Home |
Search |
Today's Posts |
#1
|
|||
|
|||
Power output meter amp schematic.
Most amps with "power" meters have VU meters
or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. |
#3
|
|||
|
|||
|
#4
|
|||
|
|||
I've been using Burr Brown MPY634KP analog multiplier ICs, seems they
would output an AC waveform in proportion to power from an amplifier with about the only external component being the current sensing resistor. You would still need a precision FW op amp rectifier to drive (a DC) meter. Jeff Goldsmith Yves wrote: "Patrick Turner" a écrit dans le message de news: ... Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Hi Pat, When I was active as ham, I widely used the MC1496 wich is a "doubly balanced mixer", more exactly a "four quadrants multiplier" (If you can "Google" you'll find tons of papers about). It could be a good choice for implementing the U x I function according than: -- It can work at vy low level (tens of millivolts) and thus does not need a large value of resistor to sample the current, nor some sort of "preamplification". So, minimal resistor is to be added in serie with the loudspeaker (althought it'is certainly possible to include it in the feedback loop). -- It easy to use it at AC level rather than DC, removing risk of thermal drift and the need of rectifying low level signals. -- It has some "gain" and thus ... -- It (perhaps) delivers sufficient level to directly drive a sensitive meter thru a simple rectifier. Except for the PSU, I bet a "single IC" project is feasible. My two pence ! Merry Xmas, Yves. |
#5
|
|||
|
|||
Yves wrote: "Patrick Turner" a écrit dans le message de news: ... Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Hi Pat, When I was active as ham, I widely used the MC1496 wich is a "doubly balanced mixer", more exactly a "four quadrants multiplier" (If you can "Google" you'll find tons of papers about). It could be a good choice for implementing the U x I function according than: -- It can work at vy low level (tens of millivolts) and thus does not need a large value of resistor to sample the current, nor some sort of "preamplification". So, minimal resistor is to be added in serie with the loudspeaker (althought it'is certainly possible to include it in the feedback loop). -- It easy to use it at AC level rather than DC, removing risk of thermal drift and the need of rectifying low level signals. -- It has some "gain" and thus ... -- It (perhaps) delivers sufficient level to directly drive a sensitive meter thru a simple rectifier. Except for the PSU, I bet a "single IC" project is feasible. My two pence ! I already have the psu required and the current sensing R can be the return wire of the common terminal for the speaker; at several amps there is enough voltage present to mweasure the current and if not a little step up tranny may do. Maybe this is the way to go. Thanks and have a good one, Patrick Turner. Merry Xmas, Yves. |
#6
|
|||
|
|||
Yves wrote: "Patrick Turner" a écrit dans le message de news: ... Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Hi Pat, When I was active as ham, I widely used the MC1496 wich is a "doubly balanced mixer", more exactly a "four quadrants multiplier" (If you can "Google" you'll find tons of papers about). It could be a good choice for implementing the U x I function according than: -- It can work at vy low level (tens of millivolts) and thus does not need a large value of resistor to sample the current, nor some sort of "preamplification". So, minimal resistor is to be added in serie with the loudspeaker (althought it'is certainly possible to include it in the feedback loop). -- It easy to use it at AC level rather than DC, removing risk of thermal drift and the need of rectifying low level signals. -- It has some "gain" and thus ... -- It (perhaps) delivers sufficient level to directly drive a sensitive meter thru a simple rectifier. Except for the PSU, I bet a "single IC" project is feasible. My two pence ! Merry Xmas, Yves. I could not find a meter circuit via Google for measuring audio power using the MC1496. The app notes don't have any typical application for the device used as such. Its an interesting device, and it does lots of things, but I may have to search more later. Maybe the old ham books I have may have something. Patrick Turner. |
#7
|
|||
|
|||
"Yves" When I was active as ham, I widely used the MC1496 wich is a "doubly balanced mixer", more exactly a "four quadrants multiplier" (If you can "Google" you'll find tons of papers about). ** The 1496 is NOT a four quadrant multiplier and cannot be used as one. The 1495 is such a multiplier but has been obsolete and unavailable since a decade ago. The AD 633 is a ( trim free) four quad multiplier and is available. .................. Phil |
