the bass and mid amplifiers do not need a passband that extends up to the normal RF atenuating filter frequency. You can usefully lower the input filter significantly.
This also reduces the signal that the amplifier has to process and should very slightly improve the performance of the amplifier in it's required passband.
Similarly the DC blocking filter can be raised in frequency for the mid and treble amplifiers.
Again this reduces the signal that these amplifiers have to process and could lead to slightly better performance.
The bass could have the two input filters set to 2Hz to 3kHz for a 300Hz crossover
Mid could be 30Hz to 40kHz for the 300Hz to 4kHz crossover
Treble could be 400Hz to 200kHz for the 4kHz crossover.
Note I have applied a one decade factor to all the input filters.
you can choose this, or a different factor to ensure effective filtering of the unwanted signal and passing enough signal such that you achieve the necessary crossover filtering and get the acoustic output you need/expect.
This also reduces the signal that the amplifier has to process and should very slightly improve the performance of the amplifier in it's required passband.
Similarly the DC blocking filter can be raised in frequency for the mid and treble amplifiers.
Again this reduces the signal that these amplifiers have to process and could lead to slightly better performance.
The bass could have the two input filters set to 2Hz to 3kHz for a 300Hz crossover
Mid could be 30Hz to 40kHz for the 300Hz to 4kHz crossover
Treble could be 400Hz to 200kHz for the 4kHz crossover.
Note I have applied a one decade factor to all the input filters.
you can choose this, or a different factor to ensure effective filtering of the unwanted signal and passing enough signal such that you achieve the necessary crossover filtering and get the acoustic output you need/expect.
Last edited:
TDA7293 tri-amp optimization follow-up
Thank you Andrew, your response makes a lot of sense conceptually, though being relatively new to this hobby, I can use a little more hand-holding.
I marked-up Daniel's schematic in the attachment below to highlight the area I think you are referring to. I'd appreciate any additional information to help identify the specific components that make up the Input filter and DC blocker to be optimized for specific drivers.
[/IMG]
I also found this link RC Low-pass Filter Design Tool as I started to research Frequency filters. My understanding is that to modify the input frequency cutoff points I would need to combine the right series Resistor and parallel Capacitor in the two areas highlighted in the diagram above, correct?
After looking at this for a while, a question came up in my mind, is the signal frequency not already constrained by the active crossover electronics before it gets to the Amp? How do the amp's filters work in combination with the crossover filtering action?
Thanks again, I am still learning tons every day, and this forum/thread continues to be a fountain of solid information.
Sixto.
Thank you Andrew, your response makes a lot of sense conceptually, though being relatively new to this hobby, I can use a little more hand-holding.
I marked-up Daniel's schematic in the attachment below to highlight the area I think you are referring to. I'd appreciate any additional information to help identify the specific components that make up the Input filter and DC blocker to be optimized for specific drivers.
I also found this link RC Low-pass Filter Design Tool as I started to research Frequency filters. My understanding is that to modify the input frequency cutoff points I would need to combine the right series Resistor and parallel Capacitor in the two areas highlighted in the diagram above, correct?
After looking at this for a while, a question came up in my mind, is the signal frequency not already constrained by the active crossover electronics before it gets to the Amp? How do the amp's filters work in combination with the crossover filtering action?
Thanks again, I am still learning tons every day, and this forum/thread continues to be a fountain of solid information.
Sixto.
page not found.
Attach your pic
The low pass filter tool shows the RC. Rin series with the signal and C shunting the unwanted part of the signal to Signal Ground/Signal Return.
Typically I would use 1k0 and 680pF MKP for a 0.68us RC, giving an F-3dB ~ 200kHz
This works for a wideband amplifier. It will work with each of your bass, mid and treble amplifiers. But I know that a 200kHz filter is NOT required for a bass only amplifier.
I can increase the capacitor to 6.8nF for a 20kHz filter or go even lower to provide more effective interference filtering.
Looking at the high pass filter, the gains in using a "tailored" roll-off frequency are much higher.
Typically for the CR filter one would use 22k for the Rin and this requires 3.6uF for the capacitor to give a 2Hz F-3dB.
