Retired Texas TI99/4A engineer needs help please.

[boater sam]

At my wits end with this TI power board.

I had a thread running on another UK Vintage forum as "Retired T!99/4A engineer needs help please" before I was banned and had some great help and advice from members there. Even had some replacement chips donated, thank you.

The circuit is here, Page 19 in the manual.

  Texas TI-99-4A manual..pdf (Size: 7.17 MB / Downloads: 36)

It was originally presented with R6, 100 ohm burnt out.  Q2, TIP31  & C15 were replaced, D10 was replaced with 1N4148.

I have replaced the UA723 and the TL331 ( with part of LM339 ) and now I have +5.124v on the regulated output until I put any load on it, when it collapses to a few tens of mV.
Removing the load it recovers slowly over 4 or 5 seconds to the same voltage.  The TL331 seems not to have been faulty.

I cannot understand this, any help please?

Sam

[ppppenguin]

Can you measure the current on the 5v rail? Or hook it up. to a bench psu?

[boater sam]

There is no computer attached. I have loaded the +5v with 15 ohms, it should be able to supply 1.8A minimum.
The 5v disappears straight away.

[Nick]

There is no computer attached. I have loaded the +5v with 15 ohms, it should be able to supply 1.8A minimum.
The 5v disappears straight away.

Check R9 and R10. They control the current limit. If they've gone high,  that'd explain it.

[Refugee]

I would guess that there is a broken track on the board around those pass transistors.
Measure the two VCE readings and work out what transistor is not driving its load or just go through with the low ohms range with the power off.

[boater sam]

Thanks guys.

Can find no broken tracks, being TI its a good quality board, I'll check again though.

I have checked every resistor, capacitor and diode around that regulator, all are OK.

[Refugee]

Tracks sometimes crack around the edge of the solder.
Use a meter on low ohms to check each soldered joint.

[boater sam]

The slow recovery of the voltage after the load is removed intrigues me, where is the time constant for this?

I'll pursue this board further when my patience returns. Next time it rains? Got lots of boaty things to occupy my addled mind too.
Sam.

[Refugee]

Your 15 ohm load is forming a potential divider with R6. You will indeed see less that 1 volt. The 723 has a system that shuts it down if it can't make the set voltage in order to stop overheating if a short circuit happens when it is unattended. This will give you the time delay you have seen.
There is surely a break in the circuit around B or E of the TIP31 to make that happen.
Q1 is probably all right as it is regulating to the correct voltage off load.

[ppppenguin]

Can't see any obvious time constants of the order of seconds. The recovery time constant may be thermal rather than electrical, as implied though not explicitly stated by Refugee in his last post.

If this was a "get it working" job rather than a "restore to near original" I'd be very tempted to whack a small 5V switchmode regulator in there. Or a higher power single chip analogue regulator.