#8
|
|||
|
|||
On Thu, 23 Dec 2004 14:59:51 +1100, Patrick Turner
wrote: I could not find a meter circuit via Google for measuring audio power using the MC1496. The app notes don't have any typical application for the device used as such. Its an interesting device, and it does lots of things, but I may have to search more later. Jung's _IC OP-Amp Cookbook_ includes various uses as a multiplier, and also some "precision rectifier" applications. Highly recommended. You'll still need to make some design decisions about how to sample current and whether to choose RMS and/or peak rectification. Or "VU" weighting? Certainly an interesting project. As a side topic, how important is current phase angle to your design? A loudspeaker's reflected current may be independently interesting to an amplifier builder. Good fortune, Chris Hornbeck " ** Go choke on your damn turkey." |
#9
|
|||
|
|||
Patrick Turner wrote:
Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Maybe use a dual control pentode. One grid would have the voltage waveform feeding the speakers. The other grid would have a current sampling resistor between the amp's ground (speaker wire return) and the speaker's ground. Speaker "ground" feeds the grid, amp ground would (thru a bias circuit) feed the cathode. The plate circuit (using a resistor) would have the product of the waveforms on both grids. However it would not be listenable audio! AC couple it to something similar to a VU meter. Use a pentagrid tube like a 6CS6 if you want to avoid a "triode" efect you might get from G3 to the plate when using a true pentode. See http://pw2.netcom.com/~wa2ise/radios/elvctube.jpg and http://pw2.netcom.com/~wa2ise/radios/6be6vol.html for rough ideas. Instead of a steady DC bias on the second control grid, feed the sampled current. |
#10
|
|||
|
|||
"Chris Hornbeck"
" Pat Turner " I could not find a meter circuit via Google for measuring audio power using the MC1496. The app notes don't have any typical application for the device used as such. Jung's _IC OP-Amp Cookbook_ includes various uses as a multiplier, and also some "precision rectifier" applications. Highly recommended. ** My copy of the second edition has no mention of it. There are several pages on the ( now obsolete) MC 1495 though. You'll still need to make some design decisions about how to sample current and whether to choose RMS and/or peak rectification. ** Since the aim is to measure true power in a speaker load, the use of rectification or RMS conversion after the multiplier would be **quite wrong**. The average value of the multiplier's output will be proportional to the true power in the load. As a side topic, how important is current phase angle to your design? ** Using a four quad multiplier takes care of that. A loudspeaker's reflected current may be independently interesting to an amplifier builder. ** When the voltage and current in a (reactive) speaker load have opposite polarities - then the multiplier will give a negative output signal. At all other times it will be positive. This is why it is essential not to rectify it - that would make reactive power into real power. A capacitor or pure inductor used for a load should show zero reading on a true power meter. .............. Phil |
#11
|
|||
|
|||
"Jeff Goldsmith" I've been using Burr Brown MPY634KP analog multiplier ICs, seems they would output an AC waveform in proportion to power from an amplifier with about the only external component being the current sensing resistor. You would still need a precision FW op amp rectifier to drive (a DC) meter. ** Wrong. With a resistive load, a four quad multiplier sensing voltage and current will give precision rectification and an output that is always positive. With a reactive load, there will be negative as well as positive swings and these must not be rectified - to do so would mean that pure reactances would show as dissipating power ( ie an impossibility). ............... Phil |
#12
|
|||
|
|||
You're right, Phil. It's even simpler than it seemed to me at the
