Mooly is suggesting you put reverse biased diodes IN .....something like 1N4001 (reverse biased ) from each rail to earth after the regs on the regulated output sides . This will probably cure your problem......especially if your following circuitry has rail capacitors connected directly from rail to rail.
Hi put in diodes as per national data sheet.diodes position with banded end towards positive rail.
It seems to be better will turn on first time.
but after switching off it will only come back on within 3 seconds or after 10 seconds.
the times it doesnt come on and is switched off it powers up briefly as the power drains away
It seems to be better will turn on first time.
but after switching off it will only come back on within 3 seconds or after 10 seconds.
the times it doesnt come on and is switched off it powers up briefly as the power drains away
You can also try two more diodes... it's a recognised method... by fitting one across each reg between input pin and output pin.
For the 78 reg the banded end goes to the input pin of the reg and the other end to the output pin.
For the 79 reg the NON banded end goes to the input pin and the banded end to the output pin.
Leave the other diodes in place.
This odd problem, and I must admit it's never happened to me, can also be fixed by sometimes using a different brand of reg. But that's not really the answer.
For the 78 reg the banded end goes to the input pin of the reg and the other end to the output pin.
For the 79 reg the NON banded end goes to the input pin and the banded end to the output pin.
Leave the other diodes in place.
This odd problem, and I must admit it's never happened to me, can also be fixed by sometimes using a different brand of reg. But that's not really the answer.
It's caused by the rails coming up at different rates. It all happens in the first few ms or so. Wha happens is that the rail that appears first can (via PSU decoupling caps and caps rail to rail) reverse bias the other rail for an instant and latch the other regulator. I would guess too depending on the other circuitry that it could also cause other parasitic diodes in for example any IC's to conduct and again lock the circuit up.
If you can hold the back of finger on the metal tab for 15 to 20 seconds they are OK. If they are as hot as a radiator at 60/65 C then thats OK.
NE5532's draw more current than the TL072. It can be as high as 16ma worst case per IC compared to around 2 or 3 ma (from memory) for the TL072. So it soon adds up and can cause an unheatsinked reg to run hot wheras with TL072's it would not.
Is the 18 volts the DC voltage going into the regs or the AC voltage from the transformer ?
A 15 vac transformer winding should give more than 18 volts when rectified and smoothed so that needs confirming and checking. Is it a full wave bridge rectifier or half wave (four diodes or two).
Is the 18 volts the DC voltage going into the regs or the AC voltage from the transformer ?
A 15 vac transformer winding should give more than 18 volts when rectified and smoothed so that needs confirming and checking. Is it a full wave bridge rectifier or half wave (four diodes or two).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.