Spring tank recommendation

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
zamproject said:
Still the Paia implementation of this idea should work too.
I don't think it is the same. In the AKG circuit the spring is driven from both ends so that the primary signal is nulled in the spring itself. The "boing" is from too much energy in the spring. So by sending an equal and opposite amount of energy from the other end, the two waves cancel each other in the middle. In the PAiA circuit, the energy of the spring is not reduced at all. The primary signal is nulled electrically which might help with noise but that isn't crystal clear to me either. The AKG circuit is way more interesting. Unfortunately it's not possible to do with conventional reverb tanks. You need two separate tiny coils attached to the spring and a C magnet at both ends. There are pic on the Internet but here's one of one end of the AKC circuit spring:

normal_AKGbx20-DettagliFolli4.JPG
 
squarewave said:
I don't think it is the same. In the AKG circuit the spring is driven from both ends so that the primary signal is nulled in the spring itself. The "boing" is from too much energy in the spring. So by sending an equal and opposite amount of energy from the other end, the two waves cancel each other in the middle. In the PAiA circuit, the energy of the spring is not reduced at all. The primary signal is nulled electrically which might help with noise but that isn't crystal clear to me either. The AKG circuit is way more interesting. Unfortunately it's not possible to do with conventional reverb tanks. You need two separate tiny coils attached to the spring and a C magnet at both ends. There are pic on the Internet but here's one of one end of the AKC circuit spring:

You'r absolutely right, I make it too short above.
When I say "same" idea I mean only "nulling" principle

For sure the BX20 is way more interesting with a true physical cancellation! i'm happy with mine  :)

best
Zam
 
squarewave said:
I don't think it is the same.
Indeed it's not.

In the AKG circuit the spring is driven from both ends so that the primary signal is nulled in the spring itself.
I'm not so sure about it. Thorough analysis of coil/magnet orientation and coil polarity is needed to confirm that.

The "boing" is from too much energy in the spring.
I don't think so. If it was the case, simply reducing the drive level would get rid of boing, which is not what I experienced.
Limiting the slew-rate of the drive signal makes a difference, though, because it decreases boing whilst leaving the overall reverb level almost untouched, so boing is masked.

  In the PAiA circuit... primary signal is nulled electrically which might help with noise but that isn't crystal clear to me either. The AKG circuit is way more interesting.
It is indeed, however I think the main reason why it doesn't boing is because of the electrical cancellation of the return signals. Being returned from a common spring, they cancel perfectly.
 
abbey road d enfer said:
Quote

    In the AKG circuit the spring is driven from both ends so that the primary signal is nulled in the spring itself.

I'm not so sure about it. Thorough analysis of coil/magnet orientation and coil polarity is needed to confirm that.

to be more precise:
-there is two coils per end sealed together at 90° (like blumlein...just for the mechanical representation...) in a single magnet field.

at electronic side:
-input amp is splited at output to two driver (with a balance trim), where each driver drive one end of the spring.
-the spring return amp(two of them), coming from opposite spring end coil, is also splited at is output, part of the signal go back to the inverted input of the coil driver amp, an also feed the sum/output line amp !
-at the output amp, one spring end drive the inverted input, and the other the non-inverted input, with also a "balance" trim

-the "time setting" is a variable DC which adjust gain/bias at 3 amp simultaneously, the input, and the two buffer used for the spring feedback in the inverted coil driver input

this is for one side of the reverb....you ave all this twice, stereo unit

the electro-mechanical spring assembly, use two sprigs (around 1.5m each) 4 magnet and 8 coils  ::)

So...there is chance that "boing" nulling is mainly done at sum/output amp, wile the internal feedback (inverted) is the damping
But as there is a complex mix of spring feedback and phase inversion summing in various location of the signal path (including coil ?), "boing" shaping should be complex too...

Best
Zam
 
zamproject said:
the internal feedback (inverted) is the damping
That is also a very important feature, which probably prevent the drive coil to mechanically overshoot (which is probably another cause of boing).
I don't remember having seen the original patent, that would probably explain all that.
 
abbey road d enfer said:
I'm not so sure about it. Thorough analysis of coil/magnet orientation and coil polarity is needed to confirm that.
...
It is indeed, however I think the main reason why it doesn't boing is because of the electrical cancellation of the return signals. Being returned from a common spring, they cancel perfectly.
I must admit that I don't really know for sure how it works precisely. I have never used an AKG spring reverb.

But this block diagram explains a lot:

Screenshot%2Bfrom%2B2018-10-16%2B12-51-24.png


Indeed no polarity information is provided in this diagram. But if the two drive signals from amp 2 are ultimately driving the spring so that the torsional pulses are opposite, then it seems to me that they should physically cancel in the spring itself.
 
squarewave said:
But this block diagram explains a lot:
I'm familiar with this dgm (I have the full service manual), and it's always baffled me (and continues).
Since the pick-up coil is solidary with the drive coil, a large amount of direct signal is present in the pick-up coil. According to this diagram, everything seems to be in-phase, so I don't see how the direct sound would cancel... ???
 
abbey road d enfer said:
I'm familiar with this dgm (I have the full service manual), and it's always baffled me (and continues).
Since the pick-up coil is solidary with the drive coil, a large amount of direct signal is present in the pick-up coil.
Yeah, clearly this diagram is "simplified".

The BX15 block diagram shows a little more:

Screenshot%2Bfrom%2B2018-10-16%2B13-53-24.png


So it looks like the receiver output is also fed back into the driver input to cancel out the direct component you speak of.

The schematic in the BX15 PDF clearly shows the exact circuit.
 
abbey road d enfer said:
I'm familiar with this dgm (I have the full service manual), and it's always baffled me (and continues).
Since the pick-up coil is solidary with the drive coil, a large amount of direct signal is present in the pick-up coil.

I think it's the exact opposite ?
the movement vector inducted in the drive coil by audio signal, make the adjacent pick-up (from other side electronic path) induct opposite electrical signal in the same magnet field (remember the two coil have 90° angle, both axed 45° in the field, mirror like), now if you consider this happen at both end, there is chance that low freq and large spring movement get attenuated at end, where spring mechanically hook the coils assembly ?

At 5000m/s one side can null other side in 0.3ms (for 1.5m spring) so "practically" only nonlinear movement generated by first wave left ?

don't know... this is an instinctive understanding, electromagnetism study is far behind me ...

best
Zam




 
squarewave said:
So it looks like the receiver output is also fed back into the driver input to cancel out the direct component you speak of.
No, it's motional NFB, but the signal in the pick-up coil still contains a faithful image of that in the drive coil. Cancellation must be done in the recovery circuit. I'll have a look at the BX15 schemo, it may be more understandable than that of the BX20.

EDit: actually the schemo is clearer and shows that the signals from the pick-up coils are summed with inverted polarity (i.e. substracted) before the EQ and output stage.
 
abbey road d enfer said:
EDit: actually the schemo is clearer and shows that the signals from the pick-up coils are summed with inverted polarity (i.e. substracted) before the EQ and output stage.
I see. So the direct component is present throughout. But the outputs of both receivers (of one channel) are summed with a simple differential amp so that the direct component is rejected as common mode. What's left is the uncorrelated / delayed components.
 
squarewave said:
I see. So the direct component is present throughout. But the outputs of both receivers (of one channel) are summed with a simple differential amp so that the direct component is rejected as common mode. What's left is the uncorrelated / delayed components.
Yes, that's my conclusion for the moment; I think studying the wiring of the springs in detail is on order for full understanding.  :)
 

Latest posts

Back
Top