Crest Audio PULSE SERIES Schematic

T E C H N I C A L S P E C I F I C A T I O N S

Pulse Series

Power Ratings

Cooler, Lighter, Stronger

Pulse 4x300

Into 8 Ohms 170W

Into 4 Ohms 300W

Into 2 Ohms 330W

Pulse 2x650

Into 8 Ohms 400W

Into 4 Ohms 650W

Into 2 Ohms 850W

Pulse 2x1100

Into 8 Ohms 700W

Into 4 Ohms 1100W

Into 2 Ohms 1500W

The Pulse Series com-

bines state-of-the-art

switched mode power

supplies to not only

reduce amplifier weight

by as much as 70%

compared to conventional

amps, but also to provide

solid, consistent

performance at all power

levels.

Technical Specifications

Power Ratings Pulse 4x300 Pulse 2x650 Pulse 2x1100

Measured per channel, both channels driven at 1kHz to no more than 0.1% THD+N

8 Ohms 170Wrms 400Wrms 700Wrms
4 Ohms 300Wrms 650Wrms 1100Wrms
2 Ohms 330Wrms* 850Wrms* 1500Wrms*

* Note: 2 Ohm spec is at 1% THD

Bridged Mono

16 Ohms 400Wrms 800Wrms 1200Wrms
8 Ohms 600Wrms 1300Wrms 2200Wrms
4 Ohms 660Wrms* 1700Wrms* 3000Wrms*

* Note: 4 Ohm bridged spec is at 1% THD

Input Sensitivity +1dBu for full output

Input Impedance 20kOhm

Benefits of the

Pulse Series

• Very light weight

• Switched mode

power supplies give
solid performance
at all power levels

• Microprocessor
protection system

• Massive heatsinks
for cooler operation
and higher
reliability

• Binding Post or
Speakon output
connector options

• Optional Remote
control via C Audio
CONNECT

• Internal crossover
card options

Occupying just 2U of rack

space, Pulse uses

massive heatsinks and

front-venting fans to

keep the electronics

really cool, plus a built-in

microprocessor which

continually monitors all

the protection aspects of

Pulse - these factors all

dramatically enhance

reliability.

Pulse amplifiers will

perform for longer

periods than conventional

amplifiers at high output

levels.

The combination of the

switched mode PSU and a

rugged steel chassis

means inherent strength.

Distortion <0.006% THD, 1kHz, 1dB below clip,

Frequency Resp. 20Hz to 20kHz, +0/-0.2dB;

Controls Power switch, bridge mode switching,

Indicators Mains present, Operate, Signal, Bridge,

Protection Microprocessor supervised:

Noise <-100dB ref full output 20Hz - 20kHz

Slew Rate >50 V/microsecond

Damping Factor >200 ref 8 Ohm

Output Connectors 4x300 - Binding post or Speakon

Power 115 or 230 volts AC nominal, internally

Dimensions 3.5" (89mm) x 18.2" (460mm) x 19" (483mm)

- with rear rack ears depth 19.5”, -21”, 20” (494, 511, 530mm)
Weight 24lbs (11kg)

Trade Descriptions Act: C Audio have a policy of continued product improvement and accordingly
reserve the right to change features and specifications without prior notice.

22kHz measurement bandwidth

<2Hz to >120kHz +0/-3dB

indented level controls (these may
be made tamper-proof)

Clip, Overtemperature, Protect, Remote

overtemperature, DC on outputs, output
stage overload, inrush current surge,
mains fail and brownout.

measurement bandwidth

2x650, 2x1100 - Binding post and Speakon

selectable, 2000VA, all channels
driven (4x300, 2x650); 3000VA (2x1100)

March 2000

Pulse P2X1100 circuit description

1. switching power supply

2. power amplifiers

3. protection

4. frontpanel & flashing lights

5. PIC and system protection

1. Switching Power Supply

Mains voltage is inserted via filtered IEC CN3. Earth is connected directly to chassis from the
body of CN3. Live passes through fuse F2.
Live and Neutral pass though a further filter consisting of C2 (X2 rated), L1 - a common mode
choke - and C15 & 14 (Y2 rated).
Live passes through the soft-start system, TH1, R14 and RLY1. R14 limits the initial current
surge caused by the primary reservoir capacitors charging up. TH1 protects R14 if a fault
causes excessive dissipation in R14. RLY1 shorts out TH1 and R14 when the PSU is running.

