What would sound better?
An el-cheapo spring reverb tank such as:
Spring Reverb Tank Electric Guitar Amplifier 2 Spring Medium Decay TPSB2EB2C1B 603803162274 | eBay
or a reasonable quality digital reverb such as:
Accutronics Belton Reverb Module - Adjustable Decay BTDR-3 | eBay
Presumably the reverb tank would be harder to splice into a stock non-reverb Princeton circuit?
An el-cheapo spring reverb tank such as:
Spring Reverb Tank Electric Guitar Amplifier 2 Spring Medium Decay TPSB2EB2C1B 603803162274 | eBay
or a reasonable quality digital reverb such as:
Accutronics Belton Reverb Module - Adjustable Decay BTDR-3 | eBay
Presumably the reverb tank would be harder to splice into a stock non-reverb Princeton circuit?
That reverb tank needs a drive and recovery/mixing stage to interface with your current amplifier, the tank by itself is unusable.
I suggest you look at a modern Fender amp,say a Hot Rod Deluxe or Deville or similar, and copy how they add the reverb.
You will basically need to make a miniboard with a TL072.
You will also need +/-15V to feed it.
Post both schematics so we can suggest how to "Frankenstein" the HRD reverb into the Princeton.
The Belton brick again needs some interface circuitry, it can not be used "alone", plus you need 5V to feed it.
Plus +/-15V to feed any Op Amps involved.
Both sound good, the spring is the original way, an the brick does a good job at imitating it.
If needed, *there is* an old Popular Electronics Reverb circuit from late 60´s , no ICs , which can be fed from existing Princeton supply and sounds surprisingly good.
Search the Forum, there was a thread about it not long ago.
I suggest you look at a modern Fender amp,say a Hot Rod Deluxe or Deville or similar, and copy how they add the reverb.
You will basically need to make a miniboard with a TL072.
You will also need +/-15V to feed it.
Post both schematics so we can suggest how to "Frankenstein" the HRD reverb into the Princeton.
The Belton brick again needs some interface circuitry, it can not be used "alone", plus you need 5V to feed it.
Plus +/-15V to feed any Op Amps involved.
Both sound good, the spring is the original way, an the brick does a good job at imitating it.
If needed, *there is* an old Popular Electronics Reverb circuit from late 60´s , no ICs , which can be fed from existing Princeton supply and sounds surprisingly good.
Search the Forum, there was a thread about it not long ago.
If you go with a spring reverb, the MOD tanks (MOD(R) Reverb Tanks | Mod Kits DIY) are consistently rated above the equivalent Accutronics tanks by most musicians. AND, they will give the "vintage" or "classic" reverb sound---IF that's what you are after.
Have a look at this thread in the Aussie Guitar forum.
Or use the Fender Champ 12 trick of driving the tank from the speaker output and then using a FET pulling power from the output tube bias cap.
Somewhere there's a decent sounding, single JFET "brick" thread but I can't find it right now.
Or use the Fender Champ 12 trick of driving the tank from the speaker output and then using a FET pulling power from the output tube bias cap.
Somewhere there's a decent sounding, single JFET "brick" thread but I can't find it right now.
Interesting stuff there, thanks.
I like the idea of the Fender Champ 12 method, I just need to figure out how I could interface that part of the circuit to a 5F2A.
I found a bit of info here regarding that schematic: Explain the Champ 12 | Telecaster Guitar Forum
Anybody have any thoughts about using an altronics kit like this to drive a tank:
New Spring Reverberation Unit - April 2017 - Silicon Chip Online
I'm not sure where the signal input for this would be tapped from or returned..?
New Spring Reverberation Unit - April 2017 - Silicon Chip Online
I'm not sure where the signal input for this would be tapped from or returned..?
I don't what your definition of "better" is, so that is unanswerable.
Real springs are traditional and give you "that" sound.
But the digital reverb is very good.
If your going to use real springs, make sure the input and output impedance of the tank matches the circuitry. If you don't, levels will be lower.
You can use a mixing desk (or daw with soundcard) aux send to drive the tank and use a mic preamp as return. It's not optimal, but works.
As a side note: A resonating mass spring system having high frequency content way above the resonating frequency doesn't make sense. Iow hifi and reverb tanks are two very different things.
Obviously it is a subjective question but more along the lines of 'can a digital reverb ever sound as good as a traditional spring reverb' than 'tell me what I like'.
Just in the same manner of questioning as: 'does a keyboard sound as good as a grand piano', does a spring reverb emulate well or does it sound synthetic when digitally reproduced?..
Out of interest, everything else being equal (cash, effort etc) and purely based on sound, if you had a choice between the two to apply to an old Princeton circuit (or Marshall if that's your flavour), which would you choose?
Thanks for the tip!
Last edited:
b
It's like comparing a phaser to a flanger. Duck vs Quail. Related but different.
Now, some of the wizzier digital reverbs have a "spring emulation" mode which is sort of the same, but why bother?
You might guess that for vintage amp rebuild I'd use a spring unit.
Serious "digital reverb" sounds like a small/medium/large room (depending on where the knob is pointing). It doesn't sound like a spring reverb. Nor a plate reverb, for that matter.
It's like comparing a phaser to a flanger. Duck vs Quail. Related but different.
Now, some of the wizzier digital reverbs have a "spring emulation" mode which is sort of the same, but why bother?
You might guess that for vintage amp rebuild I'd use a spring unit.
Digital reverb is much better..... when you have a close-mic recording in a small room and need it to sound more like a concert hall.
Spring reverb never sounds like a "room", it sounds like a spring, which IS the classic guitar effect. Some work has been done to emulate springs in digits but I don't recall seeing any small module which claimed that (and would have to try before I believed).
