The 195 and 200 series receivers had a dynamic noise reduction circuit that looked at RF level, audio level and audio frequency. Those three variables went through an analog multiplier and the result moved a variable high frequency filter up the frequency spectrum. Higher RF, higher audio level and higher audio frequency moved the high pass filter roll off, higher in frequency. The trick was to get the filter out of the way before the listerner could detect the roll off, so the attack time of the filter is less than 0.5 ms. In addition, for any significant high frequencies, the filter is 2 octaves above those high frequencies. There is a little different description in all the manuals for the receivers and is described as a trimode filter.
The filter helped remove compander breathing and "halo" effects around the audio in weak audio conditions. Some designs accomplish this with substantial transmitter pre-emphasis and receiver de-emphasis, such as our 187, 190 and IFB series. This can lead to other problems at high audio levels with high frequencies, most easily shown with the dreaded key test. The down side to our filter method, is that at very low audio levels with very little high frequency content, the 200 system's high frequencies are rolled off. These audio levels are generally so low that they typically show up only when listening to room noise or microphone self noise. The 200 systems also made lavalier microphones sound quieter than they really were. The other drawback, was that it was possible to hear the filter working if the transmitter gain was turned way up in a moderately quiet room, so the self noise of the lavalier and the room noise was high enough to cause the filter to move up and down in response to the random high frequency noise. This was audible as a moderate roughness to the noise.
Just for the record, all the 195 and 200 series have had a basic flat response to 20 kHz or more. The system roll off at 20kHz is not what testers are hearing. ( I base this statement on the fact that some systems with highly regarded audio, roll off before 20 kHz and in one case before 16 kHz and I'm not aware of complaints that these systems are muffled.) At high audio frequencies, the tri-mode filter itself is out past 40 kHz. The reason for the roll off above 20 kHz in the basic system is to prevent supersonics from messing up the compander circuits.
The 400 series does not use a compander or pre-emphasis and so the tri-mode filter was left out. We figured this would help is in competitive comparisons at very low audio levels by making the system sound more open and capable of accurately reproducing small sounds and room noises. However, when some of the beta testers compared 200 and 400 series systems side by side, using a common microphone, they were bothered by the fact that the 200 sounded quieter. (Heart attacks at Lectrosonics.) Further trials by the testers, suggested by a very worried engineer, convinced them that the noise was indeed self noise in the microphone and not noise in the 400 system. This is not to imply that the 400's are as quiet as a wire.
Here came the suprise to the recovering engineers; fully half the beta testers wanted the noise reduction left in, in order to reduce the self noise of their favorite lavalier. Since we don't want to lose low level competitive comparisons because of of lack of "air" or transparency, we have compromised with a menu selectable tri-mode filter. Since users have been "hearing" this noise reduction system for 15 years in all our wideband systems, we have decided to implement it with the same parameters in the 400. The noise reduction is menu selectable at three "strengths".
The power supply for the electret mic bias turns off immediately rather than delaying till the receiver mutes. It is a design error. All LMs shipped after Jan 2005 have been modified. We have a clean fix for earlier units and will modify your LM at no charge. Contact our service department for help.
In general, there is mechanical coupling from the case into the inductors in the main oscillator in the transmitter. A thump on the case moves or bends the inductor, changes the inductance value by a tiny amount and changes the frequency of oscillation. Since a changing frequency is just FM, the FM receiver picks it up as a low frequency thump. There are various ways of reducing the mechanical sensitivity. Most involve very rigid coil assemblies such as inductors wound on ceramic forms. In our case, we use solid quarter wave ceramic resonators.
The cutest trick I've seen, was a (brand) unit that used a miniature Teflon insulated coaxial line as a resonator. They wound the coax stripped off the outer insulation in a tight cylindrical coil with about 6 turns. The entire shielded coax coil was then soldered on the outside into a solid mass. This made a nice rigid assembly with the center conductor acting as the inductive element since a short coax line with one end shorted looks like an inductor.
The other way to generate a thump is to use a capacitor in the audio circuity that is sensitive to mechanical stress. The wrong kind of ceramic capacitor with DC voltage on it can really generate a lot of voltage when stressed. NPO ceramic capacitor types are as good as most film caps or tantalums but X5R types are bad and Y5Z are horrible. NPO's have the least capacity for a given size and the other types have 5 to 50 times more capacity in a given size and that's why they exist. I tried 50 Volt Y5Z type capacitors in the design of the 48 Volt phantom supply for the UH200C. You could get about as much audio talking into the transmitter PC board as you could using a microphone. Fortunately some small 50 Volt tantalums came on the market that would fit in the same space and saved my bacon. I knew the problem existed, but the severity surprised me.
My advice is to whack the case of a transmitter with both your finger and with a pencil sized object. If you know how a transmitter is going to react to mechanical shock, you can prepare for it.
On the subject of mechanical stress and audio, try the same thing with your electret mic cables. Some are much worse than others. If you tap the cable close to the mic (6") you will get mechanical noise transmitted directly to the mic element. In the middle of the cable, it is due to flexing of the mic cable. Phantom powered are sensitive to this since there is DC voltage on the cable and flexing the cable changes the dimensions and the capacitance of the cable. The pro mic manufactures have taken this into consideration in the choice of cable.