That is a big and can be expensive capacitor.
One does not need to pass 2Hz to the treble, or mid amplifiers.
You can save money and space by using a much lower value capacitor.
But there is a second saving.
The NFB RC must be longer than the input filter.
The PSU RC must be longer than the NFB RC.
If you raise the high pass filter at the input you can also raise the RC value of the NFB and the PSU. This again saves space and money.
For an 8ohms speaker I use 20mF as the PSU capacitance to give a 160ms RC, The NFB would use 120uF to give a 120ms RC to suit a 80ms RC input filter.
If I change the input filter to 400Hz (for the 4kHz crossover) I can reduce the capacitors by a factor of 400:2, i.e the 20mF becomes 100uF for the PSU and the 120uF NFB cap becomes 600nF. This can be a plastic film type and could even be MKP.
It is almost feasible to use plastic film instead of electrolytic in the PSU. Oh ! that's used up all the money you just saved on making all the other capacitors smaller.
Attach your pic
The low pass filter tool shows the RC. Rin series with the signal and C shunting the unwanted part of the signal to Signal Ground/Signal Return.
Typically I would use 1k0 and 680pF MKP for a 0.68us RC, giving an F-3dB ~ 200kHz
This works for a wideband amplifier. It will work with each of your bass, mid and treble amplifiers. But I know that a 200kHz filter is NOT required for a bass only amplifier.
I can increase the capacitor to 6.8nF for a 20kHz filter or go even lower to provide more effective interference filtering.
Looking at the high pass filter, the gains in using a "tailored" roll-off frequency are much higher.
Typically for the CR filter one would use 22k for the Rin and this requires 3.6uF for the capacitor to give a 2Hz F-3dB.
That is a big and can be expensive capacitor.
One does not need to pass 2Hz to the treble, or mid amplifiers.
You can save money and space by using a much lower value capacitor.
But there is a second saving.
The NFB RC must be longer than the input filter.
The PSU RC must be longer than the NFB RC.
If you raise the high pass filter at the input you can also raise the RC value of the NFB and the PSU. This again saves space and money.
For an 8ohms speaker I use 20mF as the PSU capacitance to give a 160ms RC, The NFB would use 120uF to give a 120ms RC to suit a 80ms RC input filter.
If I change the input filter to 400Hz (for the 4kHz crossover) I can reduce the capacitors by a factor of 400:2, i.e the 20mF becomes 100uF for the PSU and the 120uF NFB cap becomes 600nF. This can be a plastic film type and could even be MKP.
It is almost feasible to use plastic film instead of electrolytic in the PSU. Oh ! that's used up all the money you just saved on making all the other capacitors smaller.
Last edited:
I marked-up Daniel's schematic in the attachment below to highlight the area I think you are referring to. I'd appreciate any additional information to help identify the specific components that make up the Input filter and DC blocker to be optimized for specific drivers.
Let me try a link to the image and see if that works this time?
http://origin.dastatic.com/forums/gallery/data/500/medium/TDA7293_Diagram1.JPG
The link to my uploaded gallery file seems to work on the preview (but so did the image when I first uploaded it. Can't seem to find a how-to for uploading images in-line with text - Help anyone?
Sixto
use much less than maximum voltage for mid&treble amp
The woofer amp will work as is, on that schematic (with unregulated power). It is in active service for both woofer and full bandwidth use. The woofer amp would have more abusive use, and higher gain to assure stability in rough conditions. To match up the woofer amp with the lower gain regulated mid&tweeter amp, would mean a divider (pot) at the input to the woofer amp.
The schematic provides similar bass performance to the LM3886, although you'll notice I coaxed, cajoled and basically begged the TDA7293 long before excellent bass performance was achieved, as you can see by the biggie size filter values. It is thoroughly stable even while driving at maximum and requires only modest heatsinks. That's a fine thing for a party amp, but that's where the good news ends.
The bad news is that you'll have to undervolt it severely to get access to its finer qualities, including useful stable output while set to lower gain. I don't see that as a problem if you wanted a mid&treble amplifier, because it may actually be a solution.
