The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.

--
*If one synchronized swimmer drowns, do the rest have to drown too?

Dave Plowman London SW
To e-mail, change noise into sound.

On Nov 8, 10:36*am, "Dave Plowman (News)"
wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.

I know about amplifiers, but little about the pulses/impedances you
have in mind.


NT

On Nov 8, 10:36*am, "Dave Plowman (News)"
wrote:

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?


Buy yourself a stepper motor, an Arduino and a glue-gun. You'll get a
tacho built and you'll find yet another new hobby 8-)

On 08/11/2010 10:36, Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.


Just a thought that might be nonsense. RS232 drivers work down to about
3V and produce up to about 15V. Could you use one of these to convert
the voltage? Can't give you a reference number off-hand but should not
be difficult to find.

Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.


well there are more ways to skin this cat...BUT the basic problem is
where to get 15V without having to generate it. The car supply with the
engine running will be about 14.4v..

The original tacho would have run off the coil primary - 200V or so. If
its the same tacho that used to be fitted to Spridgets and the like, it
is in fact a 0-10mA moving cold meter so you can gut it and build your
own pulse counter.

Or you can make a mini ignition coil and use the flyback with a high
voltage switching transistor.

Audio transformers can be pretty small as well, you know.

You dont want an emitter follower either. No voltage gain.

What occurs to me is taking the diode bridge, smoothing it, and using a
standard inverter from the ECU to get a couple of hundred volts pulse
at the collector.

Or run it off 12V with an inductor in the collector.

On Nov 8, 1:34*pm, John Rumm wrote:
On 08/11/2010 12:02, Peter Scott wrote:



On 08/11/2010 10:36, Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way
to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the
diodes
trick.

Just a thought that might be nonsense. RS232 drivers work down to about
3V and produce up to about 15V. Could you use one of these to convert
the voltage? Can't give you a reference number off-hand but should not
be difficult to find.

One of the MAX232 charge pump TTL to RS232C line drivers would do quite
nicely. They work off single sided supply rails. Only thing to watch is
there can be ENC susceptibility issues with charge pump circuits IIRC.

Make sure the existing tcho is happy with negative inputs, or clamp
the RS232 levels to avoid them.

MBQ

On Nov 8, 12:11*pm, The Natural Philosopher
wrote:
Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.

well there are more ways to skin this cat...BUT the basic problem is
where to get 15V without having to generate it. The car supply with the
engine running will be about 14.4v..

The original tacho would have run off the coil primary - 200V or so.

I think we've been told that the tacho requires 15V.

MBQ

In message , Tim
Streater writes
In article ,
Peter Scott wrote:

On 08/11/2010 10:36, Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.
Just a thought that might be nonsense. RS232 drivers work down to
about 3V and produce up to about 15V. Could you use one of these to
convert the voltage? Can't give you a reference number off-hand but
should not be difficult to find.

My 68k SBC appears to use MAX232 chips:

http://focus.ti.com/lit/ds/symlink/max232.pdf

or just google for max232.

I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!
--
Ian

Man at B&Q wrote:
On Nov 8, 12:11 pm, The Natural Philosopher
wrote:
Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.
The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.
A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?
Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.
well there are more ways to skin this cat...BUT the basic problem is
where to get 15V without having to generate it. The car supply with the
engine running will be about 14.4v..

The original tacho would have run off the coil primary - 200V or so.

I think we've been told that the tacho requires 15V.


No, we haven't. We've been told that it will *work* off 15V.

How accurate it is at that voltage, is another matter.


MBQ

Ian Jackson wrote:
In message , Tim
Streater writes
In article ,
Peter Scott wrote:

On 08/11/2010 10:36, Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted
spark
with 4 coils, and the common way to feed the rev counter off this
is via
four diodes - one from each coil - and a zener and resistor to
limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive
enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev
counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio
transformer got
it working just fine - but wondered if there was a more sanitary
way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at
the rev
counter end - rather than have all the extra wiring on show for the
diodes
trick.
Just a thought that might be nonsense. RS232 drivers work down to
about 3V and produce up to about 15V. Could you use one of these to
convert the voltage? Can't give you a reference number off-hand but
should not be difficult to find.

My 68k SBC appears to use MAX232 chips:

http://focus.ti.com/lit/ds/symlink/max232.pdf

or just google for max232.

I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!


actually that's a good thought. If the tacho mechanism can be isolated
so as not to be connected to chassis, a single audio chip with a bridged
output will give 24v peak off a 12v supply.

