Differences in so-called ground voltages between different c

Keith Hilton · Post by **Keith Hilton** » 14 Feb 2000 5:06 pm

Today most guys have one or more AC-powered effect units, etc. in line. The difference in so-called ground voltages between chassis can be as much as the line voltage, and can zap inputs and outputs. AC units in line can be turned off ,or on, in different sequence. This can cause a sudden rush of current in or out, depending on the sequence. This rush of current can go to units not yet turned off, or to others that are off. This current can be either AC or DC. The loss of power to one unit can send either a plus, minus ,or both to other units in line.
Input and output protection can take the form of resistors, capacitors and diode clamping systems. "ARGUABLY"--some people don't use protection, or very much, and claim that it creates a better sound. The degree of protection ,and the method of protection, of input and outputs depends on the manufacturer. Actually it becomes a balancing act between protection and sound. Sound quality will be effected by many protection devices.
The problem builders like me have to face is the public plugging in anything! Having many devices hooked up in unlimited ways. Turning all this stuff off and on in random patterns. As a example, I had to make my power inputs where they would take AC or DC, any voltage up to 40volts,and any tip configuration. With musicians in a hurry, plugging in, in the dark, making mistakes, and drunk--I had to make my power inputs fool proof, where just about anything worked. A different problem exists with inputs and outputs, since protection can change tone. It is difficult where to draw the line! Do you make it sound better, and leave yourself exposed? Do you provide ultimate protection at the expense of sound quality? You must balance protection with sound quality, that's my opinion.
I would like some opinions on the following: Differential Input Breakdown,Common Mode Input Breakdown,Latch-up,Output Short-Circuit Protection, AC coupling of inputs and outputs with capacitors, clamping diodes on inputs and outputs, built in and built out protection resistors. All these can effect tone and performance. I'm not worried about supply voltage protection, so none of that please.
I would just like some opinions--please give some values----like .1uF or 10K ETC.
We are talking about sound quality-tone, and equipment failure. Then there is something no one talks about--SOFT FAILURE!
That's where something is damaged, but still works. It works, but the sound quality is not as good. This can happen very easy with NPN or PNP junctions!
I don't want to get out in left field, but after all this is the Electronics section of the FORUM. Keith Hilton

------------------
[This message was edited by Keith Hilton on 14 February 2000 at 05:08 PM.][This message was edited by Keith Hilton on 14 February 2000 at 05:09 PM.][This message was edited by Keith Hilton on 14 February 2000 at 05:10 PM.]

Blake Hawkins · Post by **Blake Hawkins** » 14 Feb 2000 8:30 pm

Keith...a most difficult subject.
Over the years, I've favored transformers for isolation. However, that is an expensive option, and getting quality audio transformers made is difficult.
Capacitor isolation is inexpensive, but affects the tone quality.
Differential inputs give good hum rejection...if you can't find an IC that is heavy enough, then building them from transistors is an option.
As for protection...given all the things that can go wrong...600v would be a good target voltage and should handle most all situations.
Complete protection of input signal lines may require several solutions.
That is, common mode rejection to a certain point with a shutdown trigger above that.
As you have pointed out, simple solutions such as zener diodes and caps will adversely affect the tone.
Output lines....a DC sense circuit and crowbar to protect speakers. Perhaps some sort of trigger circuit could be used to shut it down if connected to the wrong load or to a voltage source.
Just a few thoughts.
Good luck...
Blake

Steve Feldman · Post by **Steve Feldman** » 14 Feb 2000 9:11 pm

Then there is something no one talks about--SOFT FAILURE!

