Looking for input for my OpAmp reverb schematic???

[biggy boy]

Hi
I took the very small reverb tank out of my old Univox amp and want to reuse the tank with a new board. I didn't care for the amp, but love the reverb!!
I have the schematic for the old one and could build that circiut again. The old circuit ran at 27 volts or so and the new board I would make would run at 9 or 12 volts. But I don't know how to set the bias for the transistors for the new voltage. My other option was to make an OpAmp board to drive the tank.

So what I'm after is some pointers (help) with:
A) getting the transistors biased correctly at 9 -12 volts if I deside to rebuild the original circiut.
B) look at the schematic I made with the OPAmp to see if it would work.

Here's the original schematic section for the reverb




And this is the new OpAmp schematic

[sean k]

Okay, I've done this once before and lost it so I won't ramble this time.

This is a good circuit.
http://www.solorb.com/elect/musiccirc/reverb2/index.html

This explains how to make a really good circuit.
http://members.tripod.com/~roymal/reverb.htm

And this one allows you to measure the DC resistance of the input and output transducers and get some idea about what the impedances are.http://www.accutronicsreverb.com/ioic.htm

With all those resistors under 1k and 27V supply on the pnp drive tranny I'm thinking your input impedance may be in the region of 10 - 300 ohms so your voltage amp ain't gonna drive it with the current it needs. Do some measurements and compare them to the accutronics website and see what you get.

[sean k]

I've got a 8FB---- accutronics tank which means my input imp is 1925ohms, 2k, and the output imp is 2575ohms, 2.5k, so you can see I need 2mA to drive it whereas another reverb tank I've got from an old hammond organ, and its an accutronics tank too, has 11ohms DC resistance on the input and 1678ohms DC resistance on the output. From the accutronics figures for DC resistance I can only guess what the impedances are but most probably about a 100ohms imp on the input, which would require about 15mA to drive, and maybe 25k on the output.

I've also got a type 4 hammond tank and the DC resistance accross each winding is 1700k so I suppose this ones 25k impedance at both ends which means its going to take bugger all current to drive it but a fair whack of voltage. maybe something like you've designed.

Heres the driver and pickup circuit for the hotrod delux


Which is halfway down this page
http://studentweb.eku.edu/justin_holton/reverb.html

[biggy boy]

Thanks Sean for the great links.
I'll measure the resistance of the two coils tomorrow and take a picture of the tank.
and post the info tomorrow.

Glen

[biggy boy]

OK got some numbers the input coil resistance is 168 ohms and the Inductance is 3.2 mH
But I do not trust that  inductance value .f 3.2mH or the meter that read that value.
The ohm reading will be accurate.

The output coil is 746 ohms and the L= 2.2 mH

Here's the picture of the tank! It's small isn't it? Looks like it just got out of a cold shower








Glen

[sean k]

Can you do math Glen? Those calculations at the top of the page of reverb2 basically look like you can derive the impedance for a specific frequency but I have no idea what j is and how it's calculated. Is it j (?) x (2 x pi) x frequency which I'm guessing is ((j x 6.8k) x Inductance) + DC resistance for an impedance at 1k. I suppose that if I measure an inductance accross the coil with a known impedance then I could backtrack or reverse engineer the calculation to get the value of j.

My accutronics tank has a coil at 2575ohms Z, 200ohms DC R and 493mH L

Backwards 2575 - 200 = 2375 / .493 = 4817. 4817/ 6800 = .708. .708/1 = 1.412 so I'm guessing that j is the figure 1.414 which has something to do with sine waves and average energy content. or RMS anyways.

I have a set of figures here for an unknown impedance which has a DC resistance of 1.7k and L of 283mH.

The j being 1.414 I could simplify the equation to 1.414 x 6.8k to 9.612k

so 9600 x .283 = 2.718k + 1700 = 4.4k Zohms which is the input Z

and 9.6k x .409 = 3.926k + 1700 = 5.6k Zohms which is the output Z

My other tank has input 11ohms DC and 66mH L, output 1678ohms DC and 418mH L ( close enough to the other one so its 5.6k Z)

9.6k x .066 = 633 +11 = 644 Zohms which is input Z

Hopefully some of the regulars who know about this stuff will check whether my calcs are right though they seem to be in regions usual for the useage.  

I quite like this and hope that I'm right though so I'll do yours.

9.6k x .0032 = 30.72 + 168 = 200ohms Z
9.6k x .0022 = 21.12 + 743 = 764ohms Z

What I gather from this is that whe're both kinda in the territory where dynamic mics operate and line level sources are at home.

[sean k]

I reread some stuff within the reverb2 page and found an inductance for a 8D tank of 44mH so doing my calculation I didn't get the same figure as that found in the accutronics site. They had the impedance as 310ohms Z and I got 458ohms Z which misses by a factor of 50% I wonder if J is a reciprocal of something or some sort of phase shift differential 

But then again maybe my ballpark figures are close enough!

[sean k]

Did some looking into it and j is the square root of -1 and I must admit that when I first saw j something about imaginary numbers came to me but I disregarded it because I definitely don't want to go there. But on reading some simpler stuff reactance came up about current lagging behind voltage where inductance was concerned and being about tangents. This stuff is way above my head but I'm not in it for exact details so much as ball park figures that can get me to the place I want to be.

I guess thats why the reciprocal entered my thinking and I'm kinda thinking theres a ballpark relationship between inductance and Dc resistance that divides frequency to get j. I'm still just thinking in relationships but it reminds me of logarithms and bell curves and I suppose thats where the time thing comes in with acceration and decelleration... thrown in where the relationship of inductance and resistance defines the curve and frequency defines a tangent that'll meet the curve at a point that then defines what j is and that number is somewhere between .707 and 1.414. Nah it's simply beyond me. 

[biggy boy]

Nah it's simply beyond me. 

Me too 
                                                                              V
Ya I got all excited when I saw that formula  I = ------          Until I had to figure what "s " was, then I gave up 
                                                                          sL + R   

Math is not my strong point.
I thing I'm just going to have to make a circuit up a fiddle with it and try stuff out I'll use the layout that is used from the link you posted.
Only problem is I got four other boards on the go, plus an upright electric bass to work
on and a 82 fender to finish refurbishing.

Thanks very much for your input Sean it is greatly appreciated.
Keep me posted about your findings please.
                                                                                                                     

[Nasse]

I just tried my short tank from my past and just hooked the tank output straight to my mixer and rolled off all lows, still sounded mostly muddy.

Don´t have any input EQ, just connected a small diy headphone amp (71 opamp beefed up with transistor pair) to L output of my small modelling preamp, and took R output to other mixer channel, that headphone amp driving the reverb tank

With reverb channel turned up, nothing special to write home, muddy and not so inspiring sounding short reverb, like small room or empty steel barrel. Listened some music on only reverb channel and it was like recorded in the other end of a hall fx

Then I remembered that my mixer does have some crappy DSP fx processor built in and switched the fx processor on and turned the reverb spring channels fx level on, got some nice special fx sounds and it started really happening, tried fx like chorus phaser and pitch shift on the reverb signal, was pleased with results

My tank is not same as yours, size is similar but input impedance is very low, ouput is in the same class