Live and Neutral then pass to bridge rectifier BR2 which, for the 230V setting full-wave
rectifies mains, smoothing performed by reservoir capacitors C22,23,44,45,32,33. In the 115V
setting, this power supply is configured as a voltage doubler. Thus, the High Tension (+HT)
DC generated is approximately equal for 230V mains and 115V mains.

This will result in about 320Vdc between ‘LIVE GND’ and +HT. ‘LIVE GND’ is named as such
because it is not isolated from mains but it is the reference point for the PSU. If you need to
stick a scope probe around the primary side with the unit plugged in you must connect mains
via an Isolation Transformer. Without this, at best you will only trip your RCD breaker, at
worst you or your scope may not live to regret it. Do not forget that 320Vdc is still pretty
shocking whether it is isolated or not.

R36 and R37 ensure proper voltage sharing of the reservoir capacitors.

Power for the switching controller circuit is provided from two sources.
At start-up, the power comes from R93,113, ZD2, D22 and C75. C75 is charged up to about
47V through D22 from Zener regulator R93,113 and ZD2. The command to start the PSU
comes from the PIC (+5V for off, 0V to switch on) via R33. The LED in OPT1 is turned on
which turns the transistor on, shorting out pins 5&4. While the transformer windings are cooler
than 120
the top of C75 and will form a Zener regulator with ZD9. The output of this zener regulator is
buffered by TR49 which then powers the switching controller circuit.
C75 stores enough energy to run the controller for long enough until the second source of
power is ready.
The second source of power comes from the transformer on pins 3 and 14. This secondary is
voltage doubled by C82, D29,31 and C74 and produces about 48Vdc on VAUX. This is
connected via D32 to the top of C75 so the controller circuit can continue running.
VAUX is used to directly power RLY1 to short TH1 and R14 out while the PSU is running.
The circuit comprising R165,R172,ZD10,R178,TR47 and R180 shut down the PSU when the
voltage on C75 falls below about 32V. while the voltage on C75 is greater than 32V, TR47 is
switched on and pulls pin10 of IC6 down to 0V which ‘enables’ IC6. When the voltage on C75
falls below 32V, TR47 switches off and pin10 of IC6 is pulled up to 20V through R180,
‘disabling’ IC6.
R92,8,85,C67,ZD3 and TR23 form a time delayed drive for RYL2. At start-up, RLY2 will be
open and R127,145 will be in series with the primary of the transformer. These resistors limit
the current surge caused by charging up the secondary reservoir capacitors
C113,114,137,138. About 100ms after a successful start-up, TR23 will energise RLY2,
shorting R127,145.

0

C the thermal cut-out (TX5-B) will be a short circuit. So R166 will be connected to

The switching controller circuit is based around IC6, and SG3525 PWM controller. Switching
frequency is set by R175 and C199 to about 85kHz. R176 controls the ‘dead time’ period,
setting it to between 500ns and 1µs. Outputs appear at pins11 and 14. These two outputs are
complementary, that is when one is high (20V), the other is low (0V). Due to the dead time
control, neither outputs are high at the same time but both are low for the ‘dead time’ period.
These outputs are fed through R169,170 to high current buffers TR40,41 and TR42,43. The
buffered outputs push-pull drive the primary of Pulse transformer TX4. R160 is a damping
resistor used to minimise ringing caused by imperfections in TX4. The transformer has two
secondaries, each driving one IGBT in anti-phase. The turns ratio is 1.5:1 (pri:sec) and due to
the push-pull connection of the primary, the output of TX4 swings positive to about 15V to turn

the IGBT on, falls to 0V switching the IGBT off during dead time and falls to about -15V whilst
the other IGBT is switched on.

