Burwen TNE7000 Transient Noise Eliminator
Service Manual

Jay's Note:

The TNE7000 appeared in the early 1980s, and was the first (and in my opinion best) unit to remove ticks from phono playbacks. We put one in every studio in our facility. This copy of the service manual is posted both for public education and to help those who have one of these units. It's (c) 1980 by the inventor, Dick Burwen, and was graciously supplied by him. It was also hand-typed into html, so please forgive any typos.

Dick does not want to be bothered with questions about this device. If you can't find an answer here, you're on your own.

Other Burwen documents



The principles of operation, a block diagram, and oscilloscope waveforms are described in the accompanying Audio Engineering Society preprint "Suppression of Low Level Impulse Noise" by Richard S. Burwen, preprint no. 1388. For details of the operation refer to the 4chematic diagram in which the upper section of components numbered in the I00s is the left audio channel, the center section numbered in the 200s is the right audio channel, and the lower section numbered in the 300s is the detector circuit.


In the left channel amplifier A101A is used as a bridge amplifier to provide a gain of 1 and a low output impedance in the defeat mode. Amplifier A501A is not included on the circuit board but space has been provided for specialized applications requiring a 600 ohm load capability. Amplifier A101B, in conjunction with C103, R108, and R107, preemphasizes high frequencies above 1 kHz. Following AIOIB the signal passes through three 18 dB/octave low-pass filters AIOIC. A101B, and A102A. Each section is flat within approximately 0.05 dB out to 20 kHz and R117 Is used for fine-adjustment for flatness at 20 kHz. Next the signal is split Into two paths. In the lower path Q101 and C116 act as a sample-hold circuit having a 500 Hz low-pass action. Further low-pass filtering is provided by A103B whose feedback network consists of C117 and R143. At the output of TP101, the frequency response is down 6 dB at 500 Hz and the phase shift is 90 deg. In the upper channel, A102B serves as a 180 deg phase shifter centered at 500 Hz Hz and as a second 1800 phase shifter centered at 20 kHz. The components determining the 500 Hz phase shift are C115 and R121 and the 20 kHz components are R120 and C114. The purpose of the 90 deg phase shift at 500 Hz is to match the phase shift at TP101. The 20 kHz phase shifter is used to provide a preshoot on the delayed signal to assure that the step response is up no more than 10% at 40 us. This 40 us time delay produced by the low-pass filter and phase shifter allows the detector time to operate and remove the leading edge of any tick.

Following A102B is a single pole double throw electronic switch Q102. Normally Q102 acts as a short-circuit (approximately 50 ohms) and allows the signal to pass through Q102C acting as a follower. During a tick, diode D102 cuts off Q102 and the signal path to Q102C comes from TPIOI via R129. Low distortion in normal operation is achieved by maintaining the gate of Q102 midway between the source and drain via restators R128 and R131. The frequency response and gain effects of the small series resistance are compensated by providing a 3% loss in the feedback path by means of R130 and R142. These components, together with R129 and the series resistance of Q102 make A102C a bridge amplifier which, in normal use, accurately amplifies with a gain of 1.0 the signal at the output of A102B. The final output amplifier A102D contains a complementary deemphasis network in its feedback path consisting primarily of C120, R137, and R138.


