BTDR-3 Reverb brick discussion

[armdnrdy]

Last night, I plugged in and played through a reverb unit that I built based on the BYOC deluxe reverb. This BTDR-2 design includes a feedback path (Dwell) to adjust the decay time.

I have been contemplating building another reverb with the BTDR-3 which includes four pins two "adjust" the decay with resistors or a double gang pot.

I was looking at the info available on the BTDR-3, and a few things came to mind.

It seems that the internal feedback resistors were deleted or replaced with lower values and tap pins were provided to allow the decay to be set or controlled externally. This is the alternative to offering three different models with short, medium, and long decays.

So...is there anything really gained with this adjustable decay version if I already have an adjustable feedback path in my build?

The only difference I can think of is my circuit's feedback path is taken from the output of the brick and sent to the input.
The BTDR-3 has feedback taps between the PT2399s.

I wonder if there would be much of a noticeable difference in the end.

Any thoughts?

[Fender3D]

As far as I can remember BTDR1-2's sound, BTDR3 has a more "natural" decay if compared to BTDR1-2 + feedback.
I used 100K+100K and reverb gets really long, but it has not the usual heavy feedback's rumbling sound...
And it won't suffer for any added LP filters in FB path...

[Mark Hammer]

The filtering of the feedback is generally what creates the feel of any ambience-generating circuit, by mimicking the kinds of reflective (or nonreflective) surfaces found in whatever space one is attempting to mimic.

There are all sorts of opportunities for additional filtering of both the primary delay signal into the dry+wet mixer, as well as the feedback/dwell path, that would probably do more for the pleasingness of the resulting reverb sound than any change from BTDR1-2 to 3 modules.

[armdnrdy]

Thanks for the input guys!

I usually use a very conservative amount of reverb with a short decay.

The BYOC design looks like it has filtering for the reverb output as well as the feedback path. It sounds quite good!

At first glance..I thought that the "new" brick had a cool new parameter, (Accutronics calls it "Depth") but upon closer inspection..the "new" brick looks like the same old thing as outlined above.

[Mark Hammer]

It seems like the "depth" control is the dwell circuit, but with the mixing stages built in.  http://www.tubeampdoctor.com/images/File/BTDR-3%20DIGI-LOG%20REVERB%20MODULE.pdf

The quasi-stereo nature of the module presents some interesting possibilities.  It is a simple thing to stick bass-limiting caps in a few places and use the stereo aspect as a means of controlling tone.  So, for example, a person could stick a .01uf cap in series with one of the depth pots, and a 4700pf cap in series with the 39k mixing resistor in the same channel.  If you wanted to get fancier, you could also take the 39k mixing resistor in the unbright channel, spit it into 15k and 22k, and run a .01uf cap to ground from their junction.  The Depth and Level pots of the two channels (mixed down to mono) could then be used to achieve different ambient sounds, the one being the bright version, and the other the duller version of the reverb.

Here's an interesting little nugget at the end of the spec-sheet for the BTDR-2:

"Although Accu-Bell believes that circuits employing solely the BTDR-2 will easily pass FCC Part 15, no guarantees of compliance are made; the circuit must be tested as a whole for radiated and conducted emissions"

[armdnrdy]

Smart ideas Mark!

The design that I used is for the original larger BTDR-1 which has only one output. I used the BTDR-2 (smaller, with two outputs) tried to connect the second output...no go...distortion!

So I disconnected one of the outputs...worked fine! but...I now see where the other output could have been used.

Maybe I'll mess with another build. 

[armdnrdy]

I was looking at the drawing of the BTDR-3 circuit and noticed something interesting.

The "Depth" pins are labeled 1A, 1B, 2A, and 2B. That correlates with OUT1 and OUT2.

So...I wonder if there are two feedback paths that go from the output to the input of each "channel" instead of taps in between the three PT2399s?

[Mark Hammer]

I was looking at the drawing of the BTDR-3 circuit and noticed something interesting.

The "Depth" pins are labeled 1A, 1B, 2A, and 2B. That correlates with OUT1 and OUT2.

So...I wonder if there are two feedback paths that go from the output to the input of each "channel" instead of taps in between the three PT2399s?

I'd like a little more info on what's inside the module, and whether the two "channels", internally, are identical, or different from each other in some way.  Given that they take a mono feed, I gather that the independent dwell and level controls are really more directed at higher-end karaoke installations, where a little more reverb needs to come from those speakers than from these ones.  Still, if the channels are different in some way, it would help spark some useful ideas for how to use them.

[armdnrdy]

This brick is somewhat of a mystery...that's why I'm playing the guessing game..trying to figure some of this out through deductive reasoning! 

I tried to find the patent for this model on the net but...the only drawing available is for the 1st single output model.

I can't even find a PCB footprint layout on the Accutronics sight!

Maybe I'll email Accutronics and ask a few questions.....and just maybe...I'll get a reply.

[Mark Hammer]

I would think the simplest and most straightforward and innocent question to ask is: are the two internal channels different from each other, and if so, how?

A question like that doesn't result in any proprietary information leaking out.  If anything, it simply enables end-users to think of more applications for the module, increasing its marketability.

[samhay]

The two outputs of the BTDR-2 are identical, as far as I can tell.

What would be cool with the BTDR-3 (I think) is if DIFFERENT resistances on the Depth pins give differences in the nature of the reverb and/or of the signal in the two outputs (is this what Mark was getting at?) - i.e. don't use a dual-gang Depth pot, but use 2 x 10k (or whatever).

[armdnrdy]

Yeah...I have a BTDR-2 here...in a circuit.

