.  The RAM needs 10uA, which is chump change.  Seems like it would be safe to replace that 68-ohm resistor with another 100-ohm.  Maybe that would also cool down the transistor.  The RAM is U2 and may be marked as "RA-3-9600"  Is that the one you were tapping on?