Digital Reverb

[AdamM]

This has to be worth a look at even if only for its apparent simplicity:

http://www.profusionplc.com/pro/gex/pcatdtl0?ipartno=NemFX-SC-SP

Might buy a couple & see what I can lash up - maybe a solid state equivalent of one of the classic fender tube reverbs...

Anyone tried these yet?

[anchovie]

That wasn't on the site last time I looked a couple of weeks ago, so it must be very new as they haven't got a datasheet up yet. I suspect it's going to sound "adequate" - it is intended for practice amps after all.

Be aware that Profusion's shipping charges are horrendous if you only buy one or two chips.

[Mark Hammer]

I think this is a job for Small Bear!!

[markseel]

Here's a link to the specs (found via Google):
www.profusionplc.com/images/data%20sheets/nemfx-sc.pdf
The specs state a 60 mA current draw - pretty high for an effects pedal running off of a battery.
Looks really easy to use.  Why a 28-pin IC!?  Only 14 or so pins are used.
Anyone know how it's implemented (BBD's or ADC/Delay/DAC)?

[markseel]

I read the spec and it looks like it's a custom digital effect:

Nemesis Technology, Inc. is proud to offer the SC ("Single-Chip") Reverb Version of our
popular NemFX Series Digital Multi-Effects solutions.  The NemFX SC Reverb is an
extremely low cost unit that is designed for applications where price is a driving
factor.  Although it is ideally suited for entry-level products such as practice
amplifiers, its sophisticated sound quality enables the NemFX SC Reverb to be used in
higher applications.  The NemFX SC Reverb features a high-quality, sweet-sounding Spring
Reverb program.

I was actually hoping they put a tiny spring in that big IC   

[cloudscapes]

is it me or are there no control inputs on the chip?

[markseel]

I don't see any.  Strange. 

[armstrom]

Not really strange. Think of this as a digital equivalent of a belton module. I'm assuming this does not provide wet/dry signal mixing in the chip so it would be up to the implementer to mix the resulting wet signal back in with the dry.  I agree that it would make sense to provide a control parameter for the delay time, but who knows if that is even possible with the hardware implemented on on the chip. Any sort of analog control like that would require an additional A/D converter channel, the microcontroller would need to sample that A/D line and adjust the program on the fly. Considering this appears to be custom silicon the designer probably just wanted to keep things simple to keep costs down. When you consider all you really need to support this chip is a quad op amp, a handful of passive components and a regulated 3.3 supply it's pretty easy to integrate (although, so is the belton module, this is just smaller)
-Matt

[markseel]

OK, that makes some sense.  That does make it easy.

Looking at the schematic for the output circuitry; there's an op amp that seems to be summing the differential outputs of the reverb chip.
Perhaps that's where wet/dry signals are combined.  If so then the resisters could be used to change the wet/dry mixing.

[cloudscapes]

still would of been nice to have at least decay time.

guess that could be "faked" by varying the wet reverb only feedbacking into itself.
the wet/dry and tone control can easily be done analog
I might pick up a ocuple of these chips to see what can be done with. not nearly as flexible as the spinsemi, but for those of us not yet ready to learn a language, could be a stepping stone i nthe meantime.

[MetalGuy]

The specs state a 60 mA current draw - pretty high for an effects pedal running off of a battery.

Who cares about batteries these days anyway!

And uhhhhh.... 13 pins (almost half) unused??

[markseel]

Regarding the batteries; was that sarcasm?
The unused pins allows for a larger chip to make room for a bigger spring.

[.Mike]

The company that makes the chip doesn't list any US distributors and their site says to contact them for US sales, so I did just that.

I asked them if they would sell a small amount, or maybe even donate a handful of samples to the folks here at the forum.

I'll let you know if I hear back.

MIke

[markseel]

Would anyone be interested in some small PCB's?
If so, I'll whip one up with the analog front and back end shown in the spec sheet.
It costs me about $100 for 10 or 20 PCB's (3 to 4 week manufacturing time).
If I had confirmation in this forum for 10 or more I'll get them made and send them out for $10 each.
I can populate the PCB's with parts (excluding the reverb chip) for a little more $.
Or if people prefer I could just hand over the artwork for the PCB's for free and you can have them made yourself.
Thoughts?

[AdamM]

That wasn't on the site last time I looked a couple of weeks ago, so it must be very new as they haven't got a datasheet up yet. I suspect it's going to sound "adequate" - it is intended for practice amps after all.

Be aware that Profusion's shipping charges are horrendous if you only buy one or two chips.

I suspect "adequate" may be similar to or better (whatever that means) than a real spring even if its not up to Spin FV1 standard

And I can stretch to £3.50 shipping for the sake of experimentation

And like the Belton, or a real spring, the time is fixed - so what? Simple is best, right? I reckoned maybe an input ("dwell"?) pot, a tone pot, and a mix pot ought to be do-able in the analogue section of the circuit. Maybe even do the analogue parts using sub-mini tubes...but probably op-amps to start off with.

[Taylor]

I might pick up a ocuple of these chips to see what can be done with. not nearly as flexible as the spinsemi, but for those of us not yet ready to learn a language, could be a stepping stone i nthe meantime.

The Spin chip comes with 2 nice verb programs built in, which have 3 controls each. So you can jump right in with the built-in programs without having to do any coding.

However, the Spin language is pretty easy to learn IMO. I'd never done any programming until recently, and now, though I don't claim to be a guru, I'm pretty comfortable with it. It took me way less time to learn to code for the FV1 than to figure out what impedance really is.    And I still can't design a transistor boost from scratch.

[cloudscapes]

However, the Spin language is pretty easy to learn IMO. I'd never done any programming until recently, and now, though I don't claim to be a guru, I'm pretty comfortable with it. It took me way less time to learn to code for the FV1 than to figure out what impedance really is.    And I still can't design a transistor boost from scratch.

me neither, NOR do I still really know about impedance.

before the year is out, I'm very likely to own one of those little spinsemi boards.