hi,
   I wanted to ask you what potential problems might I come into when "modding" the AD-3208 [from GGG] echo/delay.
Because [in Poland @ least] the mn3207[3.6$] are way cheaper than the mn3208 [13$] and I can't get the mn3205 anyway I've thought that I could put in oh, say 4 to 6 mn3207's instead of the 2 mn3208's. I also can't get the sa571, but I can get ne571 or ne572. Will this work? If so which one would be better - 571 or 572.

BTW - I've got the schematics for a echo/>>reverb<< on mn3207's, though it's a bit big [half of an A4 page] and in Polish. If there will be people interested I'll translate it and put it up on the net [where ?!?!?] when I'll have a bit of time on my hands.

But concering reverb. I'm thinking of making a tube P.A. with tube and ROG preamps [switchable] and I want to add a plate reverb, mainly because springs are big, and cost a bit, while I have a lot of metal cans, and a bit of time on my hands. The plate reverb was discussed here some time ago - any progress???

cheers

SA571 = NE571 = NE570, however the 572 is NOT the same.  It is a better compander chip in some ways, but you will not be able to use it without major changes to the design and PCB layout.

You CAN use multiple MN3207's (early Memory Man's and MXR Analog Delays used multiple SAD1024's) but you will need to provide the bias voltage to each one, and should probably use a trimpot to balance the two outputs from each 3207 more precisely.  It is usually okay to simply use two fixed resistors to do this, but my hunch is that when you multiply the possible imbalance from a pair of 5% resistors over 4 or 6 chips, there is a risk of sonic deterioration.    Better to have each chip perfectly balanced at its output.

Steve Daniels at Small Bear Electronics can probably provide you with BL3208's at a reasonable price, and ship it in a way that works best for you in Poland.  I would recommend simply using two BL3208's, rather than 4 3207's.  At the very least, you will not have to make your own board layout if you use 3208's.

I think I've seen the 3207-based echo/reverb schematic you mention.  It is essentially a slapback echo (about 80msec or less) with an option to mimic longer decay by increasing feedback.  Personally, I would NOT use this for "reverb" since the resonances created by such a short delay would be objectionable.

On the other hand, I've done some deep thinking about simulating reverb with *multiple* BBD-based short delays, and there are some exciting possibilities.  The key is that each BBD would need to have:
a) its own separate clock or some other means (e.g., a divided master clock shared across chips) to provide a different delay time than the other BBDs in the circuit
b) a mixer at its input and splitter at its output to provide multiple feedforward and feedback paths.

Plate smaller than springs?  I don't think so.  To have anywhere near the sort of bandwidth and decay time that even the cheapest 20cm spring has, a plate system would have to be at least something like 80cm x 80cm.  The plate units I have seen are all over about a metre squared.  Remember, springs can be more flexible because they are longer than they are wide and are only suspended at their ends.  Plates are much much stiffer because of their shape and because they need to be suspended from multiple points.  The only way to achieve the needed degreeof flexibility is to make them BIG.

thanx 4 the reply,
answered all of my questions and a bit more.
Yes, you've probably seen the schematic, the description is right.
That reverb idea of yours looks interesting. Maybe using two clocks a bit phase shifted at close frequencies would provide an interesting effect [like when you pluck two strings which are supposed to be at the same pitch, but aren't exactly]. Then you could mix the two clocks seperately for each chip! That might need a schmitt trigger, to get the signals even, but should be OK  :idea:
Also - I didn't know the plate needs to be BIG. I was thinking of feeding the sound from the mike back to the reverb speaker to get more reverb out of a smaller box, or was that incorporated in plate reverbs and they needed to be big anyway?

For plates, pick up a baking sheet (a flat pan about 30cm x 60cm) at the hardware store, hold it up so that it hangs down loosely, and tap it in the middle with your fingernail.  That will give you some idea of the relationship between size and possible decay times.  Making the plate out of thinner material helps, but you still need a large area to provide the needed compliance.

My suggestion about the multiple BBD reverb comes from the MN3011 and MN3214 chips.  These are quite rare now, but were intended to provide a cheap analog reverb simulation.  Both chips have multiple output taps that occur at mathematically unrelated intervals so that unpleasant resonances are not created.  

Normally, when these are used as reverb simulators, the multiple delay outputs are mixed together at a common output to simulate the multiple reflections of a room or space.  There are a few problems with this.  One is that for extending the decay time, all outputs MUST be mixed back to the same entry point.  It is a bit like standing with your back to a wall a few feet behind you and shouting at 4 or 5 walls in at different distances in front of you.  This is NOT the same thing as a "natural" reverberant space where the reflections bounce off all possible surfaces and return at increasingly diverse times.  In a sense, "real" reverb gets less systematic or orderly as time passes from the initial signal being produced.  With the multi-tap chips and common return point, the systematicness is not completely eliminated but is not deliberately counteracted either.