IGBTs TR24,29 form a half bridge driver for TX5. D14,23 provide ‘flyback’ protection for
TR24,29. These are not fitted as we are currently using IGBTs with integral ‘flyback’ diodes.
Snubbers R84,C56,R117,C68 damp any ringing which may occur.
The drive for the transformer from the IGBTs is an 85kHz square wave almost hitting ‘+HT’ at
the top of it’s travel and bottoming out slightly above ‘LIVE GND’. This drive is connected to
the primary of TX5 at pins 15&16. The other end of the primary at pins 1&2 connects to ‘LIVE
GND’ through the closed RLY2 and C63,64,65.
C63,64,65 perform two important functions.

1. they prevent DC current flowing through TX5 thus preventing early saturation of the core.

2. they form a discontinuous resonant circuit with the leakage and stray inductance of the

transformer. This means that the current will not be the same shape as the voltage. It is
arranged that, for instance, positive current flow through TR24 will start from 0A, rise and
fall in a half-sinusoidal fashion, reach 0A and stop before TR24 is switched off. This means
that switching losses in the IGBTs are virtually eliminated because switching occurs whilst
0A of current is flowing.

There are two main secondaries.

1. Pins 10,11,4&5 are the low voltage centre tapped secondary winding. Output of this is full­wave rectified by D49,50,54,55 and smoothed by C126,C132. Further filtering is provided
by L2,3 and C120,130 before passing to 15V regulator IC7 and -15V regulator IC8. The
+15V and -15V outputs are ‘decoupled’ by C124,133 close to the regulators to ensure
stability. C28 and C29 provide further ‘decoupling’ close to the option connectors CN11,12.
D51 half-wave rectifies the secondary output. This is lightly smoothed by C117 and loaded
by R219. This is passed via ZD15,R34,R35 and C25 to the PIC. While the PSU is running,
this circuit will produce a digital ‘high’ at the PIC input. If the PSU stops for any reason, the
small value of C117 means that the PIC will receive a digital ‘low’ well before any of the
Power supplies have drooped significantly allowing the PIC time to prevent plops.

2. Pins … is the main Power centre tapped stacked secondary. This is full-wave rectified by
C45,48,53,61,57,60,36,47 and smoothed by C113,137,138,114 to produce HT+, HT- and
MT+, MT- so that MT+ is half HT+ and MT- is half HT-. R7,112 roughly equalise the
discharge rate when the PSU is stopped.

2. Amplifier

Refer to channel 1

The amplifier consists of a ‘Class A’ driver and a ‘Class AB’ rail switched power output stage.
The driver provides voltage gain only, the output stage provides current gain only.
Audio signal enters the amplifier through DC blocking capacitor C149, then low-pass filter R88
& C51 and onto the base of TR30. R261 provides a low source impedance in case the
frontpanel board is disconnected. R89 provides a DC path to ground for the base current of
TR30.
The output of the amplifier is fed-back through potential divider R104 and R102 to the base of
TR31. C58 decouples the feedback signal at DC so that DC offsets generated by TR30,31 are
not amplified at the output. D15,19 protect C58 in the event of a DC fault.
TR30 and TR31 form a Long Tailed Pair to amplify the difference between the input signal and
the feedback signal. The gain of the LTP is reduced by R99 and R100 to help prevent
oscillations and de-sensitise the performance of the input stage to parametric variations of the
two transistors. A bias current of about 2.8mA for this LTP is provided through R98 from
current source TR19, R57, D6,7 & R75. In the quiescent state half of this current is driven
through each of TR30 and TR31. The collector current of TR30 and TR31 pass through ZD7
and ZD5 and are loaded through D27 and D26 by R137 and R146.
The outputs of TR30 and TR31 are taken from the anodes of D27 and D26 to the bases of
another LTP - TR38 and TR37. C70 determines the frequency response to ensure stability. As
before, R139 and R140 reduce the gain of this LTP and the bias current is set to about 8mA
by R138. The collectors of TR37 and TR38 are loaded with a current mirror - TR20 & TR21 to
maximise gain and provide a push-pull output.
Some of this output is fed-back to the base of TR31 through C63 and R103. This defines the
open-loop frequency response independently of the output stage characteristics to ensure
stability.
Finally, the Vbe multiplier - TR25, R82&83 - provides the output stage with two voltage signals
which are identical except they are offset by a voltage varying between about 2.1V (heatsink
hot) and 2.4V (heatsink cold). C53 ensures that the two offset signals are identical at AC.