Please! This is a family Forum. Nobody likes to talk about E.D., although I hear they now have medication for that sort of thing...
SF

Keith Hilton · Post by **Keith Hilton** » 14 Feb 2000 9:51 pm

Blake, I figured this topic might get a response out of you. I agree that differential configurations seem to have less hum. I have not used them much lately, more inverting and non-inverting.
Steve, that wasn't the Soft Failure I was talking about. LOL
I forgot to mention temporary or sustained input or output short circuit. This is when your cord tip gets against a piece of metal. Kind of reminds me of the guy building a really fast hot rod motor. He could build the motor where it had 50 times the normal horsepower. The motor would out run anything in the state. Trouble is--how long would that motor last? Guys like me have to guarantee what we build. Do any of you guys have any suggestions on clamping diodes for audio inputs and outputs on OP Amps? The low leakage type is what is needed. How about a lN 458 or lN4148? Hartley Peavey told me to use a lN2004, but I could never find those. Shoot me some numbers of clamping diodes you like. Ones that won't bleed noise into the audio signal when clamping inputs and outputs to power sources.

------------------

Curt Olsen · Post by **Curt Olsen** » 15 Feb 2000 4:46 am

I personally use schottky diodes to protect inputs, you can obtain them with fast responce and yet soft recovery. The 4148's are an old type device (same as a 914) that are not used that much anymore. You might look into a newer way of clamping the inputs, MOSFET's that limit excessive levels. You would have to design them into your circuit, but would really protect the inputs.I really think inputs should be able to deal with several hundreds of volts to be effective.

Mark Amundson · Post by **Mark Amundson** » 15 Feb 2000 6:38 am

The venerable .022uf 600v safety caps and series 68k ohm input resistors (ala Leo Fender's amps) is the best I can think of. The caps do the DC blocking and the 68k keeps the input alternating currents from causing much damage; other than pushing op-amps or transistors one diode drop past the supply rails. My biggest "soap box" item is to make sure all inputs are one meg-ohm to keep from sucking all the tone out of pickups.

For battery operated stuff, I tend to go with discrete transistors over op-amps to reduce every micro-amp of battery draw possible. Also a series diode in the battery positive wire to prevent accidental flipping of 9v battery posts from blowing out the circuitry. Lately, every stomp box I make, I try to avoid AC adapter (wall-wart) jacks so that no one gets the big idea to share a wall-wart among multiple boxes.

Keith Hilton · Post by **Keith Hilton** » 15 Feb 2000 8:26 am

Mark, I'm like you, I like to keep impedance high to keep from milking pickups. Using the right OP amp gives you the high input impedance. Didn't know Leo Fender used 68K input resistors. I usually use 10K. Are you saying Leo Fender used a lot of .022uF to block DC on the inputs?
The big problem I have with clamping diodes is "NOISE". Even with the low leakage types.

------------------

Mark Amundson · Post by **Mark Amundson** » 15 Feb 2000 9:17 am

Keith, Leo only used the 68k/1meg network without any caps straight to the first grid. The 68k was originally for level shifting on the second input as each input had a 68k resistor in place (have to see the schematics to understand why). But accidently, the 68k became a grid stopper (or grid leak) for modest overvoltages. The first triode in his amps became a grid-to-cathode spark gap for ESD. The solid-state amp manufacturers could not endure an ESD hit in the input jack in the same way, so blocking caps are necessary. Pretty easy to build up a + or - 15 kilo-volt static charge on your body and then insert your instrument's cable into an input jack. A reasonable sized blocking cap will absorb those picofarad sized charges without punch-through of the cap's dielectric.

The 0.022uf sized capacitor (cap) is the value used Leo used on interstage decoupling when the input was one meg-ohm on bass guitar amps. I prefer Mylars on solid-state stuff and polypropylene on tube stuff.

Blake Hawkins · Post by **Blake Hawkins** » 15 Feb 2000 12:10 pm

Mark, and Keith,

I've always thought Fender's inputs were a bit strange.
I have a 1953, Fender 5B3 Deluxe. The input jacks short to ground when nothing is plugged in.
There are 75K resistors in series with each input.
When you plug only one guitar in...it is working into only 75K!
Usually in the older mixing circuits the primary reason for the two series resistors was to isolate the two inputs from each other. You will see that type of circuit in
many old mixers.
Fender's use of the grounding jacks is unique. Of course, grounding unused inputs
kills any stray noise but it also affects the tone of the amp.
The inputs in the Twin-Reverb are similar.
Does either of you know why Leo Fender chose to do it this way?
Thanx
Blake

