Hoping for some help from an expert. I'm more repair than design. ;-)

It is a 70s Lucas Rover ECU. An Oz spec car in the UK, so different part
number to the UK ones. And chances of finding a good replacement, rare. It
does work, but not correctly.

The ECU is entirely analogue.

I agreed to have a look for anything obvious like dry joints etc.

On powering it up off the car, the Vcc is low. Should be 5v, but is
reading 4.4v. Vref derived from the 5v should be 4.3 and that feeds all
the various engine sensors - CTS, ATS, AFM, TPS etc. So I'm guessing an
error there would effect the fuelling.

Here is the voltage reg circuit. Not official Lucas. But traced out by a
Triumph enthusiast in the US - the TR8 uses a similar ECU. So hopefully
the same PS.

https://i.ibb.co/9HjN34H/5V.jpg

ZD2 appears to be faulty. Removed from the board, reads 2M7 one way,
infinity the other.

But is marked Lucas ZD10248407 and Google can't find it - not
surprisingly.

I've checked every other component and all good. One alteration from this
circuit is D2 is replaced with a link.

So hoping someone can work out for me what it should be to source a
replacement.

If it were mine, I'd likely just replace the PS with a modern single IC
reg. But if I can find a replacement zener, much simpler.

--
*It's this dirty because I washed it with your wife's knickers*

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.

"Dave Plowman (News)" <dave@davenoise.co.uk> wrote:
> On powering it up off the car, the Vcc is low. Should be 5v, but is
> reading 4.4v. Vref derived from the 5v should be 4.3 and that feeds all
> the various engine sensors - CTS, ATS, AFM, TPS etc. So I'm guessing an
> error there would effect the fuelling.
>
> Here is the voltage reg circuit. Not official Lucas. But traced out by a
> Triumph enthusiast in the US - the TR8 uses a similar ECU. So hopefully
> the same PS.

Where's the input to the regulator? Is that the vertical line at the top
right? Is it battery 12V, or does it go through other components on the way?
What is the component on the left side attached to the collector of Q3?

> https://i.ibb.co/9HjN34H/5V.jpg
>
> ZD2 appears to be faulty. Removed from the board, reads 2M7 one way,
> infinity the other.
>
> But is marked Lucas ZD10248407 and Google can't find it - not
> surprisingly.
>
> I've checked every other component and all good. One alteration from this
> circuit is D2 is replaced with a link.

My guess is it's something like a 3.6v zener with another diode in series to
make 4.3v, and then a base-collector drop in Q1 to make 5v. But I'm not
sure of the feedback arrangement, nor whether it's relevant given the Q3
tap. It's very hard to tell anything without context.

Theo

In article <RJg*hVLLy@news.chiark.greenend.org.uk>,
Theo <theom+news@chiark.greenend.org.uk> wrote:
> "Dave Plowman (News)" <dave@davenoise.co.uk> wrote:
> > On powering it up off the car, the Vcc is low. Should be 5v, but is
> > reading 4.4v. Vref derived from the 5v should be 4.3 and that feeds all
> > the various engine sensors - CTS, ATS, AFM, TPS etc. So I'm guessing an
> > error there would effect the fuelling.
> >
> > Here is the voltage reg circuit. Not official Lucas. But traced out by a
> > Triumph enthusiast in the US - the TR8 uses a similar ECU. So hopefully
> > the same PS.

> Where's the input to the regulator? Is that the vertical line at the
> top right? Is it battery 12V, or does it go through other components on
> the way?

Shown as being direct from the 12v input.

> What is the component on the left side attached to the collector of Q3?

Goes to pin 11 on the processor IC. An LFC1041AE

> > https://i.ibb.co/9HjN34H/5V.jpg
> >
> > ZD2 appears to be faulty. Removed from the board, reads 2M7 one way,
> > infinity the other.
> >
> > But is marked Lucas ZD10248407 and Google can't find it - not
> > surprisingly.
> >
> > I've checked every other component and all good. One alteration from this
> > circuit is D2 is replaced with a link.

> My guess is it's something like a 3.6v zener with another diode in
> series to make 4.3v, and then a base-collector drop in Q1 to make 5v.
> But I'm not sure of the feedback arrangement, nor whether it's relevant
> given the Q3 tap. It's very hard to tell anything without context.

Thanks, Theo. If it helps, voltage with the suspect zener in place reads
supply volts negative side, 0.012v the other.

--
*Sometimes I wake up grumpy; Other times I let him sleep.

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.

