A5000 - diving in at the deep end!

[mrbombermillzy]

@GadgetUK164 Im glad you found a system that works.

After the post before yours, I looked at several other Archie mice on Ebay and a high number had either this same toffee ball syndrome, or the ball was missing completely (I'm guessing the ball had disintegrated completely, as opposed to the ball being taken out and lost).

In which case, your YT vid may have plenty more views yet to come! :)

[rubber_jonnie]

So, the A5000 PSU has been having a bit of a bad time, and I've been trying to diagnose it. It's suffering from the same issues as my Mac Classic was and starting with a low 12v rail so the HDD wouldn't spin up.

Having already done the caps, that were leaky and smelling like Billingsgate Fish Market on a hot summers day, I felt it couldn't be those, and was faced with the same issue on the Mac. Caps done and no change.

Unfortunately, I could not find a schematic for the PSU anywhere (A PhiHong PSA-1400RW) so it's been a case of best guessing what's wrong.

I started with the main switching transistor (IIRC this was the problem with the Mac), but no change, so then moved onto an SCR that's part of an array of 2 rectifier packs and the SCR attached to a large heatsink.

Most of the replacement order has been the result of the order in which parts arrive, as it's been a job and a half to identify replacements for the 2 out of the 4 items.

The SCR and switching transistor were available as original parts, so that was OK, but didn't fix the problem. I was then able to replace the SLP2040 with an MBR4045 PT. That works nicely too, but the problem is still there.

So that just leaves the FEP16CT, which has been the most difficult to find, but I can use a MUR1620CT, which is on the way. However, I did find that the MUR2045CT is extremely close, and I have one, so I'm going to try it out whilst I wait for the 1620 to come and see if it resolves the problem.

More to come.

[stephen_usher]

I wonder how close the design is to the TT/MegaSTE's PSU it is. If PhiHong is anything like Astec they'd basically just revise a previous design for new ones so you can use the basic schematic for one to diagnose another in the same family.

[rubber_jonnie]

I wonder how close the design is to the TT/MegaSTE's PSU it is. If PhiHong is anything like Astec they'd basically just revise a previous design for new ones so you can use the basic schematic for one to diagnose another in the same family.

I think the A5000 PSU is probably closer to a PC, but I'll certainly give the TT/MSTE PSU a look, you neve know, I might get lucky.

[rubber_jonnie]

I wonder how close the design is to the TT/MegaSTE's PSU it is. If PhiHong is anything like Astec they'd basically just revise a previous design for new ones so you can use the basic schematic for one to diagnose another in the same family.

Well, that's all the MSTE/TT PSU schematics reviewed, the closest is probably the 8012 variant that does indeed use a FEP16CT fast rectifier, and there are other similarities, but I don't think enough to help me.

I'm hoping that the last replacement rectifier will turn up tomorrow so I can fit it, we'll see if it makes a difference. I'm tempted to try and draw up a schematic myself from photos of the board, I know how much the schematics helped for the Mac Classic. What I won't be able to determine are the inductor values and the transformer turns, but it'd be a start.

[exxos]

Do you have images of the top and bottom of the PSU @rubber_jonnie ?

Also when you say 12v is low, how low exactly ? Some can get as low as 10v. Also is 5v OK ?

Assume you did a diode test on the diode pack ? Also scope both input sides to see if you have voltage both sides.

Traced the diodes back to the windings on the transformer ? Resoldered the pins ? Checked coil resistance ?

[rubber_jonnie]

Do you have images of the top and bottom of the PSU @rubber_jonnie ?

No but I can get some tomorrow.

Also when you say 12v is low, how low exactly ? Some can get as low as 10v. Also is 5v OK ?

5v and -5v are fine. 12v is coming up about 10.5v at a cold start (Left powered off for a minimum of 30 mins) After a few minutes it will rise to 11.5V and eventually will reach just over 12v. This stops the HDD spinning up right away and prevents booting.

Assume you did a diode test on the diode pack ? Also scope both input sides to see if you have voltage both sides.

Yes, the diode pack is new, the original had one failed diode. The SCR, switching transistor and other smaller diode pack are also new, as are all the capacitors. I haven't scoped both sides yet as there is a huge coil in front of the diode pack and it's a bit hairy running with the PCB hanging out of the PSU. I'll have to solder in some wires for that.

Traced the diodes back to the windings on the transformer ? Resoldered the pins ? Checked coil resistance ?