Bit overkill, but its a single chip solution.

In message , The Natural Philosopher
writes
Ian Jackson wrote:
In message ,
Tim Streater writes
In article ,
Peter Scott wrote:

On 08/11/2010 10:36, Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is
wasted
spark
with 4 coils, and the common way to feed the rev counter off this
is via
four diodes - one from each coil - and a zener and resistor to
limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive
enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev
counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio
transformer got
it working just fine - but wondered if there was a more sanitary
way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at
the rev
counter end - rather than have all the extra wiring on show for

diodes
trick.
Just a thought that might be nonsense. RS232 drivers work down to
about 3V and produce up to about 15V. Could you use one of these to
convert the voltage? Can't give you a reference number off-hand but
should not be difficult to find.

My 68k SBC appears to use MAX232 chips:

http://focus.ti.com/lit/ds/symlink/max232.pdf

or just google for max232.

I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!


actually that's a good thought. If the tacho mechanism can be isolated
so as not to be connected to chassis, a single audio chip with a
bridged output will give 24v peak off a 12v supply.

Bit overkill, but its a single chip solution.

I was thinking more on the lines of a couple of BC107s (and a million
other similar transistors) - or maybe a 2N1904 and 2N1906 NPN/PNP pair.

Why does the tacho have to be isolated from chassis? I would have
thought that you would feed it with the car's normal 0V (chassis) and
12V battery rails (obviously more like 14V), and it would be triggered
with pulses not exceeding a similar voltage. However, I confess almost
total ignorance on such matters!
--
Ian

Ian Jackson wrote:
In message , The Natural Philosopher
writes
Ian Jackson wrote:
In message ,
Tim Streater writes
In article ,
Peter Scott wrote:

On 08/11/2010 10:36, Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted
spark
with 4 coils, and the common way to feed the rev counter off this
is via
four diodes - one from each coil - and a zener and resistor to
limit the
peak voltage.

The EDIS ECU has a diagnostic output which will drive a sensitive
enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev
counter
needs approx 15 volts.

A quick lash up using an emitter follower and a 1:3 audio
transformer got
it working just fine - but wondered if there was a more sanitary
way to do
things?

Only reason I'd rather do this is neatness. I can hide an amp at
the rev
counter end - rather than have all the extra wiring on show for
diodes
trick.
Just a thought that might be nonsense. RS232 drivers work down to
about 3V and produce up to about 15V. Could you use one of these
to convert the voltage? Can't give you a reference number off-hand
but should not be difficult to find.

My 68k SBC appears to use MAX232 chips:

http://focus.ti.com/lit/ds/symlink/max232.pdf

or just google for max232.

I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!


actually that's a good thought. If the tacho mechanism can be isolated
so as not to be connected to chassis, a single audio chip with a
bridged output will give 24v peak off a 12v supply.

Bit overkill, but its a single chip solution.

I was thinking more on the lines of a couple of BC107s (and a million
other similar transistors) - or maybe a 2N1904 and 2N1906 NPN/PNP pair.

Why does the tacho have to be isolated from chassis? I would have
thought that you would feed it with the car's normal 0V (chassis) and
12V battery rails (obviously more like 14V), and it would be triggered
with pulses not exceeding a similar voltage. However, I confess almost
total ignorance on such matters!

well because if you connect it to chassis you can only generate a 12v
pulse between its input and earth, BUT if you connect its earth to
another output in antiphase, you can double that..


It's known as an H bridge..its how you get 20W out of a chip into a 4
ohm speaker with only 12v to play with.

On Nov 8, 3:07*pm, The Natural Philosopher
wrote:
Man at B&Q wrote:
On Nov 8, 12:11 pm, The Natural Philosopher
wrote:
Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.
The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.
A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?
Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.
well there are more ways to skin this cat...BUT the basic problem is
where to get 15V without having to generate it. The car supply with the
engine running will be about 14.4v..

The original tacho would have run off the coil primary - 200V or so.

I think we've been told that the tacho requires 15V.

No, we haven't. We've been told that it will *work* off 15V.

How accurate it is at that voltage, is another matter.

MBQ

Yes, if it works on 15v its likely to work on 14v too. But why spend a
bunch of time & error on soemthing that works fine already (the
transformer option), even if it is historic technology.


