You are here:  Home > F5SWN > Construction > RF Clipper

RF Clipper

 
 RF Clipper

Following the restoration of the FT-250 and in order to optimize a little the SSB emission of this transceiver, why not try the construction of one of these famous RF Clipper as it existed in the 70s ~ 80.

The scheme is inspired by the writings made by PA0FRI on the one hand and RA4NAL on the other. Each party has been the subject of tests and optimization to achieve a very satisfactory result, given the cheap components used.

 

 RF Clipper

In SSB, the emitted power depends on the level of speech. This having an important dynamic range, the resulting average level is much lower than half the crest level.

We can increase the average level of speech by compressing this dynamic range but it will be at the expense of quality and will generate harmonics. This drawback can be limited by clipping after transposing the audio signal to a much higher frequency. Harmonics and intermodulation products are also transposed outside the audio range.

It is an external solution that is inserted between the microphone and the transceiver. There are others where we intervene in the IF stage, modern transceivers with digital technology then having algorithms for this.

RF Clipper
RF Clipper
RF Clipper

 

The principle of this circuit therefore consists in generating DSB using a modulator (NE612) whose local oscillator is set around 453.5 KHz via a ZTB455E ceramic resonator. The DSB signal at 455 kHz is then clipped by a stage made up of two germanium diodes and a transistor. Then comes the side band filter, a 455 kHz ceramic filter, which will recover the upper side band by adjusting the local oscillator to position it in the center of the bandwidth of the filter. It only remains to demodulate this signal using a second NE612 which will have as a reference oscillation that produced by the first NE612. The audio signal thus recovered is filtered then leads to the microphone input of the transceiver.

The printed circuit layout does not integrate the preamp-AGC module for lack of space in the case. It was finally stuck on the filtering capacitors.

Power supply

RF Clipper
RF Clipper

The choice to use an battery makes the device independent of any external voltage source. This is a li-ion rechargeable 9V battery with a USB port. The battery support drawer has been modified to allow the load cord to be left in place.

Two LEDs are located on the front: the green is lit as long as the tension is greater than 8V, giving way to red LED when less than 8V.

 

NE612AN and ZTB455E

RF Clipper

This Chinese ceramic resonator with a very low Q factor required to optimize the values of the oscillator capabilities in order to obtain a beautiful sinusoidal form.

Clipping

Relatively classic and made up of two signal diodes mounted head-head, it is preceded by resistance so as not to overload the exit from the AN612. The clipped signal is taken from the collector of the BC547C transistor.

Filter

The 455 kHz LTM455HW ceramic filter has a bandwidth of 6 kHz and costs less than € 3. A better result can be obtained using a better sort or quartz filter on a higher frequency (for example 10.7 MHz).

Preamplifier stage

Two reasons to precede the modulator by an intermediate preamplifier stage :

  • The output level of the dynamic microphone used (HC-4 Heil Sound cell) is a little low in the input of the AN612.
  • It is preferable to have a limited level in amplitude so as not to cause the DSB product alteration.

I finally opted for two switched solutions :

  • A preamplifier module with AGC, the MAX9814, very effective and thus allowing to have a capped output level in amplitude whatever the input signal.
  • A preamplifier/compressor module, the SSM2167. Already used in the past, it is an excellent modulation compressor with very little distortion for a ratio of 2:1 to 10:1. Gives more presence but is also more sensitive to ambient noise.

Adjustable resistances make it possible to balance the input and output levels in order to obtain an optimal signal at the input of the clipper, as well as at the outlet of the circuit.

RF Clipper
RF Clipper

 

C20 adjustment method

Oscilloscope measurements and use of a two tone generator.

Probe on the T1 collector
Apply the two tone generator to C1 so that the clipped HF signal is not saturated.
In principle, this should correspond to an input level on C1 of 150 mV maximum.