In the detection circuit, A301A is a bridge amplifier which takes the difference between the outputs of A101B and A201B, thus the system responds to the vertical component of the phono signal. Following A301A the sensitivity control R306 is used to set the level of the incoming signal to a standard value as indicated on the CAL LED. High frequencies are further amplified by A301B and A302A producting the gain and frequency response at the typical setting of the snsitivity control as shown in the AES preprint. Having selected out the high frequency component of the vertical component of the program source, the signal at TP302 is then full-wave rectified at A302B and A303A. The signal then splits into two paths which feed the compartor A304A. In the lower path the signal passes directly from pin 12 of A303A through a 3 dB divider R35 and R36 which is biased approximately -80mV by R338. The upper path averages the signal, delays it, and amplifies it and full-wave peak rectifies it to provide inhibiting action for the comparator A304A during normal signals. Amplifier A303B is an 18dB/oct low-pass filter having a gain of 2 at dc. The output at TP304 is sampled by the comparator A303D which is biased at +10mV threshoold. The output of A303B is also amplified by A303C which contains the threshold control in its feedback path. A fourth stage of low-pass filtering is used ahead of A303C so that the total time delay on the signal is appoximately 1.5 ms. The threshold control determines the average value of signal which is fed to the comparator and is neceswsary for the instantaneous value to covercome in order to be considered a tick. Diode D303 assures that the output of A303C cannot go negative due to accumulated offsets in the prior amplifiers. This diode also provides peak rectifying action by charging C310 through R330. The charging time constant of this circuit is 330 us while th edischarge time constant is 22 ms. The purpose of this circuit is to reduce the sensitivity of the system to regularly occurrting impulses at low frequencies such as the leading edges of square waves which might otherwise trigger the switching action.

The comparator A304A is an open collector type wand uses R337 and C311 for a load. This rc combination provides a 60us off delay which is then sensed by the comparator A304B whose output is used to operate the electronic switches Q101 and Q102 in the left channel and Q201 and Q202 in the right channel. When a tick comes along, the switch off time typically varies in the range from 60 to 600 us and occassionally up to 2ms. In order to make the tick supression LED P302 visible, the pulse is stretched by network R340 and C112 at the output of comparator A304C so that comparator A304D will turn PL302 on for approximately 16ms for each tick.


When trouble is suspected, it can be isolated to the maor sections causing the trouble by simple sine and square wave tests and observing front panel LEDs. With power applied and the cover on, or else a shield over the top of the chassis which is grounded, the CAL light should light up brightly. Apply a 0.775 (0 dB) signal at 10 kHz at the input of one channel and advance the senssitivity control clockwise. The CAL light should dim. If it does not, there may be trouble in the detector circuit or in A101A or A101B. Feed into the left channel a 1v pp 100 Hz square wave. Feed into the right channel a 200 Hz sine wave at .775v rms. Observe the sine wave output. By turning the Sensitivity control to maximum and the Threshold to maximum clockwise, the Transient Noise Eleimination LED can be made to operate as the system senses the leading edges of the square wave. Small bulges will appear on the output sine wave where the system has switched over the signal from A102B to A103A. This indicates normal operation of the system. Next interchange the channel inputs and observe the sine wave on the other channel.

If the transient noise elimination LED cannot be made operate on the leading edges of the square wave, the trouble is likely to be in the detector circuit. If this LED operates and there is no audio output, look for trouble in the audio channel.

The last test is to observe the nosie output on one channel when triggered by a square wave on the other channel. Feed a 100Hz square wave at 0.775 v rms into the left channel. Adjust the Sensitivity and Threshold controls so the Transient Noise Elimination LED lights up. Observe the nosie output on the unused channel when the horizontal trace is synchronized with the leading edge of the square wave. Also, listen to the unused channel on headphones. If there is a small noise, this may indicate drift in the zero adjustment potentiometer R232 which can be reset to minimize the noise. The level of noise on the oscilloscope, neglecting a spike of a few microseconds duration, is normally approximately +/- 1 mV and lasts for 80 ms. A very large noise may indicate trouble in the switching portion of the circuit incluidng A202B, Q202, Q201, Q203A, or Q203B. If there is a large noise at no signal input, check the dc output at TP201 which should be 0. Next interchange the inputs and check the other channel.

If there is trouble in the power supply which may be indicated by no CAL light, check the +/- 15V output of the regulators. If a fuse is blown, find the cause of the blown fuse before replacing the fuse and turning the unit on again.

15 Mar 1998

To tutorials index

To Digital Playroom main page

Sorry, Dick Burwen has requested that we don't give out his mail address.

Return to top of document. 11