NT

Tabby wrote:
On Nov 8, 3:07 pm, The Natural Philosopher
wrote:
Man at B&Q wrote:
On Nov 8, 12:11 pm, The Natural Philosopher
wrote:
Dave Plowman (News) wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative. I've changed the ignition system to EDIS which is wasted spark
with 4 coils, and the common way to feed the rev counter off this is via
four diodes - one from each coil - and a zener and resistor to limit the
peak voltage.
The EDIS ECU has a diagnostic output which will drive a sensitive enough
rev-counter - but it is a nominal 5 volt pulse output, and my rev counter
needs approx 15 volts.
A quick lash up using an emitter follower and a 1:3 audio transformer got
it working just fine - but wondered if there was a more sanitary way to do
things?
Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.
well there are more ways to skin this cat...BUT the basic problem is
where to get 15V without having to generate it. The car supply with the
engine running will be about 14.4v..
The original tacho would have run off the coil primary - 200V or so.
I think we've been told that the tacho requires 15V.
No, we haven't. We've been told that it will *work* off 15V.

How accurate it is at that voltage, is another matter.

MBQ

Yes, if it works on 15v its likely to work on 14v too. But why spend a
bunch of time & error on soemthing that works fine already (the
transformer option), even if it is historic technology.


I think he explained that: to make it small enough to fit in the dashboard.


NT

On Mon, 08 Nov 2010 10:36:48 +0000, Dave Plowman (News) wrote:

The rev counter on my old SD1 was connected directly to the coil LT
negative.

I'm having a right ol' thicky day today, but for my addled brain, why
15V? Why isn't it just "whatever the alternator is putting out", which
will vary a lot between cars and at different loads, but might be
anywhere down to 11V or so and with 15V as a probably upper design limit?

Or have you fitted something that wasn't originally intended for car use?

cheers

Jules

In article ,
Ian Jackson wrote:
I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!

It would if you had a high enough HT rail. Sadly in a car you have to make
things work from about 10 volts.

--
*Marathon runners with bad shoes suffer the agony of de feet*

Dave Plowman London SW
To e-mail, change noise into sound.

In article ,
The Natural Philosopher wrote:
Yes, if it works on 15v its likely to work on 14v too. But why spend a
bunch of time & error on soemthing that works fine already (the
transformer option), even if it is historic technology.


I think he explained that: to make it small enough to fit in the
dashboard.

I was also hoping to offer it as an option for others with a similar
problem. I had a suitable ratio audio tranny lying around - but they're
quite expensive. A small PCB could be fitted inside the actual rev
counter.

--
*Succeed, in spite of management *

Dave Plowman London SW
To e-mail, change noise into sound.

In article ,
The Natural Philosopher wrote:
Only reason I'd rather do this is neatness. I can hide an amp at the rev
counter end - rather than have all the extra wiring on show for the diodes
trick.


well there are more ways to skin this cat...BUT the basic problem is
where to get 15V without having to generate it. The car supply with the
engine running will be about 14.4v..

I'd like the rev counter to work under all possible conditions. So lets
say just over 10 volts or so - when cranking on a cold day.

The original tacho would have run off the coil primary - 200V or so. If
its the same tacho that used to be fitted to Spridgets and the like, it
is in fact a 0-10mA moving cold meter so you can gut it and build your
own pulse counter.

It's already got one built in. But they're delicate things to start
hacking around. The present PCB is glued in place.

Or you can make a mini ignition coil and use the flyback with a high
voltage switching transistor.

Audio transformers can be pretty small as well, you know.

You dont want an emitter follower either. No voltage gain.

It buffers the ECU output. The transformer provides the voltage gain.

What occurs to me is taking the diode bridge, smoothing it, and using a
standard inverter from the ECU to get a couple of hundred volts pulse
at the collector.

The diodes as suggested work just fine. But require a connection to each
of the four coils. Which will make my loom look a bit untidy. One wire
from the dedicated output on the ECU, not.

Or run it off 12V with an inductor in the collector.

--
*A journey of a thousand sites begins with a single click *

Dave Plowman London SW
To e-mail, change noise into sound.

Jules Richardson wrote:
On Mon, 08 Nov 2010 10:36:48 +0000, Dave Plowman (News) wrote:

The rev counter on my old SD1 was connected directly to the coil LT
negative.

I'm having a right ol' thicky day today, but for my addled brain, why
15V? Why isn't it just "whatever the alternator is putting out", which
will vary a lot between cars and at different loads, but might be
anywhere down to 11V or so and with 15V as a probably upper design limit?