Having multiple BBDs with simultaneous variable feedforward and feedback means that one can vary the "flavour" of reflections.  You can do the math yourself, but imagine what you would be able to get after, say 500msecs of recirculated decay if you had 3 BBDs in series, set for delays of 50, 35, and 20msec, and could feed the output of each BBD both back to the start or back to earlier BBDs (e.g., from the output of the 20msec stage to the input of the 30msec) AND feed each output ahead to two chips down or to the output.  In very short order you start to have an extremely diverse number of delay/reflection times.

Athin: watch out with 2 clock-frequencies within an audio-range distance  :wink:

this is getting more and more interesting. 10 ms ago I had a brilliant idea - let's take 5000 transistors and 5000 caps, and make the BBD's !!! I mean it would be HUGE, the reverb out of that might get you to places sound hasn't been before... I've even developed the technology - half a glass of caps, half a glass of transistors, mix with molten solder, not too much, leave to cool, enjoy
It would also be a great way of getting your freezing beer just-so-cool when giging outdoors. [just like PC in the old days]
Now back to reality:
looks to me that the best way to get good reverb 'out of a [small or reasonably sized] box' would be to participate in the DIYDSP project Peter Snowberg started.
Since I haven't started assemling the amp yet, I have the comfort of striking a part of the plan and replacing it with another without the extra work  After reconsidering, a spring reverb, 100% custom [B4 you start asking why - because the whole amp will be custom made where possible [yup - even the trannies], just to see if I can, and if a little love can out-sound sophisticated technical processes :wink:  ] might be the best choice.
One more question - I've got a 'C.MOS Echo schematic', a real simple one. I don't quite get how it works, but probably doesn't give a lot of delay time. Is it possible to mod it to reverb?
I don't know where I got it from, so I'll give a link form my page:
http://kosi.website.pl/cmosecho.jpg
Edit : copy and paste the link

You got it from my site. It isn't a echo, is really a tremolo. In the original article, this was called a "repeater", so I thinked that it was an echo, but somebody told me that this is a tremolo.

I've built that and it's definately a tremolo. It's very choppy, on/off so if you pick a note and sustain it it sounds a bit like the note is echoing.

That explains why I couldn't understand how the echo is formed  :lol:

Many years ago I accidentally produced similar tremolo circuit like "c.mos echo" at Marcos´s pages (was testing bad leslie simulator circuit from a book). With clean guitar notes I had to double-check that my delay pedal was not connected... of course it was just pseudo echo but it was funny feeling when I noticed that the delay pedal was not plugged in...

But slightly tremoloing your guitar signal before feeding it to delay with some feedback/repeat it may sound more complex, more like multihead tape echo

I remember many years ago before reasonable priced digital processors local dance bands mostly used only slight delay in vocal mic, (mostly tape echos, those Dynacord echos with one playback head were popular) and it sounded just very good with real room acoustics in dance halls. Was the PT2399´s original use in Karaoke machines just for producing delay like that...

Quote from: Athinhi,
   I wanted to ask you what potential problems might I come into when "modding" the AD-3208 [from GGG] echo/delay.
Because [in Poland @ least] the mn3207[3.6$] are way cheaper than the mn3208 [13$] and I can't get the mn3205 anyway I've thought that I could put in oh, say 4 to 6 mn3207's instead of the 2 mn3208's. I also can't get the sa571, but I can get ne571 or ne572. Will this work? If so which one would be better - 571 or 572.

BTW - I've got the schematics for a echo/>>reverb<< on mn3207's, though it's a bit big [half of an A4 page] and in Polish. If there will be people interested I'll translate it and put it up on the net [where ?!?!?] when I'll have a bit of time on my hands.

But concering reverb. I'm thinking of making a tube P.A. with tube and ROG preamps [switchable] and I want to add a plate reverb, mainly because springs are big, and cost a bit, while I have a lot of metal cans, and a bit of time on my hands. The plate reverb was discussed here some time ago - any progress???

cheers

hello !!
I put together a AD3208 from GGG with 02 mn3205, but I cant find them anymore, so I thought I could use 4x BL3208 to get the same 600ms... do you thinks this is possible ??
do you have  another idea??

thanks !

very well possible, but note that each BBD needs its bias, and DC-blocking at its output/input (one big company put 8 BBDs in series, recently...)
- this should also answer your PM -

sorry: I`m a bit holiday-handicapped  

Quote from: puretubevery well possible, but note that each BBD needs its bias, and DC-blocking at its output/input (one big company put 8 BBDs in series, recently...)
- this should also answer your PM -

sorry: I`m a bit holiday-handicapped  

cool!!!I



8x BBD's ?

should I use 2x MN3102 too?
if you want you can take a look on the project

http://www.generalguitargadgets.com/diagrams/ad3208_schematic.pdf
thanks man !!!!!!!!