time. Jeff Goldsmith Phil Allison wrote: "Jeff Goldsmith" I've been using Burr Brown MPY634KP analog multiplier ICs, seems they would output an AC waveform in proportion to power from an amplifier with about the only external component being the current sensing resistor. You would still need a precision FW op amp rectifier to drive (a DC) meter. ** Wrong. With a resistive load, a four quad multiplier sensing voltage and current will give precision rectification and an output that is always positive. With a reactive load, there will be negative as well as positive swings and these must not be rectified - to do so would mean that pure reactances would show as dissipating power ( ie an impossibility). .............. Phil |
#13
|
|||
|
|||
Patrick Turner wrote:
Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. For something you could try see ABSE. One page, about 300K. Cheers, John Stewart |
#14
|
|||
|
|||
Chris Hornbeck wrote:
Jung's _IC OP-Amp Cookbook_ includes various uses as a multiplier, and also some "precision rectifier" applications. Highly recommended. You'll still need to make some design decisions about how to sample current and whether to choose RMS and/or peak rectification. Or "VU" weighting? Certainly an interesting project. As a side topic, how important is current phase angle to your design? A loudspeaker's reflected current may be independently interesting to an amplifier builder. Anyone thought of going the software route and using a PIC chip for the calculations. They have built in ADCs and everything. Adam |
#15
|
|||
|
|||
"Adam Stouffer" wrote in message news:O9Cyd.8245$rL3.6162@trnddc03... : Chris Hornbeck wrote: : : Jung's _IC OP-Amp Cookbook_ includes various uses as a multiplier, : and also some "precision rectifier" applications. Highly recommended. : : You'll still need to make some design decisions about how to sample : current and whether to choose RMS and/or peak rectification. Or : "VU" weighting? Certainly an interesting project. : : As a side topic, how important is current phase angle to your design? : A loudspeaker's reflected current may be independently interesting : to an amplifier builder. : : Anyone thought of going the software route and using a PIC chip for the : calculations. They have built in ADCs and everything. : : : : Adam At the pricing these days, you'd start to think of 'm as basic building blocks, yes. Interesting option, for sure. Rudy |
#16
|
|||
|
|||
"Adam Stouffer" Anyone thought of going the software route and using a PIC chip for the calculations. They have built in ADCs and everything. ** The ******s way for sure. One 8 pin IC (ie the AD 633) and a few resistors gives you the power function as a continuous signal with great precision and to well beyond the audio band. The display requires only a DC volt meter with a linear scale calibrated in watts - vary the series resistor to give 1, 10, and 100 watts FSD. ............. Phil |
#17
|
|||
|
|||
"Ruud Broens" wrote in message ... : : "Adam Stouffer" wrote in message : news:O9Cyd.8245$rL3.6162@trnddc03... : : Anyone thought of going the software route and using a PIC chip for : the : : calculations. They have built in ADCs and everything. : : : : : : : : Adam : : At the pricing these days, you'd start to think of 'm : as basic building blocks, yes. Interesting option, for sure. : Rudy : You could make some spectacular displays 24 LEDs, multiplexed, could simultaneously display peak-hold and mean rms or retro using something like this: http://www.franktechniek.nl starting from the oscilloscope clock idea (ruskie 8 cm tubes are quite cheap in eu Happy spare time Rudy |
#18
|
|||
|
|||
Phil Allison wrote:
** The ******s way for sure. One 8 pin IC (ie the AD 633) and a few resistors gives you the power function as a continuous signal with great precision and to well beyond the audio band. The display requires only a DC volt meter with a linear scale calibrated in watts - vary the series resistor to give 1, 10, and 100 watts FSD. Ah but the PIC can drive a seperate LCD directly. You could use a tri color LED and have 128 shades to indicate the level. But this is getting into the realm of eye candy. Oh and bite the big one will ya. Adam |
#19
|
|||
|
|||
