MicMix MasterRoom XL-305 Spring Reverb Clone

[mediatechnology]

I imagine the schematic would probably match the C series previously posted, but without the EQ

I think you'll proabably find it to be closer to the older "Studio B" circuits though 2 channel.

I think I posted the "Super C" which was an updated version.

[ethervalve]

Hi Wayne and all,

My Master Room XL-515 just died on me.

I think the power transformer is cooked. It shut down on me and I noticed the fuse was blown; I replaced the fuse with a 1A (which is what the silkscreen said) but after a few minutes it shut down again and the transformer was extremely hot. At this point I noticed that the previous (original?) fuse that had worked for years was actually 1/2A.
The transformer itself is stamped with the following two lines of numbers:
3111 1
7578206

I was wondering if you might have any idea what type of transformer I'll need? The IEC power jack is labelled 120V 50-60HZ 35VA.
Based on what I'm seeing on the power supply board I'm guessing that the voltages are +15/-15VDC and +5VCD (the last one is the only one that's silkscreened on the pcb).

There are six wires going from the secondary to the power supply board. The regulator chips are LM317K plus a 7915C and a 7815C.

Any tips or insights on this would be much appreciated. This is probably my favorite piece of outboard gear--it's pretty tough to live without.

here's the transformer:
https://e4eedbe25a376916dc64c37866ed0d3 ... dtrafo.jpg
bottom view of PSU
https://e4eedbe25a376916dc64c37866ed0d3 ... bottom.jpg
top view of PSU
https://e4eedbe25a376916dc64c37866ed0d3 ... psutop.jpg

[mediatechnology]

I don't have a lot of knowledge about the XL-515.
After John Saul died Bob Rodgers continued running MicMix and he, along with co-owner Bill Allen, did the '515.
The work I did on downward expansion to shorten reverb time that was used in the XL-515.
That same circuit, after being tweaked by Bob Rodgers, later became the DynaFex.

The transformer was likely made by a company called Texas Transformer.
I suspect that it has a 20-0-20 and approx 8V windings and could be replaced with just about anything capable of producing +/-15V and 5V outputs.

You might disconnect the transformer secondary to make sure that it's not a blown diode or filter cap causing it to overheat.
Judging from the coloring of the lacquer it has gotten quite hot.

You also might look at replacing the whole assembly with a Meanwell switcher.
http://www.jameco.com/webapp/wcs/stores ... _237367_-1

There are a few Meanwell's having both 5V and bipolar outputs.
I have about three in stock and have been wanting to try it with the Dual Class-A.
I think Roger Foote uses the Meanwell in some of the FCS series.
They are far quieter than the specs would indicate.
The above is 5V at 4A, +15V at 1.5A and -15V at 0.5A.

[ethervalve]

Thanks so much for the great info Wayne.
I'm going to try to get this back to life ASAP.
I'll to revive the built-in PSU first but that switch-mode supply looks like a good choice too!
That Orban unit looks interesting. Is that a single-ended noise reduction scheme? I love how the downward expander works on the reverb.

[mediatechnology]

Yes, the MicMix/CRL/Orban unit had both downward expansion and a sliding filter.

SSM made a Dynafex IC, the SSM2200: http://www.ka-electronics.com/images/SSM/SSM2200.pdf

[mediatechnology]

I found the box of Accutronics springs way back in the closet and returned to this project after a very long 11 year absence.

I measured the spring I cut 12 years ago and found it to be 13.76 ms just a tad longer than the 13.50 ms.

The question is how much shorter should it be?

The 13.76 ms spring measured 259 turns making the delay equal to 53.13 µs per turn.
A 13.50 ms spring should have 254 turns.


Impulse Delay of a 259 Turn Accutronics Reverb Spring

Hopefully we'll see the results without another decade passing.

[mediatechnology]

I got the 12 spring assemblies cut and calibrated the delays.

Once I established an approximate delay per unit length, about 3.8 ms per inch, I rough-cut the remaining lines slightly long and then shortened them based on "TDR" measurements to the exact delay.

There appear to be 72 turns per inch which makes each turn about 53µs delay.

My measurements are based on original Accutronics springs produced around 2007.

The top two sleds are for the left channel the bottom two are for the right.
When mono'd the two channels delays interleave to provide 12 early reflections.


MasterRoom XL-305 Clone Reverb Sled Assemblies

Now I need to build the second channel of the preamp and cut some test files.

[mediatechnology]

Reverb sleds paired Left/Right:


MasterRoom XL305 Clone Reverb Sleds

At the moment masking tape holds them together...
Clip leads ground the frames.

Recovery amplifier:


MasterRoom XL305 Clone Balanced Recovery Amplifier Using the Flat Moving Magnet Preamp

The balanced flat moving magnet preamp makes an excellent reverb recover amplifier.
The reverb transducer output has a large amount of inductance, 390 mH, similar to a phono cart.
The pickup transducers are wired in series making the inductance 780 mH.
Low current noise is a benefit in this application.
An unequalized phono preamp has better input characteristics than a mic preamp for use as a recovery amp.

Drive amplifier:


MasterRoom XL305 Clone Balanced Drive Amplifier Using the Dual Class-A II Headphone Amplifier

I originally used constant current drive to maintain consistent drive current versus frequency in the original XL-305 with HF rolloff. See: XL-305 Schematics https://www.proaudiodesignforum.com/for ... f=12&t=197 XL-305 Constant-current driver: https://www.ka-electronics.com/MasterRo ... _Tile2.JPG

At the moment a Dual Class-A II headphone amp is providing constant-voltage zero Ohm output.
I tried a 47R build-out to simulate constant current drive to the 4Ω transducers (2X 8Ω wired in parallel).
Constant voltage sounded cleaner but bass-heavier; the 47Ω was a bit thin.
An equalized 0Ω output to simulate constant current may be the best option or a true constant current output using op amp feedback.

Sound files coming...

[JR.]

Springs were always a major PIA in production and even more so for design. Just think about managing hum pickup inside a tight powered mixer top box. I was pretty happy when digital reverbs finally became cheap enough to replace them...

JR

[mediatechnology]

As promised here are some test samples, mostly percussion.

The left channel and right channels each have six springs.
On a couple of tracks, the woodblocks for example, you can hear a slight flutter in the early reflections.
When mono'd the left and right channels interleave to provide a 12 line reverb.
The XL-305 had 12 lines; the entire MasterRoom series had 16.
The XL-205, which I did not do, only had six.

MasterRoom XL-305 Clone Sound Files: https://proaudiodesignforum.com/content ... Stereo.mp3

(Dry test tracks courtesy of THAT Corporation.)

The drive should also have a low pass filter.
At high frequencies the springs stop providing delay and have a small amount of direct feed-through.
I don't know the mechanism of this.

Springs were always a major PIA in production and even more so for design. Just think about managing hum pickup inside a tight powered mixer top box. I was pretty happy when digital reverbs finally became cheap enough to replace them...

The XL-305 required two layers of magnetic shielding just to prevent hum from its own transformer.