I estimate that the regulators will need to drop the rail voltage down to +-26vdc for TDA7293 or +-20vdc for TDA7294. My suggestion for the TDA7294 was proofed; however, the +-26vdc for the TDA7293 was merely extrapolated, so the optimal figure could be lower or higher.
Also, The mid&treble amplifiers are likely to need a capacitor in series in-between the speaker driver and the 0v rail, regardless if other crossover components are used or not. That added accessory part will prevent dc build-up in crossover components, drive the speaker more reasonably and possibly dramatically drop the load resulting in less or no clipping. There is significant odds for the additional filtering resulting in dramatically higher fidelity or at least more durability parameters of the speaker driver itself. I do need that added part if driving 4" speakers at very high output levels, and I suppose that the need of output filtering could be similar for little midranges and tweeters.
Sorry that was all the notes I have at this time. The schematic you mentioned is attached. On that schematic is an ear-friendly party/concert sound amplifier mindful of full blast performance. For the task described (economical concert tickets) it is a fantastic performer; perhaps at the cost of doing nothing else well. Let's mod it for other uses too.
Here it is. Also, I like this one. With TDA7293, gain should be set dependent on operating voltage (it is less stable at higher voltage). . . and that gain setting would be, approximately somewhat too high but fortunately lower voltage does allow for lower gain (a great fix, but only when you didn't need a lot of power). Therefore, it looks Very good to me if he regulated the voltage down Lower for the midrange and treble amplifiers, and then dialed the gain down lower for those smaller current, lower voltage amplifiers. I hope it would be as simple as replacing the series diodes with voltage regulator chips.
The woofer amp will work as is, on that schematic (with unregulated power). It is in active service for both woofer and full bandwidth use. The woofer amp would have more abusive use, and higher gain to assure stability in rough conditions. To match up the woofer amp with the lower gain regulated mid&tweeter amp, would mean a divider (pot) at the input to the woofer amp.
The schematic provides similar bass performance to the LM3886, although you'll notice I coaxed, cajoled and basically begged the TDA7293 long before excellent bass performance was achieved, as you can see by the biggie size filter values. It is thoroughly stable even while driving at maximum and requires only modest heatsinks. That's a fine thing for a party amp, but that's where the good news ends.
The bad news is that you'll have to undervolt it severely to get access to its finer qualities, including useful stable output while set to lower gain. I don't see that as a problem if you wanted a mid&treble amplifier, because it may actually be a solution.
I estimate that the regulators will need to drop the rail voltage down to +-26vdc for TDA7293 or +-20vdc for TDA7294. My suggestion for the TDA7294 was proofed; however, the +-26vdc for the TDA7293 was merely extrapolated, so the optimal figure could be lower or higher.
Also, The mid&treble amplifiers are likely to need a capacitor in series in-between the speaker driver and the 0v rail, regardless if other crossover components are used or not. That added accessory part will prevent dc build-up in crossover components, drive the speaker more reasonably and possibly dramatically drop the load resulting in less or no clipping. There is significant odds for the additional filtering resulting in dramatically higher fidelity or at least more durability parameters of the speaker driver itself. I do need that added part if driving 4" speakers at very high output levels, and I suppose that the need of output filtering could be similar for little midranges and tweeters.
Sorry that was all the notes I have at this time. The schematic you mentioned is attached. On that schematic is an ear-friendly party/concert sound amplifier mindful of full blast performance. For the task described (economical concert tickets) it is a fantastic performer; perhaps at the cost of doing nothing else well. Let's mod it for other uses too.
Attachments
![TDA7293_Diagram1[1].JPG](/community/data/attachments/462/462785-5ea9de4beb4a18293f5e0288c678486c.jpg)
Last edited:
I think I know how I got myself confused...
AndrewT, I apologize: I think I was not clear enough with my initial post when I said..