Phil Allison wrote: "Yves" When I was active as ham, I widely used the MC1496 wich is a "doubly balanced mixer", more exactly a "four quadrants multiplier" (If you can "Google" you'll find tons of papers about). ** The 1496 is NOT a four quadrant multiplier and cannot be used as one. The 1495 is such a multiplier but has been obsolete and unavailable since a decade ago. The AD 633 is a ( trim free) four quad multiplier and is available. ................. Phil The amp where I wanted to have power meters is an old Phase Linear. It has VU meters. I found that 10k, 220 ohms and one germanium diode gave the right voltage levels for the dB indicated for the meters, so i don't have to re-calibrate them. The meters are pretty awful, very "bouncy", so not worth spending much time on. However, maybe I won't have to do a thing since I managed to apparently damage a few power transistors during a repeat test after they survived nicely during the first test, even bridged into 4 ohms, where I measured 600 watts continuous, which was planned from the 55 volt rails I have. The problem seems to be cross conduction where both N and P output transistors appear to be drawing a shirtload of current even at low voltage levels once F raised above 20 kHz, plus the doubtful channel oscillates all to easily where it didn't at first and I suspect the output devices have changed their characteristics. Some of the driver transistors were damaged so I rebuilt the whole driver amp and tested that independantly, and its fine, then replaced the driver emitter followers, same problem so the OP devices seem to be sus. I have the tricks employed by Douglas Self to prevent latch up. A bigger dog of an amp I have never experienced. If I cannot manage to make this amp stable, reliable, I can't charge to repair it, which don't worry me because it was brought to me as a POS that maybe I could do something with. In hind sight, R&D on a 300w/channel amp isn't as easy as it first was thought to be. Next time it will be mosfets instead of BJT outputs, I seem to take far less time to build a better measuring mosfet amp that can be stable and reliable. Who needs Xmas with little challenges like this one! Patrick Turner. |
#20
|
|||
|
|||
Chris Hornbeck wrote: On Thu, 23 Dec 2004 14:59:51 +1100, Patrick Turner wrote: I could not find a meter circuit via Google for measuring audio power using the MC1496. The app notes don't have any typical application for the device used as such. Its an interesting device, and it does lots of things, but I may have to search more later. Jung's _IC OP-Amp Cookbook_ includes various uses as a multiplier, and also some "precision rectifier" applications. Highly recommended. You'll still need to make some design decisions about how to sample current and whether to choose RMS and/or peak rectification. Or "VU" weighting? Certainly an interesting project. As a side topic, how important is current phase angle to your design? A loudspeaker's reflected current may be independently interesting to an amplifier builder. Good fortune, Chris Hornbeck " ** Go choke on your damn turkey." The original amp just had VU meters to show signal voltage. I thought power meters would be nice, just measure the voltage across the speaker return lead in the amp to get current, and the out voltage is already there. As long as the needles flicker around the top end of the dial when the sound is extremely loud, and down the low end when not so loud, all will be well. Patrick Turner. |
#21
|
|||
|
|||
robert casey wrote: Patrick Turner wrote: Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Maybe use a dual control pentode. One grid would have the voltage waveform feeding the speakers. The other grid would have a current sampling resistor between the amp's ground (speaker wire return) and the speaker's ground. Speaker "ground" feeds the grid, amp ground would (thru a bias circuit) feed the cathode. The plate circuit (using a resistor) would have the product of the waveforms on both grids. However it would not be listenable audio! AC couple it to something similar to a VU meter. Use a pentagrid tube like a 6CS6 if you want to avoid a "triode" efect you might get from G3 to the plate when using a true pentode. See http://pw2.netcom.com/~wa2ise/radios/elvctube.jpg and http://pw2.netcom.com/~wa2ise/radios/6be6vol.html for rough ideas. Instead of a steady DC bias on the second control grid, feed the sampled current. Ther must be a way they did it using tubes. But I am trying to re-engineer a Phase Linear 700B. Valves would look silly in that. Patrick Turner. |
#22
|
|||
|
|||
Popular Electronics, October 1979 had an Audio Power Meter project. I
built a variation of it, using an Analog Devices AD633 analog multiplier. It worked quite well. I only have a faded photocopy of the article, and no scanner, but hopefully that will give the OP a possible source. Paul return email paulbabiak at rogers dot com |
#23
|
|||
|
|||
