Problems with +5V regulated power supply

[hafrse]

Hello,
I have an Otari R2R tape recorder and had problems with the +5V regulated power supply (around 10A) where it passed through 9.1V to the control circuit electronics and burned the EPROM (27256) + other IC:s.
Attached an image of the power supply :

Q1,Q2 is a 2SA1020 transistors.
IC-1 is a M5231L regulator.

On of the 2 2SA1020 was shorted which caused a voltage of 9.1V and replaced , now the power supply is ok.

My question is, how can I build an overvoltage protection like above +6V the the +5V supply will be shorted and the built in automatic fuse in the recorder will disengage because of the short ? I was thinking to add a zener diode but could not find with high current support (?) ?
Thanks

 

[crutschow]

Below is the LTspice sim of an SCR crowbar overvoltage circuit that shorts the output when an overvoltage is detected:
The TL431 voltage reference conducts and turns on the transistor Q1 when its REF voltage reaches 2.5V, which triggers the SCR to short the output.
This is 6V at the output for the values of R2 and R4 selected.

Removing power resets it.

 

[KlausST]

Hi,

Crutschows solution is ruggedand valid. However it causes a short circuit.
I recommend to add a fuse in series with the power path that trips when the triac is active.

Klaus

 

[crutschow]

Crutschows solution is ruggedand valid. However it causes a short circuit.
I recommend to add a fuse in series with the power path that trips when the triac is active.

But the TS wanted a OVP that would short the output (below), which has an automatic fuse, so I don't see that an added fuse is needed.

how can I build an overvoltage protection like above +6V the the +5V supply will be shorted and the built in automatic fuse in the recorder will disengage because of the short ?

 

[hafrse]

Thanks!
It is actually a circuit breaker at 10A . I will add a fuse for more security , shall I add it between the triac and the emittor of the 3906 ?

 

[crutschow]

I will add a fuse for more security , shall I add it between the triac and the emittor of the 3906 ?

No.
It's an SCR not a triac.
But add it between the supply and the OVP circuit where I show a fuse in my schematic.

 

[KlausST]

But the TS wanted

You are right.
I focussed on the given schematic only.

Klaus

 

[D.A.(Tony)Stewart]

Did it fail on abrupt turn-on? Or while in operation?

As I recall this series bypass regulator design might overshoot voltage with no load when turned on near the peak AC voltage. But more likely the pre-driver shorted from overheating.

It's hard to guess why yours failed.

But my guess is the 2SC2010's overheated then shorted-out driving the bigger 2SB763's full input 11.7V voltage on the 5V bus.

So my suggestion is to replace these 1W Q's or replace them with TO-220 IC's with high hFE like the
PNP https://rocelec.widen.net/view/pdf/surzqsf4wb/SNYOS08273-1.pdf?t.download=true&u=5oefqw

So my suggestion prevents the potential cause of failure rather than a crowbar.

 

[KlausST]

Hi,

I wouldn´t be surprised if an old electrolytics capacitor (reduced capcitance) causes oscillation.

Klaus

 

[hafrse]

@D.A.(Tony)Stewart : Yes, it failed on turn on, probably the 2sa1020 was already exhausted, I will also replace to TO-220 thanks

 

[D.A.(Tony)Stewart]

FWIW: If it were 10V in and 5Vout @ 10A out the regulator draws 100W and outputs 50W with the bypass PNP and pre-driver about 10% of the total wasted power. Recalc for 9.1V in

 

[hafrse]

Hi,

I wouldn´t be surprised if an old electrolytics capacitor (reduced capcitance) causes oscillation.

Klaus

Yes, that could be the case, will check the filter capacitors, however, OVP is necessary to protect the digital circuits,

FWIW: If it were 10V in and 5Vout @ 10A out the regulator draws 100W and outputs 50W with the bypass PNP and pre-driver about 10% of the total wasted power. Recalc for 9.1V in

If you please look at the power supply schematics below, +11 is applied to the regulator, the requirement is to monitor the output of the regulator for voltage above +6V and cut when that happens. The 9.1V was discovered at the output of the regulator when one of the 2SA1020 shorted (the overvoltage incident) but in the future if similar incident happens (hope not ), the voltage can be different above +6V.

 

[hafrse]

Did it fail on abrupt turn-on? Or while in operation?

As I recall this series bypass regulator design might overshoot voltage with no load when turned on near the peak AC voltage. But more likely the pre-driver shorted from overheating.

It's hard to guess why yours failed.

But my guess is the 2SC2010's overheated then shorted-out driving the bigger 2SB763's full input 11.7V voltage on the 5V bus.

So my suggestion is to replace these 1W Q's or replace them with TO-220 IC's with high hFE like the
PNP https://rocelec.widen.net/view/pdf/surzqsf4wb/SNYOS08273-1.pdf?t.download=true&u=5oefqw

So my suggestion prevents the potential cause of failure rather than a crowbar.

Today I had time to check the power supply, Q2 which was replaced together with Q1 as described before and Q2 is getting hot so that I can not leave my finger on, the 2SB863 and Q1 is cold. Is that normal for Q2 to get that hot by this regulator design ? a leakage in the 2SB863 ? if I disengage the load from the +5V , it cools up.

 

[D.A.(Tony)Stewart]

Not normal unless no heatsink. Both Q2 + Q3 need a heatsink & insulator.

Q2 is base driver should only draw <=10% current of main series bypass (call it Q3) and thus draw < 10% Pd of Q3. Retest part.

Q1 does very little with 150 and acts as surge input voltage booster.

 

[D.A.(Tony)Stewart]

It may be much cooler and more efficient to drop in a 25W SMPS if that works for you.

Pololu 5V, 6A Step-Down Voltage Regulator D24V60F5

This synchronous switching step-down (or buck) regulator takes an input voltage of up to 38 V and efficiently reduces it to 5 V with an available output current of around 6 A. Typical efficiencies of 80% to 95% make this regulator well suited for higher-power applications like powering motors...

www.pololu.com