It took me a while to figure this out, because I initially assumed your response was directing me to tweak the resistors and capacitors connected to pins 3 and 4 on Daniel's schematic, (thank you for posting it Daniel) but now I realize that based on my post, you were guiding me to build RC x-over networks to narrow the frequencies going to each amp board. I agree completely with aiming for better Amp performance by narrowing the frequencies the Amp sees, but instead of using traditional passive RC networks, I'm opting for active x-over between the preamp and TDA7293 inputs. (I hope the net result is better via active X-over than passive)
Sixto - Minneapolis.
I think my tag-line will be: Simple, straight-forward communication is anything but!
AndrewT, I apologize: I think I was not clear enough with my initial post when I said..
I should have clarified that I'm planning to use an active crossover for each speaker box, either a Behringer SUPER-X PRO CX 3400 stereo 3 way / mono 4 way (on-hand, to be used at first for testing), or a DIY acrive crossover board (when I have the system sorted out).
It took me a while to figure this out, because I initially assumed your response was directing me to tweak the resistors and capacitors connected to pins 3 and 4 on Daniel's schematic, (thank you for posting it Daniel) but now I realize that based on my post, you were guiding me to build RC x-over networks to narrow the frequencies going to each amp board. I agree completely with aiming for better Amp performance by narrowing the frequencies the Amp sees, but instead of using traditional passive RC networks, I'm opting for active x-over between the preamp and TDA7293 inputs. (I hope the net result is better via active X-over than passive)
Sixto - Minneapolis.
I think my tag-line will be: Simple, straight-forward communication is anything but!
Thanks Daniel, I will try reducing the voltage (this will take me a while to do, I hope workload at my dayjob is unsustainable for the long-term) Are there any other variables I should test or change to prevent oscillation when I reduce from 35v rails?
Agreed, both the Behringer and the DIY crossover I'm looking to use have +/-12db ranges on each of their outputs. I plan to tune the gain of all the drivers once I have the whole thing running by using RoomEQWizard.
Yes, I am planning to do this to keep my tweeters and mids healthy as long as possible.
Thanks!
The line level active (opamp) cross-overs will filter the signals before sending them to the respective amplifiers.
That is OK.
It's when the signals arrive at the amplifiers that I am concerned about and advising that you do more to help amplifier performance.
I advise to everyone that the Receiver ALWAYS has a passive filter/s at the inputs. I also recommend that power amplifiers should have a DC blocking capacitor at it's input and this also limits the LF response of the amplifier. This is a passive filter
My previous post was recomending that the passive filtering at the power amplifier input be narrow banded to better exclude any interference and to reduce cost/size of amplifier components, or use an ultra quality component (at sensible cost) instead of a mediocre component.
eg. instead of using a 4u7F 50V bi-polar electrolytic at the input of the treble amplifier you could use a 100nF teflon dielectric and still pass ALL the treble signal to the power amplifier.
I'm getting there...
Yes, now I get it. (At least some of it) I like both the ideas of narrow-banding filters for my application, and using better components for the amp. This will take some experimenting to learn what works and why? Where I think I miss the train now, is at the relationship between the filters at the Input, NFB, and PSU in the quote below. i understand that they are all potential sources of noise. I don't understand yet why the filters at each of those points ought to be "longer" than the next, or exactly what "longer" means in this context.
Thank you for your patience and willingness to share,
Sixto
Yes, now I get it. (At least some of it) I like both the ideas of narrow-banding filters for my application, and using better components for the amp. This will take some experimenting to learn what works and why? Where I think I miss the train now, is at the relationship between the filters at the Input, NFB, and PSU in the quote below. i understand that they are all potential sources of noise. I don't understand yet why the filters at each of those points ought to be "longer" than the next, or exactly what "longer" means in this context.
Thank you for your patience and willingness to share,
Sixto
longer in terms of an RC time constant means more seconds (time)
A 100ms (milli-second) is longer than a 60ms time constant.
RC time constant is another way of specifying a passive RC single pole filter.
100ms is used in the standard formula
F-3dB = 1 / {2 Pi RC}
100ms = 1.6Hz
60ms = 2.7Hz
So longer equals lower frequency for the filter.
A 100ms (milli-second) is longer than a 60ms time constant.
RC time constant is another way of specifying a passive RC single pole filter.