"Adam Stouffer" Phil Allison wrote: ** The ******s way for sure. One 8 pin IC (ie the AD 633) and a few resistors gives you the power function as a continuous signal with great precision and to well beyond the audio band. The display requires only a DC volt meter with a linear scale calibrated in watts - vary the series resistor to give 1, 10, and 100 watts FSD. Ah but the PIC can drive a seperate LCD directly. ** Wow - use a digital display for a rapidly varying value ??? You could use a tri- color LED and have 128 shades to indicate the level. ** Even bigger wank than the PIC idea. But this is getting into the realm of eye candy. ** It is getting close to the Twilight Zone too. Oh and bite the big one will ya. ** Hell hath no fury like a ****** scorned....... ............ Phil |
#24
|
|||
|
|||
"Patrick Turner" The original amp just had VU meters to show signal voltage. I thought power meters would be nice, just measure the voltage across the speaker return lead in the amp to get current, and the out voltage is already there. ** True power meters on a audio amp are a COMPLETE WANK !!!!! No bloody use at all to the owner. They will not even show how close you are to clipping. Amps deliver voltage into a load - then reach a sudden limit which is not related to actual power. ............... Phil |
#25
|
|||
|
|||
Phil Allison wrote: "Patrick Turner" The original amp just had VU meters to show signal voltage. I thought power meters would be nice, just measure the voltage across the speaker return lead in the amp to get current, and the out voltage is already there. ** True power meters on a audio amp are a COMPLETE WANK !!!!! No bloody use at all to the owner. It may well have been some use to the owner. But I have since decided that the complexity of a power meter is not needed, and that if the owner has 4, 6 or 8 ohm speakers, then he can easily tell about what power he's using, ie, "not much", "a lot", and "****, the speakers are smoking..." They will not even show how close you are to clipping. Amps deliver voltage into a load - then reach a sudden limit which is not related to actual power. Indeed Phil, and a simple VU scale is all that's needed. I worked out how to get a corelation between signal voltage and the dB scale of the existing VU meters. I only need two R and one diode. Patrick Turner. .............. Phil |
#26
|
|||
|
|||
In another thread I said :-
However, maybe I won't have to do a thing since I managed to apparently damage a few power transistors during a repeat test after they survived nicely during the first test, even bridged into 4 ohms, where I measured 600 watts continuous, which was planned from the 55 volt rails I have. The problem seems to be cross conduction where both N and P output transistors appear to be drawing a shirtload of current even at low voltage levels once F raised above 20 kHz, plus the doubtful channel oscillates all to easily where it didn't at first and I suspect the output devices have changed their characteristics. Some of the driver transistors were damaged so I rebuilt the whole driver amp and tested that independantly, and its fine, then replaced the driver emitter followers, same problem so the OP devices seem to be sus. I have the tricks employed by Douglas Self to prevent latch up. A bigger dog of an amp I have never experienced. If I cannot manage to make this amp stable, reliable, I can't charge to repair it, which don't worry me because it was brought to me as a POS that maybe I could do something with. In hind sight, R&D on a 300w/channel amp isn't as easy as it first was thought to be. Next time it will be mosfets instead of BJT outputs, I seem to take far less time to build a better measuring mosfet amp that can be stable and reliable. Who needs Xmas with little challenges like this one! Not anyone. But I found out what my problem was. It wasn't degenerated power transistors. What I had done was move a 2 uF cap shunting the main power rails in addition to the 10,000 uF caps and when i soldered it into the revised position, one lead went to the wrong sub-rail, one of the the base inputs of the power transistors, the other to ground. Thus the HF gain was severely down, and the amp was severely unstable. Once I re-connected the cap to its rightful role the problem vanished and I regained some faith in the progress I have made. No doubt many of you will have time to spare for tinkering over the next few days. Be ou****ching for the forgotten mistake that is then mis-diagnosed, be not thus much swearing, and be not distracted by the sound of chimney struggles as Santa descends late this evening, only to find to his surprise that some bustard has fixed a plywood panel and a large subwoofer over the front of the now disused fireplace. I still think tubes are best, followed by mosfets, and finally BJTs. Patrick Turner. |
#27
|
|||
|
|||
robert casey wrote:
