Hi!
I have a 4x10" Ashdown ABM bass cab driven by a Hughes&Kettner bass top. I just had the idea of modifying the bass top to act as a current amplifier with 8 ohms output imepdance. Has anyone tried this so far? Obviously we're not talking about HiFi here!
I once had the cab driven by a Tube Amps 8Ohm output and although i didnt hear that much difference as i expected, i think, the sound was a bit less harsh even with the amp not driven into saturation.
I have a 4x10" Ashdown ABM bass cab driven by a Hughes&Kettner bass top. I just had the idea of modifying the bass top to act as a current amplifier with 8 ohms output imepdance. Has anyone tried this so far? Obviously we're not talking about HiFi here!
I once had the cab driven by a Tube Amps 8Ohm output and although i didnt hear that much difference as i expected, i think, the sound was a bit less harsh even with the amp not driven into saturation.
Sorry, but that sounds a completely silly idea - the speakers won't be damped correctly, and you're throwing away most of your power.
If you REALLY want to do it?, just feed the speaker via a LARGE 8 ohm resistor.
The valve amp sounded less harsh because of it's limited anf unlevel frequency response.
If you REALLY want to do it?, just feed the speaker via a LARGE 8 ohm resistor.
The valve amp sounded less harsh because of it's limited anf unlevel frequency response.
honestly, i don't think so. even when cutting all the top end on my amp, it doesnt sound as soft as the tube amp.
is running the cab through an 8 ohm resistor not wasting much more power than driving the cab by current with having only a resistor of about half an ohm in the signal path?
It's exactly the same, running from an 8 ohm output impedance wastes at least 50% of the power in the feed (in this case the resistor) - in a valve amp it's wasted in the valves, the transformer etc.
If you want to run it from half an ohm, place a half ohm resistor in series, but the same reasons apply (just less).
You surely must have seen this scheme in numerous musical instrument amps. It has been pretty much a standard feature since early 1990's, though earliest examples of its implementation in SS guitar amps date to mid 1960's. That's how known it is.
Anyway, load current is sensed across a low-ohm resistor in series with the speaker. The resulting signal is fed back (in negative phase with the input), which subsequently increases the output impedance and naturally skews the amplifier's frequency response when its driving a reactive load such as a loudspeaker. Unlike with the usual only-voltage-feedback setup, the amp's voltage gain now becomes affected by the load impedance, like it does in tube amps with low amounts of NFB and inherently high output Z.
The technique is basically just an inversion of the positive current feedback scheme that was developed in 1950's to linearize response of a tube amplifier, which suffered from the effects of high output impedance. Basically, instead of positive current feedback, which decreases output impedance, you use negative current feedback, which has an opposite effect.
The schematic example, I believe, is from some Fender amp. You'll find a similar setup from hundreds of other musical instrument amps as well.
Anyway, load current is sensed across a low-ohm resistor in series with the speaker. The resulting signal is fed back (in negative phase with the input), which subsequently increases the output impedance and naturally skews the amplifier's frequency response when its driving a reactive load such as a loudspeaker. Unlike with the usual only-voltage-feedback setup, the amp's voltage gain now becomes affected by the load impedance, like it does in tube amps with low amounts of NFB and inherently high output Z.
The technique is basically just an inversion of the positive current feedback scheme that was developed in 1950's to linearize response of a tube amplifier, which suffered from the effects of high output impedance. Basically, instead of positive current feedback, which decreases output impedance, you use negative current feedback, which has an opposite effect.
The schematic example, I believe, is from some Fender amp. You'll find a similar setup from hundreds of other musical instrument amps as well.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Absolutely not. Particularly in guitar amps it's a common solution for emulating the low damping factor resulting from rather small amounts of NFB present in a lot of tube amps. You might say it allows the "character" of the speaker to develop fully.
No, it´s not silly at all ... in a musical instrument amplifier.
1) that´s the point: the speakers won´t be damped (or *much* less than before) , which is good.
There will be a bass peak at the resonant frequency: good.
2) there´s no power wasting at all, because there´s no physical 8 ohm power resistor in series with the load, but the feedback is so configured for the power amp to *sound* like if it was there.
Well, you will lose *a little* power in the current sensing resistor.
I have made such amps for almost 40 years , almost 10000 of them, and I typically use a 0.1 ohm sampling resistor in series with the 4 ohm speaker load, so in a 100W amp I lose 2.5W.
I can live with that.
Well, in the musical instrument world, the peak is good.
And it´s not "one note" (like those horrible bass systems so popular in Car audio).
The effect is like rising 6dB (which is not *that* much) a single slider in an otherwise flat Graphic EQ.
And, after all, that´s exactly what classic tube bass amps do (Fender, Ampeg, etc.).
And the audible effect is not at all the same as rising the bass control by 6dB, because the regular tone control works on a very wide band.
Good, but different.
