Amiga 600 Rev 1.1 Op-Amp problem

[turrican9]

So I recently got hold of an Amiga 600 Rev 1.1 (The one that says A300) motherboard.

Motherboard is recapped and is working fine except for one thing. The sound is slightly distorted on both channels.

things lead me to the op-amp. On my working A600 Rev 2D board I have the +12v and -12v on the op-amp. On all other pins on the op-amp I measure around 2.3 Volts. On this board the sound is fine.

On the Rev 1.1 board I also measure the correct +12 and -12v on the op-amp. However, all other pins on this op-amp is in the 1.3v ish range.

Time to change the op-amp?

Edit: R301 must have been somewhat working at the time of this first post, as I had the +12v to the op amp.

[turrican9]

So I tranplanted the LF347 from the Rev 2D board. Rev 1.1 board also had the LF347 but it's a bit smaller.

Got even worse distortion this time. +12v and -12v still fine and feeding the chip. Other voltages not where they should be on LF347 legs. so I was sure I had broken this LF347 when desoldering it but I have transplanted it back to the 2D boad and it still works fine.

Very strange. I might try and put the op-amp from the 1.1 board in the 2D board to see if it is working.

[turrican9]

As soon as I put the good LF347 back in the Rev 1.1 board I'm missing the positive voltages on all pins. All shows as negative. And the +12v pin shows just a few negative volts. If I remove the OP amp then I have +12v on the pad.

I also put the op amp that came from the rev 1.1 board into the Rev 2D board. It seems it's not working correctly in this board either.

The op amp in the rev 1.1 board just says '347' and is a little bit smaller package VS the LF347 that came from the 2D board. Maybe they are not compatible with each other?

I should try to put the original one back in the 1.1 board to see if I get the positive voltages back and the +12V

[Mick]

Is this any help? it sounds possibly related.

[ Show youtube player ]

[turrican9]

Originally Posted by Mick

Is this any help? it sounds possibly related.

[ Show youtube player ]

I was actually just looking at that video. And I noticed these resistors looks to be the same between the Rev 1.1 and 2D boards. 1.5K and the same he has on his 1A board in the video. which he changes to bigger one. But on the 2D board I have no distortion.

What I could try is to put the original op amp back into the 1.1 board then inrease these two resistors. But very strange both rev 1.5 and 2b/d boards have these exact same value resistors.

Also, very strange how the op amps between Rev 1.1 and 2D doesn't seem to be interchangeable.

Edit: So put back the original op amp in the Rev 1.1 board and now I measure the same symptoms I did with the op amp from rev 2D board. Missing the +12v and negative voltages on all legs. Sound is low and distorted.

[turrican9]

So got it working now. Don't know where to begin....

Been working all night long on this issue. Trial and error. Think I must have been on this for 2 days non stop.

First of all I found out that resistor R301 that was supposed to be 10 Ohm was infact showing thousands of Kilo Ohms. And that is why I had no +12V on the op amp. Resistor R302 which is set on the side that feeds the op amp with -12v was also supposed to be 10 omh. It showed like 50 ohm.

I changed both. I currently don't have smd resistors and had to just use some regular resistors (ghetto style).

So when this was done I finally had the +12v to the op amp. However, I still had the problem where I was getting too low voltages on the input and output pins. showed around 1.6v. These are supposed to be in the 2.5v range for the op amp to work properly. So I still had distortion in the sound.

So I started researching... Some investigations lead me to R304 and E301. I probed between pin 10 on the Op amp to the closer side of R304. Then from pin 10 to E301. Then I measured the resistance from pin 10 to ground. It was like in the 0.4K Omh range. Had a similar result on my 2D board.

R304 had an R303 besides it. They are 1K ohm resistors but they seem to be paralled and the result is about 0.4K ohm from pin 10 and 12 on op amp to ground (which goes through these and E301. All looked very similar to the 2D board where the Op amp was working correctly with 2.3 volts on the input and output pins.

So I started experimenting with changing R303 and R304 to different value resistors (again ghetto style due to lack of smd resistors). At first I tried to add two 1.5K resistors. Then I measured the OP amp pins. I noticed the voltage had dropped even lower.

So I then started to try resistors with much lower values than the two original ones that was 1K. I think I first tried two 220 Ohm resistors. Anyway, I noticed the voltage on the op amp pins had increased to around 2.1 volts. I couldn't hear any distortion anymore. But I wanted more voltage. So I ended up with two 47 ohm resistors which brought the voltage up to around 2.3 volts. Which the op amp is happy with. So reduced from two 1K to two 47 ohm resistors and the op amp is working perfectly. Resistance between pin 10 and 12 to ground went from 0.4K ohm in the original config, down to about 37 ohm. And again, the 2D board is working fine with the original 1K ohm resistors at R303 and R304 and around 0.5K ohms to ground. So maybe this is some kind of flaw with the old rev 1.1 board or maybe I'm missing something else causing the voltage drop, and "cheated" ghetto style to get get it working again? Could it be a ceramic cap or something that failed between the op Amp and these resistors? And I forced the voltage up again by putting much smaller resistors at R303 and R304 and severely reducing resistance to ground? I don't know. Any one wanna shed some light on this?

