• Please review our updated Terms and Rules here

I have a iic that won’t boot. Need help

M

Mochatea396

Experienced Member
Joined
Feb 9, 2021
Messages
148
Location
Hudson Valley, New York, USA
Hello members,
I have an A2S4000 iic that I only get the green led on the keyboard. Here’s the history up to now. When I got it, it booted. The board had already been modified for ROM0 so I burned a 4X ROM. I was using it running diagnostics and playing all kinds of games. Passed the onboard “system ok” many times. Then it started locking up, messed up video. The onboard now was giving me ram errors. Different errors every time I’d run it. Then I’d get a system ok and the machine would boot again. Would be fine for a little while and then do it again. Then one day it just stopped booting altogether and just the green led.
What’s been done:
The PSU (internal and external) are good.
I replaced the CPU, MMU, IOU with known working chips, as well as tried the originals in a working machine. All good.
I replaced the sockets on the CPU, MMU and the monitor. (Nice removals with a gun no damage)
I socketed the 311 and the 555 and those tested good as well.
I socketed all 16 ram and replaced them with known working chips. The others all tested fine in another machine.
I’m getting no logic reading at all from pin 40 on the CPU on boot. I am however getting 5v to the correct pins on the CPU, MMU, IOU and the monitor. 5v to the 555, IWM, TMG, GLU and the AY3600 as well. I was getting 15.3v to the LM311 which I think may be normal. CR2,CR3 tested ok in circuit, CR1 has to be removed to test and I didn’t. I’m getting the correct ohms on the resistor networks that SAMs says I should. And all 16 RAM show 5v. Strangely any pinout for 4164 ram I could find online the GND and the VCC are on different pins than I’m getting. Even on a working machine they don’t match what I found online. They match the board I’m working on. This board also has the factory mod replacing the LS161 with the oscillating crystal.
Any direction would be appreciated
🙏
 
Well, the failure mode of slowly decreasing in functionality suggests it is probably not a digital logic chip - it's possible, but much less likely.

I'd suspect either a power issue, failing solder/PCB connection, or a bad socket. It sounds like you already took care of the sockets. Did you check/touch up solder on other chips?

What I find odd is that you got RAM errors, and yet the RAM tests good elsewhere.

You say it has +5v - is there perhaps "noise" on the power?

But what exactly do you mean the GND and VCC are on different pins? That is very standard. There are truckloads of spec sheets for 4164 chips here: https://minuszerodegrees.net/memory/4164.htm

It should be Pin 8 VCC, pin 16 ground. If that is not what you are getting, then there is your problem.
 
SomeGuy said:
Well, the failure mode of slowly decreasing in functionality suggests it is probably not a digital logic chip - it's possible, but much less likely.

I'd suspect either a power issue, failing solder/PCB connection, or a bad socket. It sounds like you already took care of the sockets. Did you check/touch up solder on other chips?

What I find odd is that you got RAM errors, and yet the RAM tests good elsewhere.

You say it has +5v - is there perhaps "noise" on the power?

But what exactly do you mean the GND and VCC are on different pins? That is very standard. There are truckloads of spec sheets for 4164 chips here: https://minuszerodegrees.net/memory/4164.htm

It should be Pin 8 VCC, pin 16 ground. If that is not what you are getting, then there is your problem.
Click to expand...
Thanks for a quick reply,
And thank you for properly identifying those 2 pins. I don’t know what they had posted that I was looking at but what I was looking at had pin 8 as the GND and pin 9 as the VCC. Pin 16 showed it being the VSS. That’s where I was getting screwed up. Your posted link concurs with your info. Let me run through all 16 tomorrow again using pin 8 and 16. With that bad info who knows if I was even on the VCC. Someone on another forum also said inconsistent ram errors sound like a power issue. I wouldn’t think there’s any noise in the power, I threw another board on with the same internal and external PSUs and no issues with another board working fine unless something on this particular board could be effecting the power. Let me check the ram voltage again and I’ll reply back.
 
That would be the pinout for a 4116 (16k bit x 1) dynamic ram chip. Those were totally different beasts.
 