Dave Plowman (News) <dave@davenoise.co.uk> wrote:
> Hoping for some help from an expert. I'm more repair than design. ;-)
>
> It is a 70s Lucas Rover ECU. An Oz spec car in the UK, so different part
> number to the UK ones. And chances of finding a good replacement, rare. It
> does work, but not correctly.
>
> The ECU is entirely analogue.
>
> I agreed to have a look for anything obvious like dry joints etc.
>
> On powering it up off the car, the Vcc is low. Should be 5v, but is
> reading 4.4v. Vref derived from the 5v should be 4.3 and that feeds all
> the various engine sensors - CTS, ATS, AFM, TPS etc. So I'm guessing an
> error there would effect the fuelling.
>
> Here is the voltage reg circuit. Not official Lucas. But traced out by a
> Triumph enthusiast in the US - the TR8 uses a similar ECU. So hopefully
> the same PS.
>
>
> https://i.ibb.co/9HjN34H/5V.jpg
>
> ZD2 appears to be faulty. Removed from the board, reads 2M7 one way,
> infinity the other.
>
> But is marked Lucas ZD10248407 and Google can't find it - not
> surprisingly.
>
> I've checked every other component and all good. One alteration from this
> circuit is D2 is replaced with a link.
>
> So hoping someone can work out for me what it should be to source a
> replacement.
>
> If it were mine, I'd likely just replace the PS with a modern single IC
> reg. But if I can find a replacement zener, much simpler.
>

Would a bit of help trying to determine the kind of zener (ZD2) you need
help?

I’ll talk you through some basic analysis of part of the circuit:

Start with Q3: The base / emitter voltage will be between 0.6 and 0.7 V.

Therefore the voltage across R9 is between 0.6 and 0.7 V.

Therefore current in R9 between 0.6 and 0.7 mA

We can ignore the base current, so this flows through R6, R8, R9, and ZD2.

We can now find the voltage at the junction of R6/ZD2

R6+R8+R9 = 15300 ohms.

At R6/ZD2 between Vmin=15300x0.0006
and Vmax=15300x0.0007

So between: 9.18V and 10.71 V

The other end of the zener is at nominal battery / alternator voltage.

The question is, what have they assumed that is.

Well, zeners come is standard values and you have a rough idea what the
battery voltage is.

A 5.1 V one on the 10.71V would suggest a battery / alternator of 15.81V -
bit high.

A 3.3 V one on the 10.71 would be 14.01 V
& 12.48 on the 9.18 limit.

Promising but possibly a bit low.

A 3.9V zener on 10.71 gives 14.61V
& 13.09 on 9.81 - more promising.

A 4.3 zener take you over 15V, which I think is high but it could be…

I think it is probably a 3.9V zener. As for rating, it doesn’t seem to be
passing much current but look for one of similar size.

As for the existing zener, it should ‘look’ like a diode if tested on a
meter with a diode test function BUT it will appear to be miss-labelled (ie
the wrong way around). That is normal - seriously.

Brian

In article <t3c4kr$5io$1@dont-email.me>,
Brian <noinv@lid.org> wrote:
> As for the existing zener, it should look‘ like a diode if tested on a
> meter with a diode test function BUT it will appear to be miss-labelled (ie
> the wrong way around). That is normal - seriously.

Can you explain that, Brian? All the zeners I have here are marked with a
band at one end - same as diodes. And on my DVM diode test function band
to negative works with both.

--
*What was the best thing before sliced bread?

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.

Dave Plowman (News) <dave@davenoise.co.uk> wrote:
> In article <t3c4kr$5io$1@dont-email.me>,
> Brian <noinv@lid.org> wrote:
>> As for the existing zener, it should look‘ like a diode if tested on a
>> meter with a diode test function BUT it will appear to be miss-labelled (ie
>> the wrong way around). That is normal - seriously.
>
> Can you explain that, Brian? All the zeners I have here are marked with a
> band at one end - same as diodes. And on my DVM diode test function band
> to negative works with both.
>

Ok, first why, then a suggested demo.

Zeners are used as references, they ‘break down’ (conduct) in the REVERSE
direction when the marked voltage is reached. In the other direction, they
will show a 0.6 to 0.7 volt drop, like a normal diode.

Now, take a diode and a zener.

Using a multimeter with a diode test range, apply the probes to each, one
way and the other.

You should see the forward voltage, typically, 0.6 to 0.7 V.

Note which end the probes are applied to in each case.