I have no clue what is happening here but it's working. I will try to put back the original op amp now to see if that one is still working aswell.

Edit: Sound is perfect in games I tested so far. But if I just let it be idle at the kickstart screen and turn the volume quite high I can hear some noise and crackling. If I push on the op amp with my finger I can hear a high frequency sound until I let go of it.

Edit2: Seems there was some bad solder joints on the OP amp causing the sound and crackling and the sound when I was pressing on it. I soldered in the original OP amp and it works. Looks like I got good solder joints this time. The resistor mod seems to work. 2.23v range on the in and outs on this op amp. Looks like they must be very resistant to heat. I don't know how many times I've unsoldered and soldered them back and fourth. At least 4 or 5 times for each of them.

[8bitbubsy]

The "digital" background noise is normal, I get it on both my Amiga 1200 and Amiga 600 if I crank the volume up. It's even more audible with accelerators and stuff connected, I guess it adds more noise to the voltage rails?

[turrican9]

Originally Posted by 8bitbubsy

The "digital" background noise is normal, I get it on both my Amiga 1200 and Amiga 600 if I crank the volume up. It's even more audible with accelerators and stuff connected, I guess it adds more noise to the voltage rails?

Yeah, but it was quite bad. Look at my last post. I just updated it with an 'Edit2'

[8bitbubsy]

Ah I see.

[demolition]

Originally Posted by turrican9

First of all I found out that resistor R301 that was supposed to be 10 Ohm was infact showing thousands of Kilo Ohms. And that is why I had no +12V on the op amp. Resistor R302 which is set on the side that feeds the op amp with -12v was also supposed to be 10 omh. It showed like 50 ohm.

Are you measuring those resistors in-circuit? Because you can only trust the measurement if they are unconnected (or at least one side). And it seems odd that resistors would go bad as they don't usually do that. If they do, it is usually from overheating and then it is normally easy to spot the discoloration.
If the opamp is shorted somehow, it could drive its supply low (because you have resistors in the path which increases the impedance), making it look like the supply is the culprit when in fact it is caused by something in the opamp circuit. - Thinking about it, perhaps a shorted opamp might kill the supply resistors if it had been left running like that for a long time.

Originally Posted by turrican9

So reduced from two 1K to two 47 ohm resistors and the op amp is working perfectly. Resistance between pin 10 and 12 to ground went from 0.4K ohm in the original config, down to about 37 ohm. And again, the 2D board is working fine with the original 1K ohm resistors at R303 and R304 and around 0.5K ohms to ground. So maybe this is some kind of flaw with the old rev 1.1 board or maybe I'm missing something else causing the voltage drop, and "cheated" ghetto style to get get it working again? Could it be a ceramic cap or something that failed between the op Amp and these resistors?

AVref should be ~2.5V even if you used 100kOhm resistors for R303+R304 as AVref is not loaded at all DC-wise. It is only loaded by C305+C306 and two opamp inputs (which is >megaohm). So if the voltage is driven low then either one of the two caps are bad (high leakage current), the opamp burned somehow or you have a short on the PCB. Making those resistors smaller may seem to be working, but it is just 'feeding' the leak and allowing it to continue and perhaps something will overheat and go pop at some point. Actually with 47+47 ohm on 5VDC, you are pushing 1/8W in each resistor and that is probably just at the limit of what they were designed for (the small leaded resistors are rated for 1/8W).

I would go back to 1kOhm and then hunt for the current leak by removing components until the leak disappears. My guess would be C305 as I have seen this one go bad before and have heard from others with the same experience. Also a bit weird though since ceramic caps don't go bad very often unless they were not soldered correctly or experience physical stress or vibrations.

[turrican9]

Originally Posted by demolition

Are you measuring those resistors in-circuit? Because you can only trust the measurement if they are unconnected (or at least one side). And it seems odd that resistors would go bad as they don't usually do that. If they do, it is usually from overheating and then it is normally easy to spot the discoloration.
If the opamp is shorted somehow, it could drive its supply low (because you have resistors in the path which increases the impedance), making it look like the supply is the culprit when in fact it is caused by something in the opamp circuit. - Thinking about it, perhaps a shorted opamp might kill the supply resistors if it had been left running like that for a long time.


AVref should be ~2.5V even if you used 100kOhm resistors for R303+R304 as AVref is not loaded at all DC-wise. It is only loaded by C305+C306 and two opamp inputs (which is >megaohm). So if the voltage is driven low then either one of the two caps are bad (high leakage current), the opamp burned somehow or you have a short on the PCB. Making those resistors smaller may seem to be working, but it is just 'feeding' the leak and allowing it to continue and perhaps something will overheat and go pop at some point. Actually with 47+47 ohm on 5VDC, you are pushing 1/8W in each resistor and that is probably just at the limit of what they were designed for (the small leaded resistors are rated for 1/8W).