That pinout picture I was looking at was clearly marked 4164, figures I use the one that’s marked wrong. 🤨 But using the correct pins 8 and 16 I’m getting 4.99v on the money on all 16 ram. Those must have been the pins I was using before and not marching that incorrect info was boggling me. Ok, so all ram as well as everything I’ve mentioned previously seem to be getting 5v where it should. I was kind of hoping there was going to be a problem there but the chase continues. Lead me to where I should be looking next.
 
Today I rechecked for logic at pin 40 of the processor. Now I have low logic on power up but doesn’t go to high. Stays low. Ctrl reset while the probe is connected doesn’t change the logic, stays low. Before I had no logic, now at least I have logic. But why is it not going to high? I also am now getting high logic readings on pins 2,4, and 6 of the CPU which according to SAMs, I should.
 
Last edited:
Have you tried a different power brick with the computer, or tested the power brick on another machine other than a volt meter?
 
Have you checked the keyboard? I was recently working on a IIC+ and it would do all sorts of crazy things if the keyboard wasn't connected, or had a bad connection.
 
Yep, I’ve tried another keyboard also. And these keyboards have no problems in other machines. I was also told that with no keyboard it will still boot it will just run onboard diagnostics. But I did have a working keyboard connected when I was performing these tests.
 
pin 40 should be 5V high after power on via a 1K pull up resistor in resistor pack RP1. Measure pin 40 to ground with an ohmmeter while powered off. Should read at least 1K.
Either you have a short to ground or an open to the 5V line. It takes both ctrl and reset keys down to ground the pin so not likely a bad key.
 
I tested pin 40 to ground and I initially was getting around 570ohms. Just poking with the probes. Then I hooked on with a mini grabber and hooked on to several ground points and it was holding at 437. I picked the board up while the leads were connected and moved it around to see if the reading would change with a slight flexing and it stayed at 437. No where’s near 1k.
 
As a reminder, make sure your DMM is not in diode mode for resistance checks since that lowers the reading. I own a 2c but it's all together so not opening it up. I do have a 2e motherboard so
measured pin 40 to ground and get reading in the megaohms. It has the resistor pack RP1 next to the 65c02 but of course no power supply is plugged into the 5V line on the board so the 1K pullup
goes to nothing other than the other chips. There should not be 437 ohms on that pin. You should probably pull out the 65c02 and measure pin 40 again on the board socket.
 
The meter wasn’t in diode mode and I checked it with 2 meters. I did what you did also, grabbed a iie board (that was easiest to get to) and the same test showed 1.17k on the iie board. I removed the 65C02 from the iic as you suggested and got the same 437 readings coming directly off the socket. Think it’s worth swapping out the RP1? RP1 is a 1K network correct?
 
Last edited:
Mochatea396 said:
The meter wasn’t in diode mode and I checked it with 2 meters. I did what you did also, grabbed a iie board (that was easiest to get to) and the same test showed 1.17k on the iie board. I removed the 65C02 from the iic as you suggested and got the same 437 readings coming directly off the socket. Think it’s worth swapping out the RP1? RP1 is a 1K network correct?
Click to expand...
Yes it's a 1K network. Just de-solder it to check resistance at pin 40 without it. If it was bad it would be an open resistor preventing a pullup to 5V at 40. Looks like it's pins 1 and 4 of the network if it's numbered on the board. Or just add your own 1K resistor from pin 40 to 5V supply to see if that brings the voltage up.
 

Attachments

  • pin 40.JPG
    pin 40.JPG
    11.8 KB · Views: 2
Yes I checked for 5v on all the ICs wherever the VCC was. I removed the RP1 and checked pin 40 to ground with the network out and I’m getting 900ohm. I checked all the pins on the network and they showed about 1.1k. When the resistor network is in is when pin 40 shows 500ish. When the network is in circuit 8 of the 10 pins show 1.1k. Pin 1 shows 310ohm and pin 4 shows that ohm reading I’m getting on pin 40.
 
Sure can't imagine what else is keeping that pin 40 low. The only other sleeper I know about is the fiber metal washer(s) that go under the floppy drive can dislodge when apart and end up causing a random short.
 
You must log in or register to reply here.
Back
Top