The Current source consisting of TR19, R57, D6,7 and R75 determines the operating point of
the whole class A driver. Therefore, one can mute the amplifier by switching this currrent
source off.
The current source is switched off by TR18, R68 and C37. When TR18 is switched on, D6,7
are shorted out through R68 which mutes the current source. C37 is discharged in the
process. When TR18 is switched off, C37 charges up through R75 until D6,7 are fully
conducting which activates the current source.
TR18 is controlled by TR11, R56, R46, ZD1, R47 and R48.
If the PIC is absent or its +5V supply has failed the ‘MUTE1’ line will be in a high impedance
state, i.e. nothing doing. The 4.7V reference supplied by R46 & ZD1 and emitter resistor R47
set the current through TR11 to about 200µA. This is enough to switch TR18 on and mute the
amplifier. This is the default state.
When working, the PIC controls the state of the ‘MUTE1’ line. To mute the amplifier, the
‘MUTE1’ line is set to 0V. The current through TR11 is then set to about 4.2mA because R48
is now essentially in parallel with R47. This mutes the amplifier as before.
To activate the amplifier, the ‘MUTE1’ line is set to +5V. This, through R48, reverse biases the
base-emitter junction of TR11. Thus TR11 is switched off as is TR18 so the amplifier
becomes active.

Under normal conditions the signals at the bases of TR30 and TR31 will be the same.
However, under fault conditions, such as a DC offset at the output, the base voltages will
become offset also. For example, in the event of a large DC offset of +50V at the output, a
positive DC voltage will appear at the feedback point and hence at the base of TR31. This DC
voltage will make D65 conduct protecting C126, so the voltage at TR31 base should be 0.6 +
50V x R102/(R102+R104), about 4V. However, the important issue is that the voltage is
positive. In the event the voltage is negative this indicates that the feedback divider is faulty.
The voltage at TR31 base being positive whilst the base of TR30 is close to 0V will then
reverse bias TR31 base-emitter, turning off the transistor. Therefore, no voltage should

appear across R100 and R146 whilst double the normal voltage will appear across R99 and
R137. Should this not be the case, it indicates a fault in the input stage itself.
The same process should now cause TR37 and TR20 &21 to be off and TR81 to be full on…

Class AB output stage

The input of the Output stage is loaded by C46. This defines the HF input impedance and thus
averts very nasty oscillations that are caused by the variable, non-linear and sometimes
negative raw input impedance. Resistors R60 and R182 ensure that output offsets are
minimised when the amplifier is muted. D5 and D39 stop the Class A driver over-saturating
TR12 and TR55. D11 and D40 prevent the output exceeding the power supply rails in the face
of ‘flyback’ pulses from reactive loads.

The output stage consists of a symmetrical Siklai follower - TR12, TR13, R61, TR55, TR56,
R181, R81, R80, C52, D13, D17 - generating the high current drive required for the parallel
connected symmetrical follower output stage - TR26, TR14, TR34, TR46, R115, R94, R121,
R124, TR72, TR59, TR88, TR79, R147, R141, R168, R159. V-I limiting is controlled by D8,
D33, TR22, TR44, R96, R95, R116, R119, C60,R129, R149, R164, R163, C81, R64, C48,
R128, C77, D10, D25, R212, R214, TR71, R217, R228, TR75.

The Pulse 2x1100 amplifier uses a rail switched (class G or commutated) output stage to
increase efficiency. This works by effectively rolling two amplifiers into one, switching between
them as required.

While Vo is between +MT and -MT, SW1 and SW2 are open and power is drawn from +MT
and -MT through D1 and D2.
If Vo gets too close to +MT then comparator1 closes SW1. Power is now drawn directly from
+HT and D1 is reverse-biased.
Similarly, if Vo gets too close to -MT then comparator2 closes SW2. Power is now drawn
directly from -HT and D2 is reverse-biased.