In article <t3cdru$cur$1@dont-email.me>,
Brian <noinv@lid.org> wrote:
> Dave Plowman (News) <dave@davenoise.co.uk> wrote:
> > In article <t3c4kr$5io$1@dont-email.me>,
> > Brian <noinv@lid.org> wrote:
> >> As for the existing zener, it should look‘ like a diode if tested on
> >> a meter with a diode test function BUT it will appear to be
> >> miss-labelled (ie the wrong way around). That is normal - seriously.
> >
> > Can you explain that, Brian? All the zeners I have here are marked
> > with a band at one end - same as diodes. And on my DVM diode test
> > function band to negative works with both.
> >

> Ok, first why, then a suggested demo.

> Zeners are used as references, they break down‘ (conduct) in the
> REVERSE direction when the marked voltage is reached. In the other
> direction, they will show a 0.6 to 0.7 volt drop, like a normal diode.

> Now, take a diode and a zener.

> Using a multimeter with a diode test range, apply the probes to each,
> one way and the other.

> You should see the forward voltage, typically, 0.6 to 0.7 V.

> Note which end the probes are applied to in each case.

That's what I did. I use the diode test function often.

And on both an ordinary diode and a zener, you connect the meter neg lead
to the band end of the diode. And get a reading of the forward voltage
drop - about 0.6v Reverse the connections and you may get a rubbish
reading or none at all.

--
*It was recently discovered that research causes cancer in rats*

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.

Dave Plowman (News) <dave@davenoise.co.uk> wrote:
> In article <t3cdru$cur$1@dont-email.me>,
> Brian <noinv@lid.org> wrote:
>> Dave Plowman (News) <dave@davenoise.co.uk> wrote:
>>> In article <t3c4kr$5io$1@dont-email.me>,
>>> Brian <noinv@lid.org> wrote:
>>>> As for the existing zener, it should look‘ like a diode if tested on
>>>> a meter with a diode test function BUT it will appear to be
>>>> miss-labelled (ie the wrong way around). That is normal - seriously.
>>>
>>> Can you explain that, Brian? All the zeners I have here are marked
>>> with a band at one end - same as diodes. And on my DVM diode test
>>> function band to negative works with both.
>>>
>
>> Ok, first why, then a suggested demo.
>
>> Zeners are used as references, they break down‘ (conduct) in the
>> REVERSE direction when the marked voltage is reached. In the other
>> direction, they will show a 0.6 to 0.7 volt drop, like a normal diode.
>
>
>> Now, take a diode and a zener.
>
>> Using a multimeter with a diode test range, apply the probes to each,
>> one way and the other.
>
>> You should see the forward voltage, typically, 0.6 to 0.7 V.
>
>> Note which end the probes are applied to in each case.
>
> That's what I did. I use the diode test function often.
>
> And on both an ordinary diode and a zener, you connect the meter neg lead
> to the band end of the diode. And get a reading of the forward voltage
> drop - about 0.6v Reverse the connections and you may get a rubbish
> reading or none at all.
>

By rubbish you mine it indicates open circuit or high resistance? If so, it
is probably ok.

I think I may have confused you.

Let me try again.

On semiconductors, the ‘arrow’ indicates the direction of flow of
conventional current.

In a normal diode, generally, you have conventional current flowing in the
direction of the arrow most of the time. In never follows the other way
(ignoring leakage current).

A zener is normally used so no conventional current flows in the direction
of the arrow. (Some circuits may differ, most don’t.) Instead, the zener
is used in reverse, so it ‘breaks down’ and maintains a constant voltage.

You will note the zener you asked about APPEARS the wrong way up in terms
of the arrow and the flow of conventional current UNTIL you accept it
operates as above.

(Emphasis not for you, it is for the idiots who stalk me from the Radio
group. )

There are actually two modes of break down, one below about 6.7 V one above
it. One has a positive temp coef, one a negative. Around 6.7v you see both
effects and the temp coefficients cancel out. They are still all called
zeners.

Sorry I confused you. Not my best explanation. A busy day.