That's in progress, but I see no signs of bad solder joints yet using a magnifying glass, but I can check with my microscope. I haven't checked coil resistance (I presume you mean the transformer). There are a lot of large inductors on it, I'll see if they have good solder joints too.

It is literally a case of it not getting up to 12v and the HDD not spinning because of it. Leave it a while and it's fine.

The UC3844 PWM controller and PC111 photocoupler are also new BTW.

[exxos]

@rubber_jonnie sounds odd the 12v drifting like that. Regulation is probably only on the 5v side . so if 5v never rises like the 12v then it rules out a lot of stuff. It almost impossible for the 12v to drift on its own.

The diode pack will have 2 inputs back to the tranforimer winding, trace those back and see if there's a center tap on the transformer. Either way, measure the coil resistance when its cold, and outputting low voltage. Then let it warm up and measure the coil resistance to see how much its drifting by. I would repeat on the 5v winding as well as a frame of reference.

Also check the diodes inputs and outputs on your scope. Cold and when warmed up. See what the voltages are like.

There isn't much to them really. Indeed you can get a idea from the MSTE PSU schematic on now things are wired up.

I'm wondering really if the transformer core has a partial internal short circuit. Hence measuring voltages and resistances on it.

Also does the 12v drop under load either hot or cold ?

[rubber_jonnie]

@rubber_jonnie sounds odd the 12v drifting like that. Regulation is probably only on the 5v side . so if 5v never rises like the 12v then it rules out a lot of stuff. It almost impossible for the 12v to drift on its own.

The diode pack will have 2 inputs back to the tranforimer winding, trace those back and see if there's a center tap on the transformer. Either way, measure the coil resistance when its cold, and outputting low voltage. Then let it warm up and measure the coil resistance to see how much its drifting by. I would repeat on the 5v winding as well as a frame of reference.

I'll take a look thanks.

Also check the diodes inputs and outputs on your scope. Cold and when warmed up. See what the voltages are like.

I will, but running out of time this week for various reasons, and away next week Monday-Wednesday for Dads funeral :( Also not easy to get a probe on them from the topside due to other components being in the way, so I need to figure out something that doesn't end up shorting stuff out or giving me a shock.

There isn't much to them really. Indeed you can get a idea from the MSTE PSU schematic on now things are wired up.

I'm wondering really if the transformer core has a partial internal short circuit. Hence measuring voltages and resistances on it.

I guess it could be, I will see if I can do some testing this week.

Also does the 12v drop under load either hot or cold ?

Ok, so I just did a cold start off load and got: -3.64v, 5.35v and 12.6v.

On load (Mainboard, HDD and FDD connected), cold, 12V started low (10.2v) and jumped up to 12.15v within about 5 seconds. Machine booted and drive spun up just fine within 30 seconds. 5v started at about 4.8/4.9v and eventually settled at bang on 5v. Machine booted just fine. -5v was settled at -4.85v.

I'm going to leave it for an hour or so, make sure it's 100% cold and all caps discharged, then I will try again.

Top of PSU:

PXL_20221122_084628322.jpg

Bottom of PSU (Flipped horizontally - left end of PCB = right end of topside):

PXL_20221122_085041687.jpg

The messy area of darker green is an earlier repair for the 7905 which had clearly been knocked and the pads loose under the board. It is not a quality PCB, that's for sure.

And here are the areas of interest:

Separation.jpg

Red=mains side
Blue=low voltage side
Pale blue=5v
White=12v
Brown=GND
Yellow=transformer connections from high to low voltage side
Purple=2xDiode packs and 1xSCR attached to the large heatsink marked 'Front'. (From the bottom up - SCR, TIC126M, new item replaced with same, Main diode pack, SLP2040P replaced with MUR3040PT and FEP16CT replaced with new MUR2045CT)

The Mains side has a K1117 switching transistor on the smaller heatsink on the other side of the transformer when viewed from above, this has also been replaced with an identical new one.

From what I can determine from the PCB and topside, there is a switching section on the high voltage side that goes through the transformer where it's rectified by the 3 components on the large heatsink.

Incidentally, whilst the -5V line has a 7905 regulator, the 5v side has no regulator so must be derived from 12v somewhere, or maybe directly rectified from the transformer.

[exxos]

Will reply more after lunch but I would resolder the transformer just to be on the safe side.

Looks like the 5v is then feed in series with the 12v coil.. I've seen similar somewhere., will look shortly.