Coil primary starts at +12v and 0v, top end normally held to +12v but
bottom end swings up to about 200-300V when the points open,.

Dave Plowman (News) wrote:
In article ,
The Natural Philosopher wrote:
Yes, if it works on 15v its likely to work on 14v too. But why spend a
bunch of time & error on soemthing that works fine already (the
transformer option), even if it is historic technology.


I think he explained that: to make it small enough to fit in the
dashboard.

I was also hoping to offer it as an option for others with a similar
problem. I had a suitable ratio audio tranny lying around - but they're
quite expensive. A small PCB could be fitted inside the actual rev
counter.

all you need to do is make a 'little' ignition coil and drive it from a
400v capable transistor

wind a few hundred turns on any old lump of ferrite or similar.

In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!

It would if you had a high enough HT rail. Sadly in a car you have to make
things work from about 10 volts.

Why do you say that? With the engine running, and the battery charging,
the battery will be around 14.4V. Even with a bit of R-C decoupling to
keep the alternator rubbish out of the amplifier, you should still be
able to get 12 or 13V pulses. Surely that's enough to trigger the tacho?
--
Ian

In article ,
Jules Richardson wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative.

I'm having a right ol' thicky day today, but for my addled brain, why
15V? Why isn't it just "whatever the alternator is putting out", which
will vary a lot between cars and at different loads, but might be
anywhere down to 11V or so and with 15V as a probably upper design limit?

Or have you fitted something that wasn't originally intended for car use?

The pulse counter in the tacho is connected to the car's 12v supply, but
the trigger feed to that comes from the coil negative. Which produces a
pulse of many volts as the flux collapses. Tests on the tacho shows this
must be a minimum of about 15 volts to trigger the pulse counter. The
output I'd like to use is a diagnostic one and limited to about 5 volts.

--
*A chicken crossing the road is poultry in motion.*

Dave Plowman London SW
To e-mail, change noise into sound.

Ian Jackson wrote:
In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!

It would if you had a high enough HT rail. Sadly in a car you have to
make
things work from about 10 volts.

Why do you say that? With the engine running, and the battery charging,
the battery will be around 14.4V. Even with a bit of R-C decoupling to
keep the alternator rubbish out of the amplifier, you should still be
able to get 12 or 13V pulses. Surely that's enough to trigger the tacho?


cold crank on a flattish battery is a measly 7V or less ;-)

In article ,
The Natural Philosopher wrote:
I was also hoping to offer it as an option for others with a similar
problem. I had a suitable ratio audio tranny lying around - but they're
quite expensive. A small PCB could be fitted inside the actual rev
counter.

all you need to do is make a 'little' ignition coil and drive it from a
400v capable transistor

Think you're missing the point...

The transformer I'm using at the moment works just fine. A single IC would
be even better.

--
*Being healthy is merely the slowest possible rate at which one can die.

Dave Plowman London SW
To e-mail, change noise into sound.

In article ,
Ian Jackson wrote:
In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!

It would if you had a high enough HT rail. Sadly in a car you have to
make things work from about 10 volts.

Why do you say that? With the engine running, and the battery charging,
the battery will be around 14.4V.

I'd like it to work while cranking. As it does at the minute. Would give
an indication if the system wasn't triggering.

Even with a bit of R-C decoupling to
keep the alternator rubbish out of the amplifier, you should still be
able to get 12 or 13V pulses. Surely that's enough to trigger the tacho?

The existing pulse at the coil is many many volts. Nothing to do with the
12 volt supply.

--
*In "Casablanca", Humphrey Bogart never said "Play it again, Sam" *

Dave Plowman London SW
To e-mail, change noise into sound.

Dave Plowman (News) wrote:
In article ,
Jules Richardson wrote:
The rev counter on my old SD1 was connected directly to the coil LT
negative.

I'm having a right ol' thicky day today, but for my addled brain, why
15V? Why isn't it just "whatever the alternator is putting out", which
will vary a lot between cars and at different loads, but might be
anywhere down to 11V or so and with 15V as a probably upper design limit?

Or have you fitted something that wasn't originally intended for car use?

The pulse counter in the tacho is connected to the car's 12v supply, but
the trigger feed to that comes from the coil negative. Which produces a
pulse of many volts as the flux collapses. Tests on the tacho shows this
must be a minimum of about 15 volts to trigger the pulse counter. The
output I'd like to use is a diagnostic one and limited to about 5 volts.

