Metropoulos Forum

Hey guys, I finally found time between work and school to put together a layout for the compressor board from the Echoplex.

However I'm running into a few issues putting the schematic together. This is something Good Guest alluded to in the EPPos thread. I noticed there is a mistake in the EP-comp's design vs. the MXR dynacomp. At the output of pin 6 of the 3080 in the EP-3, you can see a simple RC network of a 150k, a 22k, a 1µF and a .001µF. As per ampsnob's corrections there is also a 47k that comes from the 9v rail to this network.

http://forum.metroamp.com/viewtopic.php ... 45#p374235" onclick="window.open(this.href);return false;

if you check the dynacomp schematic this resistor network is set up very differently. There is a 56k (47k in EP3) off the 9v rail that connects to the 150k and .001µF, and this connection continues to the base of Q1. However the 150k and .001µ also connect to the output of pin 6, making this a feedback loop of sorts. Since the 27k (22k EP3) and 1µF also connect directly to the 56k (47k EP3) they are also part of the feedback loop. Q1 for the dynacomp is a buffer, and since this circuit is likely run after the EP-3 preamp circuit then it is probably unnecessary, and was omitted from the EP-3 compressor circuit.

In the Echoplex, all of these components are located off pin 6, more or less. I'm very puzzled as to why they decided to include these components when they are part of a circuit they omitted.

So my question to you guys, is should I bother to include these in my version? I feel they would actually be detrimental to the sound of this thing, even if it was accurate. Maybe I should rework the circuit so they are part of a feedback loop for the opamp? I'm not sure if i want to trade accuracy for perhaps poorer performance or tone.

I don't think it's used as an RC circuit. In the correct MXR schem, the 56k and 27k set up a voltage divider off the positive and negative(grnd) power supply lines creating a constant dc value (no audio). This is used to bias MXR Q2 through the 150k and .001. Since the MXR has the input buffer Q1 they take a line off for that as well. With the correction, the ep3 schem is set up the exact same way with slightly different values.

The 150k and .001 are not part of Q1's feedback, they are part of Q2's bias. (and Q1's bias in the MXR through the 1m) I don't think there will be any feedback happening there in the MXR because c6 grounds it all out (through c4 .001). With the correction, they work exactly the same way in the ep3 circuit and therefore ARE necessary.

To state this a different way: the corrections to the ep3 schem make it identical to the mxr schem in that area after pin 6 (with slightly different values but nothing critical) Without the corrections, the circuit doesn't work at all (I did try it both ways) Hope this helps.

Alright. If they're involved in biasing Q2, it should definitely remain in the design. it didn't come across to me that was their purpose, but on second look you seem to be right.

I drew up a very messy and unlabelled schematic in what little free time i'll have this weekend. i decided to include the sensitivity control from the dynacomp. just ignore the incorrect labels im just looking to make sure my connections are good before cleaning up the schem. let me know what you think.

http://www.flickr.com/photos/57775906@N ... otostream/" onclick="window.open(this.href);return false;

if it all looks right i'll spiff up the schematic and make up a layout and post it sometime next week. If we still have power on the east coast anyway

There is 1 error in your schem. The collectors in Q3 and Q4 should connect to positive through a 100k and those collectors also connect direct to the base of your Q1. That is to say: R15 is in the wrong place. it should be on a missing line to the positive rail.

Otherwise, looks great. Go for it