More volume from Demeter Compulator?

[pappasmurfsharem]

I built the demeter compulator using the layout at tagboard effects.

I used all the sames parts , VTLC510, LF353 Opamps, the only difference is the 620 ohm ressistors instead of the 619 Ohm (AKA not a difference)

It works great the tapers aren't quite perfect but its not like I am adjusting the knobs mid Riff.

I lose unity volume when the Comp pot is at the last 15% of its rotation.

Is there any way to get it above unity without compromising the effect or using a recovery stage in place of the volume control?

Rolling the comp knob below 85% gets it JUST at unity volume (assuming average pick attack) obvious if you beat the hell out of the strings in bypass it can be a much larger difference.

I've seen the occasional post that others have a similar issue, I don't know if it effects the real deal or not, but I would definitely like to have the max squish when "needed"

http://tagboardeffects.blogspot.com/2012/03/demeter-compulator.html

[Morocotopo]

I don´t have an answer for you, but I can confirm the same behaviour from my DIY build. It seems even a little bit under unity gain at max compression.

[pappasmurfsharem]

I don´t have an answer for you, but I can confirm the same behaviour from my DIY build. It seems even a little bit under unity gain at max compression.

I suppose it is nice to have first hand reassurance that it is not just me. Did adjusting the trim help you at all? I set mine at 1.104V (factory 1.1) Most things I read said it didn't change their issues. It has great sound in the areas that I get above unity or even a little boost, so I don't want to turn it and change its character if  I don't have too. I get no distortion at all. It's a  really clean sound I can't wait to try it with my bass in my gallien krueger 1001RB, I've only used it with my guitar on my practice amp so far and love it.

If the smarter powers that be can assure me that it will increase the effect output volume I will play with it.

[PRR]

A compressor/limiter is a VARIABLE-Gain device. Asking for "unity gain" is nonsense.

What it sounds like: you play loud, signal levels run higher than this limiter's limit.

Add a third diode in series with the other two. Instant 50% higher limit.

[samhay]

PRR is right - unity volume is only going to happen sometimes with a compressor. That said, you usually get enough makeup gain somewhere in the circuit to give you plenty of volume at max compression.
It looks like the trimpot should set the gain of the second op-amp. This will increase the uncompressed sound and drive the next op-amp harder, which will compress the sound more. What happens when you set it lower? If you want less compression = louder, then you could also try increasing the 620R resistor to the vactrol's LED.

[pappasmurfsharem]

A compressor/limiter is a VARIABLE-Gain device. Asking for "unity gain" is nonsense.

What it sounds like: you play loud, signal levels run higher than this limiter's limit.

Add a third diode in series with the other two. Instant 50% higher limit.

Basically what I am saying is.

Striking the strings a little harder until I reach that point where you no longer increase volume from attack alone/where it hits the volume limit threshold (which is still fairly light picking)

Then rolling up the compress knob continues to drop signal level at about 15%.

I am aware that unity volume is not a technically applicable term here, but it is the terminology to get the point across. I suppose I mean unity volume for my velocity of playing

Will adding a 3rd diode change the effect? will it have the same type of character and range of compression, just louder? or would I need to add a booster to the end to do that?

Like with my keeley 4 knob clone there was so much volume on tap that even at max compression I could overdrive the front of my amp with it. like a booster.



EDIT:
I found this in the Demeter manual:

Compress affects the amount of Gain reduction (compression) of the input signal. Turning this clockwise
will give you up to 30 dB of gain reduction (depending on input gain). Please note that the Compulator's
maximum gain is 26 dB (see Trim Pot), so in some
circumstances if your instrument is very hot you
could achive less than unity gain if you turn up the
Compress knob too much

So that explains the volume loss, and if I have it set to 1.1V on the trimmer it's according to the manual "stock" which is 20dB of gain.

So what is the best plan here, throw a recovery stage at the end?
Any recommendations I suppose I would only need a light amount of boost since the gain reduction and the maximum gain of the circuit isn't too far off?
Also how would I go about that, just take the wire from Lug 3 of the volume and run it to the input of a booster?

[samhay]

The best plan is to try turning the trimmer up - that's why it's a trimmer.

[pappasmurfsharem]

 

The best plan is to try turning the trimmer up - that's why it's a trimmer.
quote]


It  still has a volume cut.

I reiterate 30dB of gain reduction 26dB of gain

[PRR]

> Will adding a 3rd diode change the effect?

Easy to try.

[Kimster]

Hi! When I built this I changed to lin pots instead of log, both the comp and the vol. It behaved much better that way, but that doesn't make sense for your comp pot, but I remember that the vol pot behaved much better with a lin one cause otherwise all the volume came in the end of the pot.