On 16/04/2022 12:46, Brian wrote:
> Dave Plowman (News) <dave@davenoise.co.uk> wrote:
>> In article <t3cdru$cur$1@dont-email.me>,
>> Brian <noinv@lid.org> wrote:
>>> Dave Plowman (News) <dave@davenoise.co.uk> wrote:
>>>> In article <t3c4kr$5io$1@dont-email.me>,
>>>> Brian <noinv@lid.org> wrote:
>>>>> As for the existing zener, it should look‘ like a diode if tested on
>>>>> a meter with a diode test function BUT it will appear to be
>>>>> miss-labelled (ie the wrong way around). That is normal - seriously.
>>>>
>>>> Can you explain that, Brian? All the zeners I have here are marked
>>>> with a band at one end - same as diodes. And on my DVM diode test
>>>> function band to negative works with both.
>>>>
>>
>>> Ok, first why, then a suggested demo.
>>
>>> Zeners are used as references, they break down‘ (conduct) in the
>>> REVERSE direction when the marked voltage is reached. In the other
>>> direction, they will show a 0.6 to 0.7 volt drop, like a normal diode.
>>
>>
>>> Now, take a diode and a zener.
>>
>>> Using a multimeter with a diode test range, apply the probes to each,
>>> one way and the other.
>>
>>> You should see the forward voltage, typically, 0.6 to 0.7 V.
>>
>>> Note which end the probes are applied to in each case.
>>
>> That's what I did. I use the diode test function often.
>>
>> And on both an ordinary diode and a zener, you connect the meter neg lead
>> to the band end of the diode. And get a reading of the forward voltage
>> drop - about 0.6v Reverse the connections and you may get a rubbish
>> reading or none at all.
>>
>
> By rubbish you mine it indicates open circuit or high resistance? If so, it
> is probably ok.
>
>
> I think I may have confused you.
>
>
> Let me try again.
>
> On semiconductors, the ‘arrow’ indicates the direction of flow of
> conventional current.
>
> In a normal diode, generally, you have conventional current flowing in the
> direction of the arrow most of the time. In never follows the other way
> (ignoring leakage current).
>
> A zener is normally used so no conventional current flows in the direction
> of the arrow. (Some circuits may differ, most don’t.) Instead, the zener
> is used in reverse, so it ‘breaks down’ and maintains a constant voltage.
>
>
> You will note the zener you asked about APPEARS the wrong way up in terms
> of the arrow and the flow of conventional current UNTIL you accept it
> operates as above.
>
> (Emphasis not for you, it is for the idiots who stalk me from the Radio
> group. )
>
> There are actually two modes of break down, one below about 6.7 V one above
> it. One has a positive temp coef, one a negative. Around 6.7v you see both
> effects and the temp coefficients cancel out. They are still all called
> zeners.
>
> Sorry I confused you. Not my best explanation. A busy day.
>
well done brian you are always helpful

On 15/04/2022 12:35, Dave Plowman (News) wrote:
> Hoping for some help from an expert. I'm more repair than design. ;-)
>
> It is a 70s Lucas Rover ECU. An Oz spec car in the UK, so different part
> number to the UK ones. And chances of finding a good replacement, rare. It
> does work, but not correctly.
>
> The ECU is entirely analogue.
>
> I agreed to have a look for anything obvious like dry joints etc.
>
> On powering it up off the car, the Vcc is low. Should be 5v, but is
> reading 4.4v. Vref derived from the 5v should be 4.3 and that feeds all
> the various engine sensors - CTS, ATS, AFM, TPS etc. So I'm guessing an
> error there would effect the fuelling.
>
> Here is the voltage reg circuit. Not official Lucas. But traced out by a
> Triumph enthusiast in the US - the TR8 uses a similar ECU. So hopefully
> the same PS.
>
>
> https://i.ibb.co/9HjN34H/5V.jpg
>
> ZD2 appears to be faulty. Removed from the board, reads 2M7 one way,
> infinity the other.
>
> But is marked Lucas ZD10248407 and Google can't find it - not
> surprisingly.
>
> I've checked every other component and all good. One alteration from this
> circuit is D2 is replaced with a link.
>
> So hoping someone can work out for me what it should be to source a
> replacement.
>
> If it were mine, I'd likely just replace the PS with a modern single IC
> reg. But if I can find a replacement zener, much simpler.

It is a really odd circuit as drawn. You might want to try downloading
LT spice (free) and run it in the simulator. My best guess from looking
at it is that the whole thing looks a little implausible.

My best guess of intended operation is that Q2/Q3 is intended to be a
soft start so that there is a short delay before it powers up pads E,K.
The chain R6,R8,R9 & C5 forming the shorter time constant and R3/C2 the
longer one. The circuit seems to be very clunky to me.