And it´s not "one note" (like those horrible bass systems so popular in Car audio).
The effect is like rising 6dB (which is not *that* much) a single slider in an otherwise flat Graphic EQ.
And, after all, that´s exactly what classic tube bass amps do (Fender, Ampeg, etc.).
And the audible effect is not at all the same as rising the bass control by 6dB, because the regular tone control works on a very wide band.
Good, but different.
Hi Juan Manuel, thanks for the information. I know it's an old thread, but I'm intrigued in using current feedback in a tube bass amp, as I like the idea to modify the impact of the cabinet on the sound of the amp. I would build an amp for myself with a control for current feedback (being able to go from positive to negative).
I've used the same 100 mOhm resistor by now, in the simulations, for a 8 Ohm cab. The "variable CFB" control would be there also to adapt to 4 Ohm cabs.
Thanks in advance,
Roberto
Jumping into a Zombie thread in a forum a rarely tread into, but i am very interested in current amps.
To the OP’s points.
It is very likely the 8Ω on the tube amp refers to the expected nominal speaker impedance matching not the output impedance.
If the speakers are 8Ω (nominal) and the amp has an 8Ω output importance you do not have a current amp, as for that the Rout has to be greater than the speaker impedance. This would be an amplifier right at the boundary and has maximum power transfer (aside: i 1st started thinking about current amps when i talked to an old WE engineer who thot it was silly to do anything but build an amplifier where Rout = Zspeaker).
As an amplifier approaches that zero crossing from the voltage side and transitions to current side, more and more of the speaker’s impedance is convolved with the end frequency response. A real issue on the hifi side where one of the aims is flat FR, but on a MI system perhaps a suitable effect.
This would be a situation where a variable Transimpedance amplifier would be very useful where. adial on the front can dial the Rout from near zero (voltage amp) to near infinisty (current amp) so that the appropriate effect for the situstion could be dialed in.
dave
To the OP’s points.
It is very likely the 8Ω on the tube amp refers to the expected nominal speaker impedance matching not the output impedance.
If the speakers are 8Ω (nominal) and the amp has an 8Ω output importance you do not have a current amp, as for that the Rout has to be greater than the speaker impedance. This would be an amplifier right at the boundary and has maximum power transfer (aside: i 1st started thinking about current amps when i talked to an old WE engineer who thot it was silly to do anything but build an amplifier where Rout = Zspeaker).
As an amplifier approaches that zero crossing from the voltage side and transitions to current side, more and more of the speaker’s impedance is convolved with the end frequency response. A real issue on the hifi side where one of the aims is flat FR, but on a MI system perhaps a suitable effect.
This would be a situation where a variable Transimpedance amplifier would be very useful where. adial on the front can dial the Rout from near zero (voltage amp) to near infinisty (current amp) so that the appropriate effect for the situstion could be dialed in.
dave
Well, he didn't think it through, or wasn't very well informed, or was entirely the wrong sort of engineer (maybe he was a soils engineer or something like that) !
There are three different types of electrical impedance matching, each of which is the best choice in some specific circumstances.
Consider an electrical power plant feeding the homes and businesses in a city, say, drawing, say, 10 MW of power. Would it be a good choice for the electrical station to have an output impedance matching the electrical load of all the homes it was feeding? That would waste 50% of the power generated at the sub-station! Not only would five million watts of precious power be wasted, but this prodigious waste would also fry the electrical generators!
So in this case, the ideal match is to have very low output impedance at the power plant, much less than the load impedance being fed. This minimizes power wasted inside the plant itself, and provides a relatively constant voltage to the entire load (city) as homes turn on and off electrical devices.
This is also the way we design audio preamps and power amps. The preamp is designed to have a low output impedance, the power amp a much higher one. In this case, the goal is to transfer all the voltage from the preamp to the power amp, and to prevent any reactive component of input or output impedance from having any significant effect on the frequency response. In other words, we design for maximum voltage transfer, not maximum power transformer.
On the other hand, when designing radio frequency amplifiers, matching generator and load impedances (Zout = Zload) is the norm. The goal is maximum power transfer from one gain stage to the next, and also, mismatched impedances result in voltage standing waves and all sorts of unpleasant side effects.
Finally, there is a third kind of impedance matching. Here, the source impedance is made much higher than the load impedance. This results in the load current becoming almost independent of load impedance. The old (and virtually obsolete) DIN 5-pin audio connector standard was of this type, with signal sources designed to put out a 1-uA signal current at high source impedance. The matching input would then receive 1 millivolt of audio signal per kilo-ohm of input impedance; using a 100k input impedance, for example, would result in a 100 mV audio signal. This type of design produces maximum current transfer from source to load.
This last type of impedance matching is also what you get, by default, when you use pentodes to drive a loudspeaker. The output resistance of a pentode (or push-pull pair of pentodes), even after step-down through the output transformer, tends to end up at 50 to a 100 ohms, much greater than the impedance of any typical guitar loudspeaker available on today's market.
