Silicon Chip – June 2019

siliconchip.com.au Australia’s electronics magazine June 2019 109

I built both versions of the Sena-

tor Loudspeakers – the fully-fledged

set using the Celestion drivers (Sep-

tember & October 2015; siliconchip.

com.au/Series/291) and the “budg-

et” set using the Altronics drivers

(May & June 2016; siliconchip.com.

au/Series/300).

I have used Jaycar-sourced induc-

tors in the crossovers for both, after

scouring the branches across Aus-

tralia to get them.

I notice that the budget set sounds

a little better than the original ver-

sion with regards to the bass repro-

duction.

I have mounted the crossovers on

the reflector plate behind the woof-

er in both versions and was won-

dering if this is correct, or should I

have mounted them on the floor of

the cabinets?

Both have the specified acoustic

wadding but I noticed that in both in-

stances, it covers the bass reflex port.

Is this a problem? I have tried both

sets of speakers being with the same

Class-A amplifier and the same CD.

The cheaper set definitely has

better bass reproduction and I’m

wondering why. I love the maga-

zine; keep up the good work. (P. C.,

Woodcroft, SA)

• We asked Allan about this and
  he responded: it is always pleasing
  to get feedback about my speaker
  projects!
  In regards to the bass response,
  the Altronics C3065 driver does
  have a slightly lower resonance
  because of its softer compliance,
  but it also has a much lower pow-
  er handling ability and sensitiv-
  ity (60W/92dB) compared to the
  Celestion NTR10252OD/E drivers
  (250W/96dB).
  The original Senator design is
  designed to handle 250W/channel.
  They can fill an auditorium and han-
  dle the heat because of the dual 2.5-
  inch voice coils used in the woofers.
  But if your room is small, then the

Budget Senator Loudspeakers have more bass

budget speakers with the Altronics

drivers will be fine.

I listen to the Celestion-based ver-

sions nearly every day for watching

TV and movies and am amazed at

their smooth response and big dy-

namics.

I never tire of the sound where-

as I find the “budget” version a bit

boomy in the bass. We all have dif-

ferent ears, so it becomes a personal

choice which is better if high sound

levels are not needed.

Your crossover mounting position

is fine, and the wadding can be kept

away from the port by rolling it up

and loosely tying it up with a bit of

thin insulated wire. But we don’t

think it will make too much differ-

ence either way.

Just check that all your cabinet

joints are airtight because this can af-

fect the bass response and you need

to ‘run in’ the Celestions to get the

best sound because of their tighter

surrounds.

wire link. Change the 16V supply rail
bypass capacitors to 25V types. Con-
nect the junction of your two batteries
to the Vcc ÷ 2 split rail and then pow-
er the circuit from the 18V across the
two batteries.
The second option is the Balanced
Microphone Preamplifier & Line Mixer
in the April 1995 issue. There is Al-
tronics kit for this project still avail-
able (Cat K5531).
It’s designed to run from ±12V, but
it would work OK at ±9V. You would
need to bypass the 12V regulators.
Connect the junction of the two 9V
batteries to the ground rail.
You can purchase a scan of that ar-
ticle at the following link: siliconchip.
com.au/Shop/?article=5163 Note that
the low-frequency response of this
design can be improved from -3dB at
180Hz to -3dB at around 34Hz by in-
creasing C7 from 100nF to 470nF.

Modifying amp power

supply for lower voltage

I am building a kit for the 50W Au-
dio Amplifier Module from the March
1994 issue of Silicon chip (Jaycar Cat
KC5150). I want to use it to turn a

passive subwoofer into an active sub-

woofer. The speaker impedance is 4W,

and the instructions say to swap the

LM3876T chip with an LM3886T chip

and reduce the power supply rails

from ±35V to ±28V.

Unfortunately, it doesn’t say how to

do that. I’ve been scouring the internet

for help and can’t find anything rel-

evant. I’ve sourced an LM3886T but

don’t know how to modify the power

supply circuit to get the required volt-

age. I need to know if it requires a dif-

ferent transformer and/or rectifier and

the value of the required capacitors.

(M. H., via email)

• For ±28V (nominal) supply rails you
  need a transformer with two 18V wind-
  ings (18-0-18V or 36V centre-tapped).
  This will give you about ±25V at full
  load and about ±28V at light loads. The
  Altronics Cat M5118C 80VA toroidal
  transformer is suitable. You don’t need
  to change any other components in the
  power supply.

Using Insulation Meter

to test long cables

I purchased the June 2010 issue and

the Power Supply PCB you designed

to build a Digital Insulation Meter

(siliconchip.com.au/Article/186). I

built it and tested it on resistances

from 500kW to 500MW at 500V, and it

worked fine.

I then tested the meter on the same

resistances but using 30m cables to

connect the resistances to the meter.

It worked OK below 10MW, but start-

ed giving incorrect results above that.

For example, the reading was 190MW

with a 100MW resistor at the end of a

30m cable.

I measured the current inside the

cables, and it varied a lot more with

long cables than with the short ones.

I think this is due to the capacitance

of the longer cables affecting the unit’s

operation.

I want to make insulation meas-

urements on a cable a few kilometres

long. Do you have any advice on how

I can achieve this? (M. deR., Toulouse,

France)

• The June 2010 Insulation Meter was
  not designed to check long cables. We
  think you’re right that it’s the cable ca-
  pacitance that is causing the problems.
  Overcoming this might be tricky.
  We suggest that you try fitting an
  additional LC filter between the out-