1. Positive half-cycle

The output voltage is sensed at the top of current sharing resistor R124 to correct for errors
caused by the varying voltage drop across current sharing resistors. This is divided down by
R221 and R234 and enters pin6 of IC10-B which is used as a comparator. C123 compensates
for time delay through the circuit. +MT is divided down by R229 and R230 and enters pin5 of
IC10-B. R237 provides hysteresis around the comparator.
Whilst pin6 is lower than pin5, the output of the comparator is at +13.5V on pin7. This
switches TR67 on with its collector current set by R216 to about 2.5mA.
When pin6 exceeds pin5, the output of the comparator goes to -13.5 on pin7. This switches
TR67 off.

Rail switching is performed by FET1 and FET3 with commutation diode D52. FET1 and FET3
are connected in parallel except for the gates which have separate ‘gate stopper’ resistors
R192, R195. A floating 10V supply for the drive circuit is provided by R199, ZD13 and C112.
The drive circuit consists of push-pull buffer TR68,69, pull-up resistor R201 and C110.
When TR67 is on, 2.5mA flows through R201, dropping 12.75V. therefore, TR69 will be off
and TR68 will be on, pulling FET1 and FET3 gates low which switches them off. Power for the
amplifier thus flows from +MT through D52.
When TR67 is off, R201 pulls the bases of TR68,69 up, which pulls the gates of FET1 and
FET3 up and switches them on. Power for the amplifier now flows from +HT through FET1
and FET3. C110 slows the rise and fall times of the +overdrive rail to about 1µs.

2. Negative half-cycle

The output voltage is sensed at the bottom of current sharing resistor R159 to correct for
errors caused by the varying voltage drop across current sharing resistors. This is divided
down by R207 and R248 and enters pin2 of IC10-A. C111 compensates for time delay
through the circuit. -MT is divided down by R259 and R256 and enters pin3 of IC10-A. R246
provides hysteresis around the comparator.
Whilst pin2 is higher than pin3, the output of the comparator is at -13.5V on pin1. This
switches TR74 on with its collector current set by R225 to about 2.5mA.
When pin2 exceeds pin3, the output of the comparator goes to +13.5 on pin1. This switches
TR74 off.

Rail switching is performed by FET5 and FET6 with commutation diode D62. A 10V supply
referenced to -HT for the drive circuit is provided by R239, ZD20 and C139.
The drive circuit consists of push-pull buffer TR81,82 pull-up resistor R251, C145 and current
mirror TR80,D63,R258,R250.
When TR74 is on, the current mirror reflects 2.5mA through TR80 which pulls down on R251.
Therefore, TR82 will be off and TR81 will be on, pulling FET5 and FET6 gates low which
switches them off. Power for the amplifier thus flows from -MT through D62.
When TR74 is off, the current mirror reflects this by turning TR80 of. R201 pulls the bases of
TR81,82 up, which pulls the gates of FET1 and FET3 up and switches them on. Power for the
amplifier now flows from -HT through FET5 and FET6. C145 slows the rise and fall times of
the -overdrive rail to about 1µs.

3. Protection

Output stage

Output stage protection is accomplished by a three-slope V-I limiting circuit which has limiting
characteristics chosen to emulate the Safe operating area of the output stage transistors at
their maximum operating temperature.
As the output stage is symmetrical, the positive half only will be described.
The V-I limiting works by controlling TR22: when the base-emitter voltage of TR22 exceeds
about 0.65V then TR2 turns on and steals current, via D8, from the input of the output stage
and thereby limiting the output. So, V-I limiting is controlled by controlling the base-emitter
voltage of TR22.
Each output device has its own current sharing resistor - R115 R94, R121, R124 - the voltage
across which is proportional to the current flowing in the output device. These voltages are
sampled and summed by R96, R95, R116 and R119. C60 improves stability when V-I limiting
is activated.
Thus the amplifier is protected for short circuits because the base-emitter voltage of TR22
reaches 0.65 if the output current is large and the output voltage is less than about 1Vpk
For output voltages exceeding about 1Vpk, D10 conducts connecting R212, R214 & TR71 to
sense the output voltage. In this case, as output voltage increases, the base-emitter voltage of
TR22 reduces, thus the current limit is increased as the output voltage increases, defining the

nd

2

slope of the limiting characteristic.
The third slope is controlled by TR71, R214 and the rail switching comparator. If the
+overdrive rail is low, then TR71 is switched on and shorts R214, leaving R212 to define the

nd

2

slope. When the +overdrive is high, TR71 is switched off and the 3rd slope is defined by
R214 in series with R212.
C48 and R64 desensitise the current allowing brief peaks of output current which exceed the
normal current limited amount to allow proper operation into reactive loads (ie loudspeakers).

