The 292M (no letter) Quad Lopass Gate is the original version which does not use vactrols but instead diodes. This is an EMS panel legends. On the no letter schematic the three positions are 20 dB/Oct.,  6 dB/Oct., and gate.

The original had a hand wired panel but this version uses a panel PCB designed for this module and the 292A. Neither had an All-Output but the summing op-amp was included on this panel PCB if the user so desired.

The white wire is -15V to the all-out op-amp.

The main PCB shows the typical symmetry and you can see the 6 diodes in each filter section.

There are a lot of wires in close proximity around the switches. I chose to use shielded cable as indicated on the schematic for the one side of the switch. I chose to not use a common 10R resistor for all four shields and instead wired the shields to ground. You can see the short black wires from the switch to the ground trace. There are only three standoffs for mounting the PCBs.

Calibration

There are two trimmers per channel. Set both to the center to start. Monitor the output voltage. I find an oscilloscope or analog meter easier than a DMM since the calibration is to minimize small voltage changes. There are no inputs needed for calibration.

Measure the Output voltage and adjust R15 (10M) to set the output to 0V.

Adjust R13 (5K) so there is minimal change in the voltage as you vary the CV control over the full range. Don't worry if the output is not at 0V. You want to minimize any change in output voltage over the full range of the control.

Once R13 (5K) is set then readjust R15 (10M) to bring the output back to 0V.

The All Out, which is not on the original, is the DC sum of the four outputs so will likely have some DC offset. At first calibration mine was about 0.2V. The first channel seemed to be the culprit.

Operation

Wow, this is a neat filter. I like it. This is a 440 Hz square in 20 dB/Oct. mode.

This is the same signal at the same control setting in 6 dB/Oct. mode. See my frequency response plots at the end for why this setting is more heavily filtered.

This is the same signal at the same control setting in gate mode.

A bit more control at 3 o'clock yields a better formed signal.

As to be expected with no vactrols, there is no decay.

I replaced my 292MB with this module (I'm out of space) and ran this module to the left and my EMS 292C to the right and compared them. They were very similar. It was hard to tell on speakers so I switched to headphones. The only significant difference between them was the 292nl seems to need a bit more CV to open up in gate mode. I haven't done any analysis but this could require 15V CV. I just added a bit more control which equalized the response.

I did a frequency sweep to compare the rolloff of the 20 dB/Oct. and 6 db/Oct. modes. The channels appear quite well matched so I setup channel 1 (yellow) to 20 dB/Oct. and channel 2 (cyan) to 6 dB/Oct. Both controls were set to mid position. I swept a 258C with a 281 to show the resulting frequency response. You can see the 20 dB/Oct. response is much sharper and the 6 dB/Oct. more gradual with a lower Fc.

The initial frequency of the VCO is ~80 Hz.

Fc of the 20 dB/Oct. is at ~1190 Hz.

Fc of the 6dB/Oct. is nearly half at ~610 Hz.

I then patched my 292C EMS with the exact same settings. It was hard to see both responses so I superimposed them in this image so the cyan is green when overlaid. They are fairly similar.

To better compare these I inverted and overlaid the 292nl (red & blue) on the 292C EMS. They have similar shapes but differ in amplitude and Fc. No wonder my headphone test sounded similar.

back