I can put a ballpark estimate on the maximum voltage that ZD2 can be
before it becomes impossible for Q3 to switch on as 14v - 0.6v*15.3 ~
4v. So my guess is that it is meant to be a 3.9v zener diode and that
the resistor chain is supposed to have ~0.6 mA flowing down it.
(varying slightly with the state of charge of the battery)

For the purposes of a quick suck it and see test replacing ZD2 with 6.8k
ought to get something like the right voltages at pads E & K. Regulation
will be terrible though it really wants a zener diode in there.

Check the current draw looks acceptable!

BTW If C2 should ever fail open circuit it will fry ZD1 by putting Q2 &
Q1 into hard conduction.

The more normal classic old school series regulator would have the zener
diode and capacitor in the leg of the base to ground and the beefy
transistor configured as an emitter follower. Or alternatively as a
shunt regulator which is less efficient. Your circuit is neither.

This isn't a bad intro to the two sorts of transistor based regulator:

https://www.tutorialspoint.com/electronic_circuits/electronic_circuits_regulators.htm

or Wiki

https://en.wikipedia.org/wiki/Linear_regulator#Shunt_regulators
>
Hope this is some help.

--
Regards,
Martin Brown

In article <t3jiai$11b$1@gioia.aioe.org>,
Martin Brown <'''newspam'''@nonad.co.uk> wrote:
> On 15/04/2022 12:35, Dave Plowman (News) wrote:
> > Hoping for some help from an expert. I'm more repair than design. ;-)
> >
> > It is a 70s Lucas Rover ECU. An Oz spec car in the UK, so different part
> > number to the UK ones. And chances of finding a good replacement, rare. It
> > does work, but not correctly.
> >
> > The ECU is entirely analogue.
> >
> > I agreed to have a look for anything obvious like dry joints etc.
> >
> > On powering it up off the car, the Vcc is low. Should be 5v, but is
> > reading 4.4v. Vref derived from the 5v should be 4.3 and that feeds all
> > the various engine sensors - CTS, ATS, AFM, TPS etc. So I'm guessing an
> > error there would effect the fuelling.
> >
> > Here is the voltage reg circuit. Not official Lucas. But traced out by a
> > Triumph enthusiast in the US - the TR8 uses a similar ECU. So hopefully
> > the same PS.
> >
> >
> > https://i.ibb.co/9HjN34H/5V.jpg
> >
> > ZD2 appears to be faulty. Removed from the board, reads 2M7 one way,
> > infinity the other.
> >
> > But is marked Lucas ZD10248407 and Google can't find it - not
> > surprisingly.
> >
> > I've checked every other component and all good. One alteration from this
> > circuit is D2 is replaced with a link.
> >
> > So hoping someone can work out for me what it should be to source a
> > replacement.
> >
> > If it were mine, I'd likely just replace the PS with a modern single IC
> > reg. But if I can find a replacement zener, much simpler.

> It is a really odd circuit as drawn. You might want to try downloading
> LT spice (free) and run it in the simulator. My best guess from looking
> at it is that the whole thing looks a little implausible.

> My best guess of intended operation is that Q2/Q3 is intended to be a
> soft start so that there is a short delay before it powers up pads E,K.
> The chain R6,R8,R9 & C5 forming the shorter time constant and R3/C2 the
> longer one. The circuit seems to be very clunky to me.

> I can put a ballpark estimate on the maximum voltage that ZD2 can be
> before it becomes impossible for Q3 to switch on as 14v - 0.6v*15.3 ~
> 4v. So my guess is that it is meant to be a 3.9v zener diode and that
> the resistor chain is supposed to have ~0.6 mA flowing down it.
> (varying slightly with the state of charge of the battery)

> For the purposes of a quick suck it and see test replacing ZD2 with 6.8k
> ought to get something like the right voltages at pads E & K. Regulation
> will be terrible though it really wants a zener diode in there.

> Check the current draw looks acceptable!

> BTW If C2 should ever fail open circuit it will fry ZD1 by putting Q2 &
> Q1 into hard conduction.

> The more normal classic old school series regulator would have the zener
> diode and capacitor in the leg of the base to ground and the beefy
> transistor configured as an emitter follower. Or alternatively as a
> shunt regulator which is less efficient. Your circuit is neither.

> This isn't a bad intro to the two sorts of transistor based regulator:

> https://www.tutorialspoint.com/electronic_circuits/electronic_circuits_regulators.htm

> or Wiki

> https://en.wikipedia.org/wiki/Linear_regulator#Shunt_regulators
> >
> Hope this is some help.

It was designed in the 70s by, I'm told, Lucas Aerospace.

--
*Do infants enjoy infancy as much as adults enjoy adultery?

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.