Patrick Turner wrote: Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Maybe use a dual control pentode. One grid would have the voltage waveform feeding the speakers. The other grid would have a current sampling resistor between the amp's ground (speaker wire return) and the speaker's ground. Speaker "ground" feeds the grid, amp ground would (thru a bias circuit) feed the cathode. The plate circuit (using a resistor) would have the product of the waveforms on both grids. However it would not be listenable audio! AC couple it to something similar to a VU meter. Use a pentagrid tube like a 6CS6 if you want to avoid a "triode" efect you might get from G3 to the plate when using a true pentode. See http://pw2.netcom.com/~wa2ise/radios/elvctube.jpg and http://pw2.netcom.com/~wa2ise/radios/6be6vol.html for rough ideas. Instead of a steady DC bias on the second control grid, feed the sampled current. The recommendation by Robert Casey is a good one, not for the audio but rather for the multiplying function to drive a power meter. That is backed by a reference in the text 'Electron Tube Circuits (1950)' by Samuel Seely. He recommends the 6L7 as a tube suitable for that function. The transfer characteristic shown is very accurate for the multiplying function. Not sure how a 6CS6 or other common pentagrid tube would work but sure would be interesting to try the idea. Cheers, John Stewart |
#28
|
|||
|
|||
Phil Allison wrote:
"Adam Stouffer" Phil Allison wrote: ** Wow - use a digital display for a rapidly varying value ??? Yes. Faster readings than a bouncing needle. Or have it take samples for a second and then compute the total. You could use a tri- color LED and have 128 shades to indicate the level. ** Even bigger wank than the PIC idea. Uh, the PIC drives the LED and you get a quick visual indication of power being dissapated. ** Hell hath no fury like a ****** scorned....... Your words... they burn so much... can't take it.. *WAH* Adam |
#29
|
|||
|
|||
If you don't mind spending a little money, check out the AD834
from Analog Devices (Digi-key probably has the part). The data sheet has a circuit for a power meter, but AD also makes a lot of parts for the power industry. Steve -- Steven D. Swift, , http://www.novatech-instr.com NOVATECH INSTRUMENTS, INC. P.O. Box 55997 206.301.8986, fax 206.363.4367 Seattle, Washington 98155 USA |
#30
|
|||
|
|||
John Stewart wrote: robert casey wrote: Patrick Turner wrote: Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Maybe use a dual control pentode. One grid would have the voltage waveform feeding the speakers. The other grid would have a current sampling resistor between the amp's ground (speaker wire return) and the speaker's ground. Speaker "ground" feeds the grid, amp ground would (thru a bias circuit) feed the cathode. The plate circuit (using a resistor) would have the product of the waveforms on both grids. However it would not be listenable audio! AC couple it to something similar to a VU meter. Use a pentagrid tube like a 6CS6 if you want to avoid a "triode" efect you might get from G3 to the plate when using a true pentode. See http://pw2.netcom.com/~wa2ise/radios/elvctube.jpg and http://pw2.netcom.com/~wa2ise/radios/6be6vol.html for rough ideas. Instead of a steady DC bias on the second control grid, feed the sampled current. The recommendation by Robert Casey is a good one, not for the audio but rather for the multiplying function to drive a power meter. That is backed by a reference in the text 'Electron Tube Circuits (1950)' by Samuel Seely. He recommends the 6L7 as a tube suitable for that function. The transfer characteristic shown is very accurate for the multiplying function. Not sure how a 6CS6 or other common pentagrid tube would work but sure would be interesting to try the idea. Cheers, John Stewart A 6BE6 would be easier and smaller for a multiplier perhaps. One og the control gids gives gm x RL at the anode, and where you have two grids, and two voltage inputs, one of which is dereived from a current sensor you get gm x gm x RL. I'm not sure though if that is the case, I only know that a 6BE6 is a good frequency converter, since an RF oscillator is made to saturate the tube current, and the small RF AM signal is then introduced to cause quite small intermodulation products at Fo + RF, and Fo - Rf, and the IF is tuned to one of these IMD products. Perhaps a dual gate j-fet or a 6DT6 may also work OK. Patrick Turner. |
#31
|
|||
|
|||
On Thu, 23 Dec 2004 16:42:32 +1100, "Phil Allison"