Spring reverb, you bump the cabinet, it goes "BWOOMMM!mmmmm...". No digi-mod does that.
So just build both. Neither system costs as much as a playable guitar, and takes less space to store.
Thanks for the advice and suggestions.
I'm going to go for the BTDR-3 (just $28 AUD) and build the circuit from the datasheet with a RC4558 with variable gain.
Would I break into the circuit at the input to the 12ax7 or would it be better to cut in at the output of the first stage of the 12ax7 before (or after) the tone controls?
I'm going to go for the BTDR-3 (just $28 AUD) and build the circuit from the datasheet with a RC4558 with variable gain.
Would I break into the circuit at the input to the 12ax7 or would it be better to cut in at the output of the first stage of the 12ax7 before (or after) the tone controls?
Attachments
Do it as late as possible, get your signal either from the output (per Fender Champ 12) or from the last 12AU7 and feed it back in to the grid of the 6V6
Use a decent op-amp and run it off as many volts as you can find. Or use a jfet or two.
The RC4558 was an alternate to a TL072 for a preamp circuit in another thread and I didn't hear any complaints from someone that knows way more than I can ever hope to understand:
I had searched for 'alternate to TL072' and that was marked as equivalent.
Part of choosing the RC4558 was the price - I don't have to worry about frying an $8.20 IC when I am using one that costs $0.66...
If the RC4558 (or TL072) really is that bad then perhaps I should just go with the OPA2134.
A few remarks.
I don't think anyone can hear it if you use a digital spring reverb, they are that good.
Spring reverb tanks are literally mass spring systems. This means that frequencies above resonance are subdued. Frequencies above 10k are non existent in spring reverb, except as noise. So using expensive opamps is not necessary.
Reverb before distortion is not going to sound musical often, so place it as far back in the circuit.
I don't think anyone can hear it if you use a digital spring reverb, they are that good.
Spring reverb tanks are literally mass spring systems. This means that frequencies above resonance are subdued. Frequencies above 10k are non existent in spring reverb, except as noise. So using expensive opamps is not necessary.
Reverb before distortion is not going to sound musical often, so place it as far back in the circuit.
Here's my two cents on this issue.
The RC4558 has bipolar junction transistors at the input; this means relatively low input impedance. It also has a relatively low slew rate; this means it struggles if asked to put out a large amplitude, high frequency sine wave.
But how bad are these problems, really? Should we spit upon the poor RC4558 and replace it with something much more modern? Throw holy water on it, shoot it with silver bullets, stab it through the heart with a silver dagger?
To find out, let's work out some numbers to see what the 4558 can and cannot do for us:
1) There is no useful e-guitar signal above, say, 5 kHz.
2) A little basic calculus says that the maximum amplitude (Vpeak) you can get from an op-amp without distorting the sine wave due to slew rate limitations is given by Vp_max = (dV/dt) / (2 pi f). dV/dt is the slew rate, which the datasheet says is 1.1 V/uS, worst case.
3) Sticking in a conservative dV/dt of 1000,000 volts/ second (the same as 1V/uS), and f = 5 kHz, we find Vp_max is about 32 volts. Maximum supply voltage limits this op-amp to less than half that voltage in any case; so the lowly RC4558 is actually fast enought to be capable of spitting out an undistorted 10 kHz signal at maximum output (roughly +/- 12 volts.)
So we see that, while the RC4558 is definitely not fast enough for Hi-Fi (where we expect a clean undistorted treble response to well above 20 kHz), it is actually more than fast enough for e-guitar.
How about noise? Specified at 8 nV/root hertz, it is equivalent to the thermal noise from a roughly 4 kilo ohm resistor at room temperature. The thermal noise from the DC resistance of a guitar's pickup windings generates more noise than this. The volume pot in the guitar generates even more noise, unless set to full volume. In other words, the 4558 has better noise performance than the guitar itself does.
Surprise! It turns out that the terribly scorned RC4558 is actually quite capable of handling guitar-bandwidth signals, and the noise performance is also more than adequate for e-guitar. However, this being a BJT input device, it is not really well suited to being fed directly from a high source impedance (more than a few kilo ohms.)
That may not be a problem at all in your application: you're going to have to divide down the signal from your preamp triode's anode anyway, as an op-amp will not cope with the hundred or more volt swing a 12AX7 can spit out. If you pick your divider resistances properly, you can end up with a low enough source impedance for an RC4558 to work very well.
All that being said, if you do decide to use a BJT, my suggestion is, just use an NE5532 and be done with it. It's faster and quieter than a 4558, Digikey USA charges only $1.36 USD for one, and more importantly, the 5532 is in every way more capable than the human ear, so if used properly, it is audibly flawless: human ears cannot hear any imperfections it produces.
In applications where a guitar is feeding an op-amp directly, the venerable TL072 is better than the 5532, because its JFET inputs produce a better noise figure at the high source impedances of an electric guitar.
There are surely some extremely demanding applications that might benefit from the various "new and improved" op-amps that came after the TL072 and NE5532. But for e-guitar? Those two really are all you need. They are the only ones in my parts-bin.
One last comment: a newer, faster op-amp is, because of its higher bandwidth, much more likely to burst into unwanted oscillation. PCB layout, shielding, and grounding will all become much more critical.
-Gnobuddy
Thanks Gnobuddy for (yet again) being so generous with your time to explain so thoroughly.
I now not only know that the 4558 will be perfectly adequate for the intended application (although the NE5532 would perform better) but also how to calculate the slew rate limitations of an IC!
Every little building block helps!
I now not only know that the 4558 will be perfectly adequate for the intended application (although the NE5532 would perform better) but also how to calculate the slew rate limitations of an IC!
Every little building block helps!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.