Probe on output L1/C16
1300 Hz on C1, about 80 mV.
Adjust C20 for the most regular wave form and the highest amplitude.
I also did it with the two tone generator, it works.

Probe on C11
We should now measure 453.7 kHz (it depends on the previous setting and the bandwidth of the filter).
The C11 output amplitude is around 500 mV at 455 kHz and quickly decreases above.
But it will be slightly greater at 453.7.
Note that 453.4 corresponds to a audio frequency centered on 1000 Hz
454 is suitable for a frequency centered on 1600 Hz (the modulation will therefore be acute).
454.7 corresponds to a compromise between the two (1300 Hz) and is suitable for most modulations.

Modules level adjustment
Input microphone, compressor position
Probe on C1,
Adjust R14/R1 so as not to exceed 150 mV (R14 is useful for limiting the sensitivity to ambient noise of the microphone).

Then microphone at the input, preamplifier/AGC position
Probe on C1,
Adjust R15/R16 so as not to exceed 150 mV (R15 is useful for limiting the sensitivity to ambient noise of the microphone).

R6 is adjusted to obtain a level consistent with the position off of the circuit (microphone alone) or depending on the desired level at the input of the transceiver if it does not have mic gain.

Some pictures

Illustrating the construction as well as some measures.

 

Recording

As a local listening on a receiver (Yupiteru MVT7100) placed in such a way that its s'meter practically does not deviate, the FT891 being on fictitious load.
The recording is made via a magnetic loop placed on its speaker. The microphone test alone, then of the two positions of the RF Clipper gives a fairly blatant result.

The second was carried out with more difficult conditions, the reception noise being greater while the received signal is even lower. Thus, with the microphone alone, it is even sometimes difficult to understand what I say. We better understand the potential of this kind of device which turns out to be very useful when the noise level is important.

Conclusion

In practice, the result is quite surprising. The compromise between quality and efficiency remains at the service of increased intelligibility and we find the 6db gain as if we had multiplied by four the power of the transmitter.

But we must not lose that if the average power increases, the final stage of the transmitter is much more requested. It is therefore preferable to relieve it by avoiding full power.

I personally adopted the following method: on fictitious load, I adjust the microphone gain of the transceiver in such a way that whistling on the microphone, the power does not exceed 70 watts with the clipper triggered.

I then see a modulation power around 60 watts while the ALC does not move at all. In use, the transmitter never comes out more than 70 watts and it does not overheat.

Glossary

DSB

DSB means "double sideband". This is generally what a modulator produces: an AM signal from which the carrier is removed. Only the two side strips remain, one of which is then filtered to choose the upper (USB) or lower (LSB) side strip. We then speak of SSB (Single Sideband).

Here is an article about the modulation of a signal (Wikipedia).

FT-250

FT-250 Sommerkamp, a Tube HF transceiver (16). This dates from 1977, probably one of the last manufactured. Its restoration has been done in several stages but it is now in a good place in the station :

FT-250 Sommerkamp

MAX9814

Microphone amplifier module with AGC :

MAX9814 module préampli à CAG

See on adafruit.com for information.

 

MVT-7100 Yupiteru

Scanner 53 kHz - 1650 MHz, AM/FM/WFM/SSB/CW

MVT 7100 Yupiteru

Description on Rigpix.

Q-Factor

The Q factor expresses the quality of an RLC circuit and is defined at its resonance. If this factor is high, the bandwidth is reduced. It also represents the ratio of the measured voltages on the circuit at the resonance and that of the source.

RA4NAL

His article about clipper :

https://ra4nal.ontvtime.ru/clipper.shtml

 

SSM2167

Preamp compressor with SSM 2167 circuit :

SSM2167

You can find it easily on ebay.

PGEgaHJlZj0iL2VuLz9GNVNXTi9Db25zdHJ1Y3Rpb24vUkYtQ2xpcHBlciZhbXA7bG9naW4iPkxvZ2luPC9hPg==