Right,. You need an inductor in the mix somewhere then. It doesn't need
to be MUCH of an inductor. Years ago those little transistor output
transformers..would do. Microphone transformer secondary might do as well.

Blimey. Amazon does audio output transformers for 3 quid.

That's just the ticket.

You can simply drive the primary from a transistor collector, and drive
the base from your 5v pulse via a suitable resistors. Collector should
have a diode to earth, reverse biassed.

Collector output should be a 50v+ spike.

In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!

It would if you had a high enough HT rail. Sadly in a car you have to
make things work from about 10 volts.

Why do you say that? With the engine running, and the battery charging,
the battery will be around 14.4V.

I'd like it to work while cranking. As it does at the minute. Would give
an indication if the system wasn't triggering.

Even with a bit of R-C decoupling to
keep the alternator rubbish out of the amplifier, you should still be
able to get 12 or 13V pulses. Surely that's enough to trigger the tacho?

The existing pulse at the coil is many many volts. Nothing to do with the
12 volt supply.

Noted. If the trigger is from across the points, I expect it's what -
400 to 600V?
--
Ian

Ian Jackson wrote:
In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
In message , "Dave Plowman (News)"
writes
In article ,
Ian Jackson wrote:
I would have thought that a simple two-stage any-old-transistor
amplifier, designed in two minutes flat on the back of a fag packet,
would suffice!

It would if you had a high enough HT rail. Sadly in a car you have to
make things work from about 10 volts.

Why do you say that? With the engine running, and the battery charging,
the battery will be around 14.4V.

I'd like it to work while cranking. As it does at the minute. Would give
an indication if the system wasn't triggering.

Even with a bit of R-C decoupling to
keep the alternator rubbish out of the amplifier, you should still be
able to get 12 or 13V pulses. Surely that's enough to trigger the tacho?

The existing pulse at the coil is many many volts. Nothing to do with the
12 volt supply.

Noted. If the trigger is from across the points, I expect it's what -
400 to 600V?

rarely over 400.

In the early days of transistorised ignition the cap as charged up to
about 250, before thyristor-ing across the coil.

On Nov 8, 7:10*pm, "Dave Plowman (News)" wrote:
In article ,
* *The Natural Philosopher wrote:

I was also hoping to offer it as an option for others with a similar
problem. I had a suitable ratio audio tranny lying around - but they're
quite expensive. *A small PCB could be fitted inside the actual rev
counter.

all you need to do is make a 'little' ignition coil and drive it from a
400v capable transistor

Think you're missing the point...

The transformer I'm using at the moment works just fine. A single IC would
be even better.

Still havent heard the specs, ie what does the unit need on is input,
what impedance, does it just want +15v pulses or what. In principle it
sounds entirely doable with a fet audio amp, or if the impedance is
quite high then even cmos type logic.


NT

In article ,
The Natural Philosopher wrote:
The pulse counter in the tacho is connected to the car's 12v supply,
but the trigger feed to that comes from the coil negative. Which
produces a pulse of many volts as the flux collapses. Tests on the
tacho shows this must be a minimum of about 15 volts to trigger the
pulse counter. The output I'd like to use is a diagnostic one and
limited to about 5 volts.

Right,. You need an inductor in the mix somewhere then. It doesn't need
to be MUCH of an inductor. Years ago those little transistor output
transformers..would do. Microphone transformer secondary might do as
well.

Blimey. Amazon does audio output transformers for 3 quid.

That's just the ticket.

You can simply drive the primary from a transistor collector, and drive
the base from your 5v pulse via a suitable resistors. Collector should
have a diode to earth, reverse biassed.

Collector output should be a 50v+ spike.

My present device does much the same - one transistor a few components and
one transformer. ;-)

--
*I never drink anything stronger than gin before breakfast *

Dave Plowman London SW
To e-mail, change noise into sound.

In article ,
Ian Jackson wrote:
Noted. If the trigger is from across the points, I expect it's what -
400 to 600V?

I've not actually measured it. Worked the other way round and found that
the tacho will trigger reliably with a 15v peak pulse.

--
*Why can't women put on mascara with their mouth closed?

Dave Plowman London SW
To e-mail, change noise into sound.

In article
,
Tabby wrote:
On Nov 8, 7:10 pm, "Dave Plowman (News)" wrote:
In article ,
The Natural Philosopher wrote:

I was also hoping to offer it as an option for others with a
similar problem. I had a suitable ratio audio tranny lying around
- but they're quite expensive. A small PCB could be fitted inside
the actual rev counter.

all you need to do is make a 'little' ignition coil and drive it
from a 400v capable transistor

Think you're missing the point...

