# Solid State Spring Tank Reverb Circuit

Discussion in 'Amp Tech Center' started by robrob, Feb 24, 2016.

1. ### Ten OverTele-Holic

Posts:
734
Joined:
May 13, 2015
Location:
Central California
C13 (470pf) and R32 (1M) form the classic Fender high pass filter that I mentioned above. In simplistic terms, it has a cutoff frequency of 339Hz so it cuts frequencies below that and passes frequencies above that. You don't want those lower frequencies because they just muddle-up the sound of the reverb.

The gain of U2A (the driver opamp) is one plus the ratio of R33 (4.7K) in parallel with the input coil to R34 (47) in series with C14 (22uf). For the frequencies that we are using, the impedance of C14 is small enough to ignore. So we can calculate the impedance of the input coil in parallel with 4.7K then divide that by 47 and add 1 to find the voltage gain. (The inductance of the input coil can be calculated from the impedance at 1KHz and is 95.5mH, needed for the calculations)

For simplicity, we will ignore the 4.7K for now below 2KHz because the impedance of the coil is small compared to it. So at 500Hz, the impedance of the coil is 300 ohms and the gain is (300 /47) + 1 = 7.4 . At 1000Hz the coil is 600 ohms and the gain is (600/47) + 1 = 13.8 . The impedance of the coil doubled from 500Hz to 1000Hz but the gain almost doubled, also. So the current in both cases is very similar.

As the frequencies get higher, the significance of the 4.7K increases and the gain does not increase as much per change in frequency. Eventually it will max out at a voltage gain of 101. Not only do you not want lower frequencies, but you also don't want higher frequencies, so this works out well. Also, the opamp would reach its limit if gain kept increasing in a linear fashion. The 4.7K resistor also prevents the gain from going to infinity if the tank is disconnected.

The driver is a constant-current device for the frequencies of interest.

2. ### mRtINYFriend of Leo's

Posts:
2,187
Joined:
May 7, 2015
Location:
Orygun
I checked my calculations again... Must have lopped off 3 orders of magnitude somewhere. You are correct about the turn-over frequency at about 1/3 of a KILOHETRZ.

It's the connection to the driver that isn't clear. You are telling me that the second connection (shown as a shield, here) is actually isolated from the Pan and connects to the low side of the driver. Right?

...that will change the behavior significantly. And will do a reasonable job of mimicking a current-mode driver.

So where does the drip go in most SS Fender reverbs? Is it the higher-impedance driver (600 ohms instead of 8)?

-tINY

3. ### Ten OverTele-Holic

Posts:
734
Joined:
May 13, 2015
Location:
Central California
The input coil of the reverb tank forms part of the NFB loop, so neither end can be grounded. The tip part of the connecting cable hooks the output of U2A to one end of the coil. The other end of the coil hooks to the shield side of the cable which connects to the junction of R33 and R34. This causes a lot of trouble for people who ground the cable through one of several methods. The phono jack on the input of the tank must be isolated from the tank so that the cable shield does not get grounded by the tank itself.

I don't know what you mean by "drip".

Last edited: Feb 26, 2016
4. ### printer2Poster Extraordinaire

Posts:
7,180
Joined:
May 24, 2010
Location:
Rob, you are thinking full wave from all the tube stuff you are used to seeing. The page I linked to does give the correct information, your just not used to thinking that way yet. Maybe Google some SS PS's, voltage doublers with transformers with center taps. One of the most widely used application of the circuit is for PC power supplies to be used on 120/240V.

Ten Over, nice of you to drop by, seems like you know your stuff (glad to see, oh no wonder, your an old fart also. Probably cut your teeth on tubes also). Not saying that you would want to use a couple of zeners to get into voltage range off a bias winding. Seem to recall we only need 5mA or something similar for a high impedance tank. Could be wrong mind you, don't have the energy to find out how much we need. (Been a rough day, week, last three years, long story but part why I took a hiatus from here). Wouldn't have done it that way anyway. Would rather steal from the heater supply, nah, just use a separate supply.

Would like to get further in this circuit along with a few others I was hoping to work on but just tired. Maybe later. Didn't know this little circuit would have been on anyone's radar but mine. I was going to do a digital reverb (sort of digital, bucket brigade) but I just did not like the sound of any I heard. Sure you guys will sort things all out, carry on nicely.

5. ### Ten OverTele-Holic

Posts:
734
Joined:
May 13, 2015
Location:
Central California
Accutronics published a saturation current of 3.1ma rms for their 4E tank. The driver under consideration here can easily deliver twice that. But the AC current delivered to the coil and the DC current drawn by the opamp are two different things.

What really determines the DC current draw is the power supply itself. Like I said before, the power supply is designed so that it draws the same current whether it is dumping it or the opamps are consuming it. My power supply drops 7.5V across the 470 ohm resistor which is 16ma. This is more than the opamps will consume under the worst case scenario, but not a whole lot more. The HRD supply drops 32V across 470 ohms which is 68ma. This is a whole lot more than needed.

The Hot Rod Deluxe or the Hot Rod DeVille use a bunch of opamps and may need that kind of DC current. The Blues Junior only has two opamps both used for reverb and it drops 11.7V across 750 ohms which is 15.6ma. Their supply and my supply perform about the same as far as voltage and current are concerned, but we sure get there in different ways.

6. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,301
Joined:
Dec 29, 2012
Location:
United States
Here's an LTSpice simulation of the bias tap power supply:

Green trace shows + output voltage of +16v at node 2 (upper right of schematic). Blue trace shows - output voltage of -16v at node 3. Red trace shows current through resistor R78.

Last edited: Feb 28, 2016
7. ### Ten OverTele-Holic

Posts:
734
Joined:
May 13, 2015
Location:
Central California
Half wave and using a big chunk of current just like I said.

What happens if you change R78 and R79 to 2K?

8. ### mRtINYFriend of Leo's

Posts:
2,187
Joined:
May 7, 2015
Location:
Orygun
It's that "surfy" sound you get when hitting the string hard with a partial palm mute.... The newer Fender amps don't sound quite the same.

-tINY

9. ### Ten OverTele-Holic

Posts:
734
Joined:
May 13, 2015
Location:
Central California
I heard that Dick got his reverb from an old water tank out behind the studio.

10. ### mRtINYFriend of Leo's

Posts:
2,187
Joined:
May 7, 2015
Location:
Orygun
OK. But, the older (early 60's) reverbs had a different response to transients. Even my Boss pedal (modeled 6g13) gets the drip.

I'm just trying to figure out what the difference is - it may be the 300Hz HPF.

-tINY

11. ### Ten OverTele-Holic

Posts:
734
Joined:
May 13, 2015
Location:
Central California
Perhaps Duane.

12. ### mRtINYFriend of Leo's

Posts:
2,187
Joined:
May 7, 2015
Location:
Orygun
Well, I got the drip from my Supersonic 22... I had to turn up both tone controls quite a bit. So, I'm guessing that it's a drive-level issue.....

Shortly afterward, it broke into a loud, 120Hz buzz and started smelling burnt. So she must not like it. We'll see if its a factory defect (amp has about 10 hours on it...).

-tINY

IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.