wroth: ** When the voltage and current in a (reactive) speaker load have opposite polarities - then the multiplier will give a negative output signal. At all other times it will be positive. This is why it is essential not to rectify it - that would make reactive power into real power. A capacitor or pure inductor used for a load should show zero reading on a true power meter. ............. Phil It would be satisfying, for me anyway, to have a 'scope type display with separate real and reactive instantaneous powers displayed. Music is quite "dynamic" and you would miss a lot of information by using a simple meter to display "power". I could use the old MK I eyeball to get averages easily from a 'scope display. Jim |
#32
|
|||
|
|||
On Fri, 24 Dec 2004 18:53:21 +1100, Patrick Turner
wroth: The original amp just had VU meters to show signal voltage. I thought power meters would be nice, just measure the voltage across the speaker return lead in the amp to get current, and the out voltage is already there. As long as the needles flicker around the top end of the dial when the sound is extremely loud, and down the low end when not so loud, all will be well. Patrick Turner. You would not gain much more information from the current measurement over a simple voltage measurement. Most speakers are designed to be as close to resistive as they can be. Reactive components are usually very small in comparison to the purely resistive part. It would only be necessary to include a switch to change the scale factor that related volts to power in order to get accuracy good enough for driving a moving coil meter. And a switch would only be necessary if you were going to use different speakers. Jim |
#33
|
|||
|
|||
On Fri, 24 Dec 2004 20:54:55 +1100, "Phil Allison"
wroth: Amps deliver voltage into a load - then reach a sudden limit which is not related to actual power. .............. Phil Not related to power? The output voltage is clipped at some level. The output voltage is directly related to the output power. How can clipping NOT be related to output power unless by "power" you mean a time averaged power measurement. Jim |
#34
|
|||
|
|||
Patrick Turner wrote:
John Stewart wrote: robert casey wrote: Patrick Turner wrote: Most amps with "power" meters have VU meters or other voltage meters which only tell you about the output voltage of an amp. A real power meter tells you the product of voltage and current, so that when the load value is halved and the output voltage is kept constant as is the case with many amps, the power will double. Who knows of a very simple circuit that is at least accurate to about 10%? There isn't a need for any greater accuracy. Patrick Turner. Maybe use a dual control pentode. One grid would have the voltage waveform feeding the speakers. The other grid would have a current sampling resistor between the amp's ground (speaker wire return) and the speaker's ground. Speaker "ground" feeds the grid, amp ground would (thru a bias circuit) feed the cathode. The plate circuit (using a resistor) would have the product of the waveforms on both grids. However it would not be listenable audio! AC couple it to something similar to a VU meter. Use a pentagrid tube like a 6CS6 if you want to avoid a "triode" efect you might get from G3 to the plate when using a true pentode. See http://pw2.netcom.com/~wa2ise/radios/elvctube.jpg and http://pw2.netcom.com/~wa2ise/radios/6be6vol.html for rough ideas. Instead of a steady DC bias on the second control grid, feed the sampled current. The recommendation by Robert Casey is a good one, not for the audio but rather for the multiplying function to drive a power meter. That is backed by a reference in the text 'Electron Tube Circuits (1950)' by Samuel Seely. He recommends the 6L7 as a tube suitable for that function. The transfer characteristic shown is very accurate for the multiplying function. Not sure how a 6CS6 or other common pentagrid tube would work but sure would be interesting to try the idea. Cheers, John Stewart A 6BE6 would be easier and smaller for a multiplier perhaps. One og the control gids gives gm x RL at the anode, and where you have two grids, and two voltage inputs, one of which is dereived from a current sensor you get gm x gm x RL. The 6BE6's grid 3 (its second control grid) has a remote cutoff characteristic. Great for radios, but not here. The 6CS6 is another pentagrid tube, but it's grid 3 is a sharp cutoff characteristic. This tube should be cheap enough, as it was a common "TV" tube. |
#35
|
|||
|
|||
