Silicon Chip – May 2019

110 Silicon chip Australia’s electronics magazine siliconchip.com.au

online version from siliconchip.com.
au/Shop/12/4078
Yes, the Sensor Modifier can change
the sensor voltage range up or down
according to the map of input versus
output that is entered. So it can shift a
1.2V input signal down to a 0.5V at its
output while leaving a 4.5V level sig-
nal unchanged. The main restriction is
that both the input and output signal
ranges must be between 0V and 5V.

Motor speed controller

with external pot

I recently built your June 1997 High-
Current Speed Controller for 12/24V
Motors from a Jaycar kit (Cat KC5225).
I’m using it to control a car wiper mo-
tor that rotates my telescope dome. The
kit worked fine right off, so thank you
for a good design and article.
Is it possible to replace the trimpot
on the board with an off-board 5kW
potentiometer? I want to be able to
make occasional changes to the rota-
tion speed once the board is housed.
If so, would I be correct in assuming
that it should be a linear pot? Thanks
for a great mag! (N. F., via email)

• Yes, you can use an external po-
  tentiometer and the resistance law re-
  quired is linear.

Multi-spark CDI

transistor failure

Thanks for the fantastic series of
automotive electronics projects! I’m
driving a 1989 Mazda B2000 ute with
points ignition and carburettor which
makes it slow to warm up, and it can
run a bit rich.

So I built the Multi-spark CDI unit

from the December 2014 and Janu-

ary 2015 issues (siliconchip.com.au/

Series/279). It bench tested fine, the

engine started straight away and I then

happily drove the vehicle for about

two months. After driving up a large

hill, the CDI unit completely failed. Q1

and Q2 had blown to pieces.

I am about to rebuild it but I would

like to know why it failed. One clue

is that before it failed, the vehicle bat-

tery became discharged twice. I had to

remove it from the ute and recharge

it. I figured the battery was old and

needed replacing, but I didn’t. Perhaps

the CDI discharged it. I’m guessing the

CDI low-voltage cut-out feature oper-

ated, switching the unit off once the

battery was flat.

So perhaps the CDI unit was draw-

ing too much current from the bat-

tery? I am suspicious of the internal

dead time comparator setting for Q1

and Q2. Could they have been short-

ing the 12V supply when switching?

When I rebuild the circuit, I will check

this with an oscilloscope.

Secondly, I came across a letter in

the June 2015 issue about a CDI failure

due to a short circuit between turns on

the transformer secondary. This would

cause the circuit to struggle to reach

300V; thus, the pulse width driving Q1

and Q2 would increase. Perhaps this

could lead to overheating?

When I get the unit running again,

what should I expect its current draw

to be? I could temporarily connect the

ballast resistor back in series with the

12V supply and measure the voltage

across it.

Regarding the coil, with the new CDI

unit installed, I kept using the same coil

that was used with the ballasted points

system. It has a primary resistance of

about 1.6W. I noticed that in the first

article, the Bosch GT40 coil is men-

tioned. These are designed for use in

ignitions without a ballast resistor and

have a primary resistance of about 3.6W.

So by using the lower impedance

coil, could I have been over-driving

the poor CDI unit until Q1 and Q2

fried themselves? That might explain

why the unit worked for a while be-

fore failing (although I reckon it should

have worked better than it did). (B. N.,

Dunedin, NZ)

• The complete failure of Q1 and Q2
  suggests that the unit overheated until
  meltdown. Possibly it is a shorted turn
  in the transformer, as you suggest. Or
  as you say, it could be that the drive to
  the Mosfets does not have the correct
  dead time. Check the gate drives to the
  Mosfets to make sure these are correct
  before re-connecting transformer T1.
  It should not matter what ignition
  coil is used, whether it is a ballasted
  points type or the Bosch GT40. The
  Multi-spark CDI should draw no more
  than 1A at low engine revs.

Muting radio during

comms activity

I have a boat with two radio systems,

entertainment and marine VHF. I want

to use Bluetooth to connect them to a

pair of headphones for me to monitor

while at the helm, particularly when

I am alone. The Bluetooth bit is OK

and I am considering using your June

2015 Champion preamp.

I built the GPS-synced Frequency

Reference (October-November 2018;

siliconchip.com.au/Series/326), but

it isn’t working properly. I have been

trying to fix it for a few weeks. All in-

dicators in the Status screen are fine

but the output frequencies are way

off the mark. I’m checking them on

my HP 5335A counter which uses

an external GPS-locked 10MHz ref-

erence.

After the unit has been run-

ning with a GPS signal for over

24 hours, the 10MHz output reads

10,000,064.89Hz and the VCO Trim

C setting is at 0. Gain is at 1000 and

Update is 3600. It seems I need to be

able to enter a negative figure for the

C value to force the 40MHz oscilla-

tor to run at the correct frequency,

but there is no way to enter a nega-

tive value.

I am wondering if the TXEAAC-

SANF oscillator is faulty. Are you

able to supply a replacement? Do

you have any ideas as to what the

problem could be? (M. T., Balgal

Beach, Qld)

• The VCO Trim C Value cannot
  be a negative number. Zero is the

GPS Frequency Reference not calibrating properly

lowest it will go. It sounds like ei-

ther the TXEAACSANF oscillator

is faulty, as you suggest, or its trim

input is not being driven by the

DAC properly.

With a C Value of 0, you should

be able to measure close to 0V at the

output of the DAC (pin 6 of IC3).

This should ramp up to 2.5V with a

C Value of 16,000,000.

If this is not what you are see-

ing, there may be a problem with

DAC IC1, op amp IC3 or one of the

signal lines between the Micromite

and IC1.

Continued on page 112