The transformer I'm using at the moment works just fine. A single IC
would be even better.

Still havent heard the specs, ie what does the unit need on is input,
what impedance, does it just want +15v pulses or what. In principle it
sounds entirely doable with a fet audio amp, or if the impedance is
quite high then even cmos type logic.


Thought I'd given it earlier.

The output from the ECU is pulses peaking at 5 volts on my scope. The
tacho needs pulses peaking to 15v to trigger reliably.

--
*According to my calculations, the problem doesn't exist.

Dave Plowman London SW
To e-mail, change noise into sound.

On Nov 9, 12:09*am, "Dave Plowman (News)"
wrote:
In article
,
* *Tabby wrote:



On Nov 8, 7:10 pm, "Dave Plowman (News)" wrote:
In article ,
* *The Natural Philosopher wrote:

I was also hoping to offer it as an option for others with a
similar problem. I had a suitable ratio audio tranny lying around
- but they're quite expensive. *A small PCB could be fitted inside
the actual rev counter.

all you need to do is make a 'little' ignition coil and drive it
from a 400v capable transistor

Think you're missing the point...

The transformer I'm using at the moment works just fine. A single IC
would be even better.
Still havent heard the specs, ie what does the unit need on is input,
what impedance, does it just want +15v pulses or what. In principle it
sounds entirely doable with a fet audio amp, or if the impedance is
quite high then even cmos type logic.

Thought I'd given it earlier.

The output from the ECU is pulses peaking at 5 volts on my scope. The
tacho needs pulses peaking to 15v to trigger reliably.

still not giving the necessary info, good luck

In article
,
Tabby wrote:
Still havent heard the specs, ie what does the unit need on is
input, what impedance, does it just want +15v pulses or what. In
principle it sounds entirely doable with a fet audio amp, or if the
impedance is quite high then even cmos type logic.

Thought I'd given it earlier.

The output from the ECU is pulses peaking at 5 volts on my scope. The
tacho needs pulses peaking to 15v to trigger reliably.

still not giving the necessary info, good luck

I'm not quite sure how you get a 15 volt plus output from an 'audio amp'
running off a 12 volt rail?

The tacho input impedance is approx 6M.

--
*If all the world is a stage, where is the audience sitting?

Dave Plowman London SW
To e-mail, change noise into sound.

Dave Plowman (News) wrote:
In article
,
Tabby wrote:
Still havent heard the specs, ie what does the unit need on is
input, what impedance, does it just want +15v pulses or what. In
principle it sounds entirely doable with a fet audio amp, or if the
impedance is quite high then even cmos type logic.
Thought I'd given it earlier.

The output from the ECU is pulses peaking at 5 volts on my scope. The
tacho needs pulses peaking to 15v to trigger reliably.

still not giving the necessary info, good luck

I'm not quite sure how you get a 15 volt plus output from an 'audio amp'
running off a 12 volt rail?


bridge mode, but we cant do that as one size of the tacho needs to be
earthed..

The tacho input impedance is approx 6M.


So as I said, we need to either produce a higher rail voltage with maybe
a capacitative charge pump, or use an inductor.

Charge pump would be good enough at that sort of impedance.

Here's the simplest form.

http://picsru.eu/keyword/charge%20pump/

On Nov 9, 9:53*am, "Dave Plowman (News)" wrote:
In article
,
* *Tabby wrote:

Still havent heard the specs, ie what does the unit need on is
input, what impedance, does it just want +15v pulses or what. In
principle it sounds entirely doable with a fet audio amp, or if the
impedance is quite high then even cmos type logic.

Thought I'd given it earlier.

The output from the ECU is pulses peaking at 5 volts on my scope. The
tacho needs pulses peaking to 15v to trigger reliably.
still not giving the necessary info, good luck

I'm not quite sure how you get a 15 volt plus output from an 'audio amp'
running off a 12 volt rail?

The tacho input impedance is approx 6M.

Now we know its 6M it becomes a lot more doable. The usual way to get
above rail votlage is to use a switched mode power supply, tiny power
ones are very simple to make. 2 transistors and an M3 bolt IIRC. At
those powerlevels you can simply use resistance to protect against
overcurrent.

Or you can use the pulses output themselves to create an above rail
second rail, driving a 2nd output from the first.


NT