"James Meyer" "Phil Allison" Amps deliver voltage into a load - then reach a sudden limit which is not related to actual power. Not related to power? ** Clipping is not related to **actual** power. Please learn to read - dickhead. Clipping will occur even with NO load and therefore ZERO power. ................ Phil |
#36
|
|||
|
|||
"James Meyer" You would not gain much more information from the current measurement over a simple voltage measurement. Most speakers are designed to be as close to resistive as they can be. ** On what planet is that the case ??????? Reactive components are usually very small in comparison to the purely resistive part. ** In some parallel universe maybe. Never looked at a real speaker system's impedance curve ??? Obviously not. Simple rule for YOU James: When you know bloody nothing - keep it to yourself. ............... Phil |
#37
|
|||
|
|||
On Mon, 27 Dec 2004 13:21:03 +1100, "Phil Allison"
wroth: "James Meyer" "Phil Allison" Amps deliver voltage into a load - then reach a sudden limit which is not related to actual power. Not related to power? ** Clipping is not related to **actual** power. Please learn to read - dickhead. Clipping will occur even with NO load and therefore ZERO power. ............... Phil I read better than you write, dickhead. You said "Amps deliver voltage into a load". Then you try to save your stupid statement about a "sudden limit not related to power" by saying zero power into NO load will produce clipping. So, which is it, dickhead? Loaded or not loaded? I think you were confused by the fact that you sometimes listen to your amps without any loads, not even a speaker. Of course you could always put one speaker lead in your mouth and the other up your ass. Given the vacuum in your head, that would also be a "no load" situation. Jim |
#38
|
|||
|
|||
On Mon, 27 Dec 2004 13:25:19 +1100, "Phil Allison"
wroth: "James Meyer" You would not gain much more information from the current measurement over a simple voltage measurement. Most speakers are designed to be as close to resistive as they can be. ** On what planet is that the case ??????? Earth. Which planet did you come from? Simple rule for YOU James: When you know bloody nothing - keep it to yourself. .............. Phil You keep nothing to yourself, do you? Even when you're wrong, you pour your verbal diarrhea out on newsgroups. Jim |
#39
|
|||
|
|||
"James Meyer" = moronic, lying, criminal pig "Phil Allison" Amps deliver voltage into a load - then reach a sudden limit which is not related to actual power. Not related to power? ** Clipping is not related to **actual** power. Please learn to read - dickhead. Clipping will occur even with NO load and therefore ZERO power. I read better than you write, dickhead. ** You are an illiterate, stinking ****wit - on a good day. You said "Amps deliver voltage into a load". Then you try to save your stupid statement about a "sudden limit not related to power". ** What happened to the word "actual " ?????? So, which is it, dickhead? Loaded or not loaded? ** Amps clip whether loaded or not and a power meter cannot be calibrated to indicate when for a speaker load. A voltage responding meter can be so calibrated. I think you were confused by the fact that you sometimes listen to your amps without any loads, not even a speaker. Of course you could always put one speaker lead in your mouth and the other up your ass. Given the vacuum in your head, that would also be a "no load" situation. ** I think the hospital made a mistake - you were meant to be aborted and flushed down the sewer. Or is that what happened and you crawled back out ?? ............. Phil |
#40
|
|||
|
|||
"James Meyer" = lying, POS asshole .
Phil Allison You would not gain much more information from the current measurement over a simple voltage measurement. Most speakers are designed to be as close to resistive as they can be. ** On what planet is that the case ??????? Earth. Which planet did you come from? ** But you do not live on earth - you inhabit the sewers on another planet. Planet Moron - where the worst scum on earth are sent for punishment. You dine on the floating excreta. Simple rule for YOU James: When you know bloody nothing - keep it to yourself. You keep nothing to yourself, do you? Even when you're wrong, you pour your verbal diarrhea out on newsgroups. ** Simple rule for YOU James: YOU know bloody nothing - so **** the hell off. ............. Phil |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
KISS 117 by Andre Jute | Vacuum Tubes | |||
KISS 113 by Andre Jute | Vacuum Tubes | |||
KISS 102 by Andre Jute | Vacuum Tubes | |||
rec.audio.car FAQ (Part 2/5) | Car Audio |