Excessive continuous power demand

Continuous maximum sine-wave operation of both channels into the minimum rated load
resistor is beyond the capabilities of this product and is treated as a fault condition.
Refer to channel1
Amplifier output current is measured by R159, 124, 183, 191, 177, 179 and IC5-A.
Output current causes a voltage drop across current sharing resistors R124 and R159. If the
voltage across both resistors is measured, positive output current will cause the voltage to
increase and negative output current will also cause the voltage to increase. This is sensed
and attenuated by differential amplifier R183, 191, 177, 179 and IC5-A.
The output at pin1 of IC5-A is the sum of: 1. full-wave rectified replica of the output current; 2.
An error signal caused by incomplete common-mode rejection. The error signal is merely
attenuated output voltage.
As the desired signal, output current, has been full wave rectified and the undesired signal,
output voltage, has not, the undesired signal can be removed without corrupting the desired
signal by passing it through an integrator. The raw output current signals from each channel
are summed, averaged and inverted by R174, 52, 53, C43 and IC4-A (sheet 4) so that the
output at pin1 of IC4-A is a negative DC voltage proportional to the total average output
current.
This voltage is compared with a threshold, set by R50 and R51 to about

-2.7V, at IC4-B. R55 provides hysyteresis. If it is higher than the threshold, pin7 will swing to -

13.5V. If it is lower than the threshold then pin7 will swing to +13.5V.
IC4-B pin7 drives transistors TR66, 64, 65 and 63. The collectors of these transistors connect
to the collectors of TR67, 86, 74& 62. When pin7 swings to +13.5V, TR66, 64, 65& 63 are
switched on, the current flow set to 2.5mA by R198 and R197. This forces rail switching off.
IC4-B pin7 is connected to the PIC pin22 through R49.

Bridge imbalance protection.

During normal operation, the bridged output is fully differential mode with little or no common­mode signal component. Activation of the output stage current limiters erratically upsets this
state, producing a large common-mode error - imbalance - which can destroy the output
stages.
The bridge imbalance detection is performed by R10,12,13, TR3 and R11.
In two channel mode, the ‘BRIDGE 1+2’ line is pulled up to about +13V. this switches TR3 on
through R11. The collector of TR3 is connected to pin 23 of IC1 (PIC). The input of the PIC is
internally protected by diodes against inputs above +5V and below 0V. so, in two channel
mode, pin23 of IC1 receives no signal.
In bridge mode, the ‘BRIDGE 1+2’ line is held at 0V which switches TR3 off. Now pin 23 of IC1
can receive a signal. R10, 12 sum the output of each channel, the result appearing across
R13 which sets the sensitivity. If the bridge is balanced, the voltage across R13 will be zero. If
the bridge is unbalanced then there will be voltage across R13 which is also connected to
pin23 of IC1.
When an imbalance is detected the PIC immediately mutes both channels for 4 seconds.
Then the channels are re-activated.

DC Protection

Each channels output is connected through R16,R15 to R21 and C16. C16 ensures that only
DC is detected.
For no DC fault, R32 pulls pin9 of IC1 towards +5V.
For a positive DC faults, D4 becomes forward biased and turns TR8 on, pulling pin9 of IC1
(PIC) towards 0V.
For a negative DC fault, D3 becomes forward biased which switches TR10 and TR9 on,
pulling pin9 of IC1 (PIC) towards 0V.
In the event of a DC fault, the PIC switches the PSU off, waits and then tries again.
C30,31,35,36 prevent the full-wave recification of PSU noise causing false alarms.