In a nutshell: we can choose to maximize power transfer, voltage transfer, or current transfer. Each one of these has its uses, and all have been used in various applications at one time or another.
-Gnobuddy
There is an interesting article on the subject of current vs voltage drive of loudspeakers here: Amplifiers: Solid State amps verses Valve amps
I found many other articles on the subject, but most were not worth reading. Some are rants, of the "Everybody else is doing it wrong, and I am the lone genius that has had the wits to realize there's a better way" sort. Others are opinion pieces, light on information and heavy on emotion. The article I linked above, however, is more balanced.
The outright rants don't stand up to critical scrutiny. We all know a loudspeaker's frequency response varies with the amount of damping the voice coil experiences near its fundamental bass resonance. The damping force is a mix of mechanical damping from the spider and surround, and electrical damping from back-emf causing current flow through the amplifier.
An amplifier with low output impedance increases damping due to voice-coil back emf. But a loudspeaker manufacturer could produce the same increased damping, and therefore, the same frequency response, with a different amplifier, one with a high output impedance, simply by choosing different spider and surround damping materials that increase mechanical damping. (By reducing Qms, if we put it in Thiele-Small terms.)
So there is no case to be made for "Current drive is the one true drive!", as loudspeaker manufacturers can in fact tailor their woofer designs to work best with current drive, voltage drive, or anything in between.
When it comes to musical instrument amplifiers, the facts (as opposed to personal opinions) appear to be pretty simple. A year or two ago, a couple of diyAudio members made measurements of the actual output impedance of various typical tube guitar amps they own, and found their output impedances to be in the range from 25 ohms to 50 ohms, more or less.
As we know, typical modern solid-state amps have output impedances much less than one ohm, and often just a few milliohms.
Quite simply, this means that if you connect the same guitar speaker to a solid-state guitar amp, and a tube guitar amp, you will not get the same frequency response, and this means the two amplifiers will sound different.
Manufacturers of solid-state guitar amps, above all, want their amps to sound like tube guitar amps, which have been the benchmark for the vast majority of serious electric guitar players for some seventy years. Therefore, these manufacturers contrived a way to raise the output impedances of their solid-state guitar amps to match, more or less, the output impedances of typical tube guitar amps. This immediately eliminates one of the reasons why their offerings do not sound like tube amps.
And that's why solid-state guitar amps often use that mix of current and voltage feedback at the output. Not because it's inherently better than having a low output impedance, but simply because it makes a solid-state guitar amp behave more like a tube guitar amp.
-Gnobuddy
I found many other articles on the subject, but most were not worth reading. Some are rants, of the "Everybody else is doing it wrong, and I am the lone genius that has had the wits to realize there's a better way" sort. Others are opinion pieces, light on information and heavy on emotion. The article I linked above, however, is more balanced.
The outright rants don't stand up to critical scrutiny. We all know a loudspeaker's frequency response varies with the amount of damping the voice coil experiences near its fundamental bass resonance. The damping force is a mix of mechanical damping from the spider and surround, and electrical damping from back-emf causing current flow through the amplifier.
An amplifier with low output impedance increases damping due to voice-coil back emf. But a loudspeaker manufacturer could produce the same increased damping, and therefore, the same frequency response, with a different amplifier, one with a high output impedance, simply by choosing different spider and surround damping materials that increase mechanical damping. (By reducing Qms, if we put it in Thiele-Small terms.)
So there is no case to be made for "Current drive is the one true drive!", as loudspeaker manufacturers can in fact tailor their woofer designs to work best with current drive, voltage drive, or anything in between.
When it comes to musical instrument amplifiers, the facts (as opposed to personal opinions) appear to be pretty simple. A year or two ago, a couple of diyAudio members made measurements of the actual output impedance of various typical tube guitar amps they own, and found their output impedances to be in the range from 25 ohms to 50 ohms, more or less.
As we know, typical modern solid-state amps have output impedances much less than one ohm, and often just a few milliohms.
Quite simply, this means that if you connect the same guitar speaker to a solid-state guitar amp, and a tube guitar amp, you will not get the same frequency response, and this means the two amplifiers will sound different.
Manufacturers of solid-state guitar amps, above all, want their amps to sound like tube guitar amps, which have been the benchmark for the vast majority of serious electric guitar players for some seventy years. Therefore, these manufacturers contrived a way to raise the output impedances of their solid-state guitar amps to match, more or less, the output impedances of typical tube guitar amps. This immediately eliminates one of the reasons why their offerings do not sound like tube amps.
And that's why solid-state guitar amps often use that mix of current and voltage feedback at the output. Not because it's inherently better than having a low output impedance, but simply because it makes a solid-state guitar amp behave more like a tube guitar amp.
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.