Keith Hilton · Post by **Keith Hilton** » 15 Feb 2000 5:41 pm

Actually input impedance depends on much more than just the audio input resistor. It also depends on what happens after the resistor also. I'm not much good on the old tube stuff, even though Hartley Peavey has sent me a book on that very subject.
As we all know, input impedance must be very high or it will milk the pickups. I like to use a 10K input resistor, then the audio signal sees the impedance of my OP amp which is 10to the 12th power II 10. I don't have electronic symbols, but you guys know what I mean. It is very high input impedance. Even with ESD--Electric Static Discharge, and other transient currents, if the input resistor is high enough, I see no need for clamping with diodes. Especially with a capacitor. I am fond of a .1uf cap. on the audio input. Hartley Peavey keeps telling me I need clamps on the input to the power. I really don't like these clamping diodes, because I can hear noise, even out of the low leakage types.
We still have not talked about the output protection of some in-line devices. Here is what I like; I like 2.2uf capacitor, any smaller attenuates the lows. I like a 2.2k output resistor. I also like a 100K resitor to ground on the output--"SOMETIMES". Why? If you plug into another device that has capacitors built up, the 100K resistor to ground can really lesson the "POP" you can hear. Again, Hartley Peavey thinks I should put clamping diodes on the audio out, to the power supplies. Again, I don't like it because I hear noise.
Mark really hit the nail on the head when he talked about ESD Electric Static Discharge. + or - 15,000 volts of body static by just walking across a carpet really tells it all.

------------------
[This message was edited by Keith Hilton on 15 February 2000 at 05:44 PM.]

Mark Amundson · Post by **Mark Amundson** » 15 Feb 2000 8:02 pm

A couple of technical points to keiths last posting:

Regarding vacuum tube input impedances, unlike junction semiconductors, tube signal grid impedance is dominated by the grid to cathode capacitance (a couple of picofards) with almost unmeasureable resistance. The exception to this would be when the grid is forward biased (when overdriven), it becomes like a second plate and exhibits a rectifier characteristic. Two must have books tube amplifier gurus are: The Radiotron Designer's Handbook (F. Langford-Smith 1953), and Principle of Electron Tubes (H.J. Reich 1941)

On output protection for in-line (guitar-to-amp) toys, a decoupling cap (around 10uf to 100uf) and series resistor from 47 to 1000 ohms are what I see in alot of pro-gear. Also, as you mentioned, a 100k to 1meg ohm anti-pop resistors are handy. I choose specific values depending on what I believe is the worst case input impedance to be driven, but still will survive shorted outputs. I see Hartley's rail protect diodes in his equipment, but you are right, they make good parasitic capacitors and noise sources. Anything beyond ESD hits that needs these diodes, should be sent to a service shop anyway for charred component inspection.

We're getting deep technical here, so I don't know much the readership can take before we have to do this off-line.

Blake Hawkins · Post by **Blake Hawkins** » 16 Feb 2000 6:46 am

On input protection: I may be a bit out of date on this as the last system I designed used 5534's.
I think most, if not all, modern IC OpAmps have zener diode protection built into the inputs.
These work pretty well but there is no easy way of testing them.
Keith, what Mark said about tube inputs is correct.
In practice, the tube input is bridging the input resistor (or resistor-cap combination)
and not drawing current from the source.
The pickup is supplying current to the resistor and on tube gear as well as transistor a low value of input resistor will cause lower level to the system.
Oh...I made a mistake in my previous post...the #1 input load is 150K instead of
75.
By the way, on the same amp, the "Microphone"
input does not have a load or isolating resistor on it. Only the coupling cap and the 5 megohm grid resistor and my steel guitar sounds MUCH better there.
Good Thread.
It's nice that you both take the time to build quality products.
Blake

SJ Russell · Post by **SJ Russell** » 16 Feb 2000 9:19 pm

I use a 4 plug outlet with 2 50 db filters and 2 75dbs filter. I installed isobar protector after I had a orange ball of electricty that traveled from where i has pluged in.It fried two power supplys
and three effects. Since that tine I have not had any problems. What I use is a computer power supplys these units can be bought from radio shack.