I have seen amp schematics with and without a capacitor right after the signal input. Usually around .1 uf
I assume that this is to stop any unwanted DC from elevating the grid of the first preamp tube and throwing it off bias. If this is the case, how come they can't be found on every amp? Are there adverse effects like filtering out low frequencies?
Thanks
I assume that this is to stop any unwanted DC from elevating the grid of the first preamp tube and throwing it off bias. If this is the case, how come they can't be found on every amp? Are there adverse effects like filtering out low frequencies?
Thanks
The best cap is no cap.. Yes, DC on the input will affect the operating point of the input stage, but IMHO this problem is best fixed at the source.
FWIW I haven't been using input caps for about 25yrs now, and have yet to hear of an instance where this has caused a problem.
FWIW I haven't been using input caps for about 25yrs now, and have yet to hear of an instance where this has caused a problem.
Filtering out low frequencies does not necessarily have an adverse effect, if it stops signals you can't hear from interfering with signals you can hear. Something will remove LF, and it may be better for a coupling cap to do it than the OPT.
If you know that all the sources you will ever use are incapable of emitting DC or subsonics then the input cap can be omitted.
If you know that all the sources you will ever use are incapable of emitting DC or subsonics then the input cap can be omitted.
I just was wondering about this as I was considering making a pan-able input to a two-channel amplifier.
and it reminded me of a protective measure recommended by a guitar shielding tutorial where they put a large value 600v capacitor in series with the guitar signal to prevent DC from coming from the amp to the guitar strings in case of a short or other catastrophy.
and it reminded me of a protective measure recommended by a guitar shielding tutorial where they put a large value 600v capacitor in series with the guitar signal to prevent DC from coming from the amp to the guitar strings in case of a short or other catastrophy.
An amp input is unlikely to be able to emit a harmful amount of DC, unless something really catastrophic has happened. If you rely on a cap for safety then it must be a proper mains-rated cap, X or Y (I forget which), which is guaranteed never to fail short-circuit.
yeah, the guy wasn't claiming that it would thoroughly stop the dc, but rather, buy the guitarist a moment of time until the cap failed, to get the guitar off. anyway, could be unfounded. here's a quote from the page:
An 0.33uf, 400V metal film capacitor. [SIZE=-1]Note, this is a change (10/06/1998) from the 1uf capacitor previously recommended. This capacitor was originally placed to prevent DC shocks from defective tube amplifiers. I hadn't given much thought to blocking AC because I always check wiring recepticals and so on. A recent discussion on the REC.MUSIC.MAKERS.GUITAR newsgroup brought it to my attention that many others don't and that some gigging musicians simply can't check/control wiring. This new value will provide some protection from the dreaded PA-and-Amp-on-different-polarities problem while providing much better noise performance than the parallel combination of a tiny capacitor and resistor recommended by some. This will limit current on a 120V, 60hz system to a maximum of about 30ma – certainly enough to get your attention but less likely to be lethal for most people. Even so, if you gig a lot and don't have control over the wiring – play it safe by either deleting the string (bridge) ground connection entirely or by using the tiny capacitor and parallel resistor method to isolate the strings. See the very important article on electrical shock[/SIZE]
An 0.33uf, 400V metal film capacitor. [SIZE=-1]Note, this is a change (10/06/1998) from the 1uf capacitor previously recommended. This capacitor was originally placed to prevent DC shocks from defective tube amplifiers. I hadn't given much thought to blocking AC because I always check wiring recepticals and so on. A recent discussion on the REC.MUSIC.MAKERS.GUITAR newsgroup brought it to my attention that many others don't and that some gigging musicians simply can't check/control wiring. This new value will provide some protection from the dreaded PA-and-Amp-on-different-polarities problem while providing much better noise performance than the parallel combination of a tiny capacitor and resistor recommended by some. This will limit current on a 120V, 60hz system to a maximum of about 30ma – certainly enough to get your attention but less likely to be lethal for most people. Even so, if you gig a lot and don't have control over the wiring – play it safe by either deleting the string (bridge) ground connection entirely or by using the tiny capacitor and parallel resistor method to isolate the strings. See the very important article on electrical shock[/SIZE]
Last edited:
In olden days some guitar amps had no mains transformer so an input cap might be the only thing stopping the guitarist from appearing at the great gig in the sky.
Perhaps it is more of a protective measure for the user rather than to block DC from the incoming component.
At least in guitar amps that makes sense. Last time I checked guitars aren't known for spontaneously producing DC 😀
But.... you never know with active pickups these days.
At least in guitar amps that makes sense. Last time I checked guitars aren't known for spontaneously producing DC 😀
But.... you never know with active pickups these days.
Audiophools like to risk their gear by not using any input caps. Just in case your valve gets a anode-grid short, you risk of getting your audio source burned out..
One cap more or less doesn´t matter.
One cap more or less doesn´t matter.
Last edited:
Wow, in 40 plus years of tinkering I have never had a small signal tube exhibit this sort of failure. (And that is probably over several thousand pieces of equipment I have repaired, modified, designed or built.) I have several thousand tubes in my stash and again I have not yet encountered a tube with this failure, unfortunately other sorts of failures have been much more common. (And I have had a few inexpensive 300B that exhibited intermittent shorts between filament and grid) Properly built tubes just don't fail this way in my limited experience, but somewhere out there there is undoubtedly a tube that has failed due to a grid to plate short. I have heard this chestnut so many times over the years and have yet to meet anyone who has had a small signal tube fail this way. (I belong to a local tube oriented audio club with ~40 members and this discussion never comes up) I would put this in the category of near urban myth.
Given the limited current available in most input stages (IT coupled excluded) and the large plate resistors generally used I would say the likelihood of damaging a source is pretty low.
Many years ago (20+ yrs) I did a double blind test between no cap and several of varying quality - everyone easily identified the one situation where there were no caps in the circuit, most identified and had a marked preference for a specific cap under cap to cap comparison with the same device.
My current power amplifiers use no coupling caps at all so why would I want to add one unnecessarily in the signal path?
I guess I'll just happily adopt my "Audiophool" moniker if avoiding caps I have found unnecessary warrants that label.. 😀
Many sources have output coupling caps, having 2 caps right is a row doesn't make sense especially when good coupling caps aren't cheap. And IMO one cap in series with the signal is definitely audible...
- Status
- Not open for further replies.