I was thinking about scoping the output pins with a sine at the input....see if I can gauge any difference.

[Fender3D]

...
The quasi-stereo nature of the module presents some interesting possibilities.  It is a simple thing to stick bass-limiting caps in a few places and use the stereo aspect as a means of controlling tone.  So, for example, a person could stick a .01uf cap in series with one of the depth pots, and a 4700pf cap in series with the 39k mixing resistor in the same channel.  If you wanted to get fancier, you could also take the 39k mixing resistor in the unbright channel, spit it into 15k and 22k, and run a .01uf cap to ground from their junction.  The Depth and Level pots of the two channels (mixed down to mono) could then be used to achieve different ambient sounds, the one being the bright version, and the other the duller version of the reverb.

BTDR3 (and I guess 1 and 2 also) doesn't work that way...
It is not quasi-stereo, it is real stereo since it (they) creates different reflections (echoes), from a mono signal, for both out 1 and 2.
The Depth pot is quite different from a feedback control. Interestingly the longest reverb is achieved with the highest resistance.
When you disconnect 1 lug, the reverb time is the longest, I used 100K on mine and it is slightly longer with lug unconnected.
Furthermore the 2 Depth paths interact each other, but only for the reverb time obtainable, and indipendently from outputs.

i.e.
if you disconnect pot 1A-1B and short 2A-2B you have a strange reverb (more like a bunch of echoes, as if it's made with a lot of predelay, but not periodical like just 1 echo), and you can hear different reflections on each output (1 and 2)

Ask, until I have it on my breadboard 

[Mark Hammer]

I'm a little disappointed that my assumptions were wrong, but even though I was wrong, I am encouraged that those controls still give us something to play with when it comes to shaping the sound by means of differential use of the two channels.

I'd still like to know what the differences are between the two channels, though.  I am a curious little devil. 

[armdnrdy]

Federico,

Thanks for spending the time to breadboard and experiment with the brick.

It almost sounds like what you are describing is some sort of tap between the PT2399s.

Three 2399s will give a reverb effect..but the tap from one or two 2399s should produce what you described as, "more like a bunch of echoes"

It is still very unclear how this is all connected.

Something of note...I just checked continuity/resistance between the two output pins of the BTDR-2 with a meter....
No continuity...no resistance reading. So...there is something more than just a basic resister mixing node connecting the output pins to the reverb output.

[PRR]

> anything really gained with this adjustable decay version if I already have an adjustable feedback path

It is a fake "room".

Delay is the size of the room.

Feedback is the hardness of the room.

Take two rooms: a concrete cellar and a large cathedral. We might even find they both have the same total reverberation time (time for an impulse to decay below hearing). But they may sound very different.

In a concrete cellar the first bounce comes in about 10mS (0.010 Seconds, 10 feet), and repeats about every 10mS thereafter. Sound may hit a wall 200 times in 2 seconds, but if walls are so hard they reflect 97% of what hits them, it will take 2 seconds to fade away.

In a cathedral the first bounce (not counting the wall near the performer) comes in about 100mS (100 feet) and repeats at the same slower rate. Sound may hit a wall only 20 times in 2 seconds, and if the average reflection is 70% (some 99% marble, some 10% drapes) then fade-away time is still 2 seconds.

Although the fade-away time is "same", your ear is not fooled. It "knows" a big room from a small room. The repeat rate is a strong clue, which our caveman brains are sensitive to. (Is that tiger toe inside the canyon or inside this cave?)

[Fender3D]

I'd still like to know what the differences are between the two channels, though.  I am a curious little devil. 

It looks like the 2 outputs are taken from 2 different delay chips.

Returning to PRR's cathedral, you can feel your position in the room by hearing the first reflections, the following later reflections (almost perceived in mono) will give you a rough guess of the room dimension.

In the Belton chip, you clearly hear the first reflections (from 2 different delay chips, one on out1 and the other on out2), then, I guess, they feed signal "criss-crossed" back to chips, obtaining the reverb tail.

[Mark Hammer]

So if there is no attempt at introducing feedback/regeneration, and we listen to the two outputs, independently, one has a longer delay than the other?

[Fender3D]

It is not necessarily a "longer delay than the other", even if a longer delay on one output is used indeed...
You'll hear first reflections clearly separated on the 2 outputs, then the reverb tail mixes in both output fading to almost mono reverb...

Imagine something like this:

[armdnrdy]

Interestingly the longest reverb is achieved with the highest resistance.

Looking at the BTDR-1 schematic, you can see how the internal connections including feedback paths and summing nodes are implemented.

Two 2399 inputs are connected to the incoming "dry" signal. The summing nodes are the 11.8K, 10K resistors. The feedback path is marked "GF." (Global Feedback) which is the return from a single summing amp. (TS321)

This is also described in the patent available on the net.

I gave some thought as to how you would get longer reverb with the highest resistance.
If the "Depth" pots are connected in parallel with the 11.8K feedback resistors, (GF) when the resistance is increased on the pot, the feedback resistance will be lowered allowing more signal to pass.
More feedback = more reflections = more apparent depth.



Some things of note:
I traced the schematic of the brick back to a Ukraine stompbox sight. It was a member of that forum that opened a brick and traced the component values. The brick that he traced was a BTDR-1 Long Decay.

I also found this schematic posted on the same thread. I was using google translate because my Ukrainian is a little rusty! I am not sure of the origin but...it shows output 1 and 2 coming from two different 2399 outputs. It also shows the single summing amp (TS321) which is not depicted in the first drawing above.