Issue 3B problem with voltages

[deano467]

Can you please check the polarity of the two electrolytic capacitors that are fitted as part of the DC/DC converter upgrade.
One (C78, 22µF) I'm not sure of the polarity from the photo. The second (C79, 1µF, in position D16) looks like it may be fitted with the incorrect polarity.

Mark

Hi Mark, the C78 is correct polarity, but it does seem I fitted C79, 1uF with incorrect polarity.

Unbelievable and well spotted by yourself. I will reverse the polarity hopefully later tonight if I get a chance, and see if that fixes my problem. If not I will do all the other tests you suggest.
Thanks very much.. you are so helpful and I appreciate this so much

[1024MAK]

You may need to fit a new capacitor. Electrolytic capacitors don't like a reverse voltage. But if you don't have a spare, it is worthwhile trying the existing capacitor, as it may have survived.

Mark

[deano467]

So having another look at this. I'm getting 0.71v on this test. Does that mean the TR5 may need replaced?Screenshot_2024-01-29-20-35-23-450-edit_com.android.chrome.jpg

This appears to be in circuit testing with the power off. I need to check this before answering.

However, I would like you to carry out these tests with the power on:
Multimeter black/negative lead/probe to C44 negative lead.

• TR5 emitter (should be the same voltage as the +5V rail) 4.94v
• TR5 base 4.14v
• TR5 collector 4.82v
• R59 left hand lead (should be the same as TR5 base) 4.14v
• R59 right hand lead 9.48v
• R58 left hand lead (should be the same as TR5 base) 4.15v
• R60 right hand lead (should be the same as TR5 collector) 4.81v
• R60 left hand lead -1.0v
• R61 left hand lead -1.0v
• R61 right hand lead -1.4v

Also check the polarity of C74. Correct

Then with the power still on, test the voltage across R62. That is with the meter positive/red lead on R62 right hand lead and the meter negative/black lead on R62 left hand lead. -0.05v

Mark

I fitted a new capacitor C79 correct orientation, started up to Sinclair logo but still no colours.

Did the power on tests with multimeter...results above which don't look right.

Now lost the Sinclair logo start up, so it's got worse

[1024MAK]

Based on those results, TR5 looks to be hard on, that is, it's trying to in increase the +12V rail voltage. So it's not faulty, but is okay.

I suspect you have a "lower" DRAM chip fault.

Are any of these 4116 (or equivalent)(IC6 to IC13) DRAM chips getting hotter than the others?

Are you okay with cutting a PCB track and cutting or lifting a pin on each of the 4116 DRAM chips?

Mark

[deano467]

I suspect you have a "lower" DRAM chip fault.

Are any of these 4116 (or equivalent)(IC6 to IC13) DRAM chips getting hotter than the others?

Are you okay with cutting a PCB track and cutting or lifting a pin on each of the 4116 DRAM chips?

Mark

I ran the diagnostics from dandantor a few days ago when I first had the monochrome screen issue and lower ram passed.

Unfortunately now I have a fuzzy screen so don't think the diagnostics are any help now.

I will test for hot to touch chips later.

[1024MAK]

Well, either there is a problem with the DC/DC converter circuitry that is limiting the amount of current flowing to the +12V line, or one or more of the 'lower' DRAM chips is drawing more current than it/they should.

The only way to tell beyond doubt is to cut the PCB track for the +12V supply going to the 'lower' DRAM chips. Or cut the +12V pin on each chip. Or undo your DC/DC modification and change it so we can separate the +12V feed to isolate the feed to the 'lower' DRAM chips (which is difficult due to the layout of the components on the board).

Unless you happen to have a bench PSU (with adjustable current limit) that can supply +12V.

Mark

[deano467]

The only way to tell beyond doubt is to cut the PCB track for the +12V supply going to the 'lower' DRAM chips.

Mark

I can do this as it is reversible, where would be the best place to do this on the 3B board?

[1024MAK]

Track to cut to isolate +12V rail to DRAM

Track to cut to isolate +12V rail to DRAM

Mark

[deano467]

Thanks I will try this tomorrow evening

Just to be sure cutting this track is enough... I don't need to lift or cut any legs on memory chips when I do this?

[1024MAK]

No, nothing else is needed at this stage. Cutting the marked track (clear of the pad for C44) should isolate all eight DRAM chips from the +12V supply. Obviously when the track is cut, this DRAM will no longer work, so the machine will not "boot".

Measure the voltages like before. The if the +12V is now back to a normal level and the -5V is better, then it's very likely that one of the 4116 (or equivalent) is faulty.

Before cutting the track, do try powering on for a few minutes and seeing if any of the 4116 DRAM are hotter than the rest first. If one is significantly hotter, that could be the faulty DRAM chip.

Do be careful when cutting the track. The cut needs to be all the way through, but the width of the cut only needs to be just enough to break the track.

Mark