"So that explains the volume loss, and if I have it set to 1.1V on the trimmer it's according to the manual "stock" which is 20dB of gain. "

Don´t you mean 1.1K ? Don´t know how to set it to 1.1V

[pappasmurfsharem]

Hi! When I built this I changed to lin pots instead of log, both the comp and the vol. It behaved much better that way, but that doesn't make sense for your comp pot, but I remember that the vol pot behaved much better with a lin one cause otherwise all the volume came in the end of the pot.

"So that explains the volume loss, and if I have it set to 1.1V on the trimmer it's according to the manual "stock" which is 20dB of gain. "

Don´t you mean 1.1K ? Don´t know how to set it to 1.1V

Correct I did intend for that to be 1.1K . Most of my posts are from my phone, which auto-corrects randomly after the fact. It also tends to jump around as I type. Chrome on Android is bleh.

The action of the pots is a little off, but its still fairly "linear in sound" on mine even with log pots I don't see utility in changing them at this point, but YMMV.

Changing to linear however won't help my volume loss. I'm probably just going to replace the volume pot with a very low gain booster, or try PRR's diode suggestion. The downside is I left very little room for splicing in the diode, and in my lack of pre-planning, removing the board from the box would be a mildly arduous task. It should be fairly simple to add a small volume daughterboard though.

[samhay]

Sorry - missed your second post.
If you don't want to change the character of the pedal as you have it, then a booster at the output sounds like a good idea. Should make for a more versatile effect.

[rankot]

I know it's quite an old story, but what if we change those 1k resistors which divide signal to output and compress to 100 ohms? Would it give more output?

[PRR]

> change those 1k resistors

The 1Ks drive 20K loads.

20K/21K= 0.95
20K/20.5K= 0.97

Increase 1.027 or a quarter of a dB.

Inaudible.

[amz-fx]

I know it's quite an old story, but what if we change those 1k resistors which divide signal to output and compress to 100 ohms? Would it give more output?

Here is an observation... you want to be aware of the dc output level of the detector as it turns on the opto/photocell to set the base compression amount.

In the stock configuration, the dc at the output of the LF353 detector is about 3.15v. This means that the opto is already turned on since the LED inside the opto has a Vf less than 1.5v. If the LED is partially on, then the cell has lowered resistance and the audio signal is being attenuated before hitting the LF353 output stage.

If you add a diode to the detector (3 total), the dc idle goes to about 4.29v so the cell is turned on even more and the signal is further attenuated at idle.

If you remove a diode from the basic detector (now 1 total), the dc bias from the opamp output to the opto falls to 1.96v so the photocell is barely turned on and more signal gets through to the gain stage, i.e. when you roll back on compression, there is more audio output. The trade off is that the maximum available compression is less, but is still probably more than sufficient for heavy squashing. The gain trim can be adjusted to manage the output volume, but it also interacts with the compression control.

Just a few notes on the detector operation and you will have to decide what works best for you. Easy to try.

Best regards, Jack

[rankot]

I have built few different compressors and I can notice that Demeter Compulator is the most quiet. Ross has the most power, although it is a bit noisy, Clean Sweep is really nice and with volume at approx. 50% is like "unity gain" (almost no change in volume with compressor on/off). I have also built Engineer's Thumb, but I don't like it, it is almost unnoticeable. However, Demeter Compulator works, but when it is on, sound is almost 50% quieter than without it. I watched Youtube video demo for DC and it doesn't behave like that there - on/off sound is almost the same in volume. I'll be very grateful if someone can help us troubleshoot this. 

[rankot]

One interesting thing, too: datasheet for LF353N says it must be run on ±6V minimum, while this is running on ±4.15V. Could this be the cause of insufficient amplification?

[amz-fx]

I don't see that voltage limit on any LF353N datasheet, though I only looked at a couple. On the TI datasheet it says that the PSRR was measured at ±6V minimum but that is just the low end of the range for that test.

The ST Electronics datasheet says that the operating voltage range is 6v to 36v, which would be ±3V minimum and ±18V maximum.

Best regards, Jack

[rankot]

What would happen if I run Demeter on 18V? More headroom, more gain?

[PRR]

> datasheet for LF353N says it must be run on ±6V minimum

Like Jack, I do not see that.

The maker does not want to promise anything at the extremes. (Actually they do, but they do the chip different, a different type, and plaster "Low Voltage!" all over the sell-sheet.)

I see supply current, OLVG, and CM input all plotted for +/-18V down to +/-5V. Voltage gain drops slightly at lower voltages. (In-circuit, voltage gain is more about the resistors than the chip!) The supply current is not dropping-off at +/-5V. I would assume (and know from experience) that +/-4V is fine for these chips, for guitar level work.

Knowing more about these chips than is revealed in current datasheets: the bias is (was?) derived from a Zener which goes hungry below 7V (+/-3.5V). Even so the circuit works at mildly reduced performance to +/-3V or a bit less.