I would go back to 1kOhm and then hunt for the current leak by removing components until the leak disappears. My guess would be C305 as I have seen this one go bad before and have heard from others with the same experience. Also a bit weird though since ceramic caps don't go bad very often unless they were not soldered correctly or experience physical stress or vibrations.

Yes, I did measure the resistors out of circuit. Same result. The seller of this board said that it had been reflowed several times. So I was actually thinking these components had been overheated. There was some slight discoulouration in that area. When I first got this board I noticed resistor R102 was snapped in half. Probably due to push-fitting the 4MB RAM expansion (mentioned by the seller) the wrong way around, or it did not have the proper cutout in the PLCC to clear R102. It was not easy to spot, so my guess is this resistor had caused alot of issues and they had tried to reflow the board several times. He told me he had been sending the board back and fourth to a repair shop for reflowing.

It's very unlikely the op amp is shorted since I also tried the op amp from my 2D board and it requires the same resistors when installed in the 1.1 board to get enough voltage. And this op amp is now back into the 2D board with the two 1K resistors and working fine.

C305 is an interesting suggestion. This board had been recapped. So all electrolytic. That includes C306 which is next to C305. There is discolouration on the soldering joints of C305. This is probably from the leakage of the old C306. So C305 could very likely be eaten by electrolyte acid.

What I will do is solder back the two 1K resistors, I will find what size C305 is and then try another cap there.

Thanks for your suggestions. This is why I contacted you. I knew my "fix" probably wasn't very good, even if it was working for now.

Edit: Looking at the audio circuitry diagram it looks like C305 is a 0.33 uF cap. And looking at the diagram it makes alot of sense now. This is what's feeding the audio circuit. So it has to be either of these caps causing the issue. C306 should be a new one. The soldering joints on it looks fresh. Says it's recapped in June of this year. But I will keep C306 in mind aswell.

[demolition]

Originally Posted by turrican9

What I will do is solder back the two 1K resistors, I will find what size C305 is and then try another cap there.

If you can't find the replacement easily, just remove it to begin with and listen - the circuit will still work albeit maybe with slightly more noise (not distortion though).

[turrican9]

Originally Posted by demolition

If you can't find the replacement easily, just remove it to begin with and listen - the circuit will still work albeit maybe with slightly more noise (not distortion though).

So I put back the two 1K resistors. I desoldered C305. It was not easy because the joints were corroded.

I borrowed C305 from the working 2D board and soldered it in the 1.1 board.

It works. It was C305 that was the culprit.

I now have 2.33 Volts on the in and out pins on the op amp. Which is the same I get in my 2D board.

Thank you! I should have suspected that cap myself. I had printed out the schematics for the audio circuitry. And it was right next to C306, and had dark solder joints. I was probably thinking it was very unlikely that a ceramic cap had gone bad. Very happy I contacted you.

Also, the resistance from pin 10 on op amp to ground are now around 0.5K ohm. Same as the 2D board. With the bad cap and the 1K resistors I had like 0.4K Ohm (bit lower sometimes) . So that was one sign.



All the trouble I could have saved myself... LOL.

Anyway, thanks again!

Edit: Of course the resistors in the +12v and -12V circuitry for the op amp was bad aswell. R301 was completey shot. should be 10 ohm but was thousands of kilo ohms. R302 was like 80 Ohm and again, should be 10 ohm. And again, I measured the resistors out of circuit aswell. So I was missing the positive voltages to the OP amp. So there was more than just that cap. Funnily enough I had the +12v to the op amp the first days after I got this board. But R301 looked quite baked so when I started working on the board it probably completely broke. So it was very confusing when I suddenly lost the positive voltages to my op amp. So my work was not in vain.

[demolition]

Originally Posted by turrican9

All the trouble I could have saved myself... LOL.
...
So my work was not in vain.

Well, it is not just about the destination but also the journey.
I have also tried spending countless hours on something that ended up being a really simple fix, but I see that time spent as a mental investment.

[turrican9]

Originally Posted by demolition

Well, it is not just about the destination but also the journey.
I have also tried spending countless hours on something that ended up being a really simple fix, but I see that time spent as a mental investment.

For sure. I have a much better understanding of the audio circuitry in Amiga 600s now. It will be almost identical in A1200s. So I should be very comfortable doing repairs of the audio circuitry in the future.

Also, this was my first time de-soldering and re-soldering smd chips (op amp). So I know how to do that comfortably now.

And this board had more trouble than just cap C305 as I said. R102 split on the middle probably because someone tried fitting a 4MB RAM expansion the wrong way and R301 and R302 was shot.