100ms is used in the standard formula
F-3dB = 1 / {2 Pi RC}
100ms = 1.6Hz
60ms = 2.7Hz
So longer equals lower frequency for the filter.
Thanks Andrew! that's just opened up a whole new area of study for me. I had no idea that RC filters for various frequencies implied a time delay as well. Is this something that can be compensated for, or is the delay so small that it doesn't impact how we hear the whole signal. I'm wondering now, will my treble signal be out of synch with the bass?
I found another site that has various ways to quantify the relationship between the R and C values, the frequency and the time delay. http://www.sengpielaudio.com/calculator-timeconstant.htm Can't wait to test it! So, for example - if I run an un-filtered XXX Hz signal at the same wavelength as a filtered signal through two otherwise identical amp boards, I should theoretically see a delay of some millisecond duration on an 2-channel oscilloscope, right?
Thanks again, Sixto.
The RC is measured in time units, but it's not a time delay.
It is a phase delay. That means the signal arrives at a later time.
A single pole passive RC with the Frequency of roll -off defined by F-3dB = 1/{2 Pi RC}
has 45degrees of phase delay at the cut-off frequency.
If you cascade two single pole filters, then the two phase delays will add.
eg. A two pole active filter will have 90degrees of phase delay at the cut-off frequency, the same a two identical single pole filters.
It is a phase delay. That means the signal arrives at a later time.
A single pole passive RC with the Frequency of roll -off defined by F-3dB = 1/{2 Pi RC}
has 45degrees of phase delay at the cut-off frequency.
If you cascade two single pole filters, then the two phase delays will add.
eg. A two pole active filter will have 90degrees of phase delay at the cut-off frequency, the same a two identical single pole filters.
Last edited:
Actually, this amplifier is more stable at Lower voltage. You do want the multi-turn cermet dial to set the gain precisely on tolerance because it is conditional stability that yields the highest resolution for any amplifier. We certainly wouldn't want to be without that dial, but what we CAN do is improve the conditions so that "conditional stability" also results in lower gain.
Lower gain is not a desirable end result if that also means compromising tonality until it sounds awful if we cranked it up loud; but, it is a fantastically informative guide.
For midrange and tweeter, it would be a good idea to insert a protection cap in series, probably on the speaker negative (for any amplifier with bi-polar supply), which can block rather a lot of errata, quite seamlessly. Nichicon's 50v ES series are far more than enough with TDA7294 mids&treble amplifiers, plentiful resolution and also not expensive at all. Generally, the approach will cut speaker distortion to half (for any size speaker) and outright put a stop to tweeter fires from DC surges (that can't pass the redundant-looking extra cap).
However, you can temper that statement down a little by assuming that I might have too many beers and get really abusive on the audio equipment, rocking the house excessively; but, if one did make that assumption than one must also consider that the experience was a lot more pleasant with the filters in place, because that's really what they're for.
Although a series capacitor is fantastic for reducing speaker distortion, driving a capacitive load can reduce amplifier stability; so, it is fortunate that there is an elementary and well documented fix. . .
Amplifiers generally drive capacitive loads more stably if both the output RC and the output inductor||resistor (look at discrete amp schemas) are used, so that would be a good plan.
Startup value for the 1k multi-turn trimmer (lovely quality cermet) is 730 ohms, and just leave it right there for 35vdc~37vdc and woofer/party amp use. You can get the gain about 2x lower by using the trimmer rather than a fixed resistor. Set the dial for the lowest gain possible without obnoxious tone/behavior. We should hope to get the gain at somewhat less than 38x else the fidelity might be hindered. Without reducing the voltage, I was able to go as low as 36x. Or if the voltage was reduced dramatically, I could get access to the datasheet listed minimum gain figure, without making the amp sound/behave awful--there is that really practical limit on how low you'd want to set the gain (also illustrates cases when more gain could be a good thing).
The mids&treble amplifiers don't require 440u per rail for bari harmonics support, or at least the treble amp certainly doesn't need that much--just 220u per rail will do for it, at the amp board. Even smaller could be done, but I haven't checked that, so just use really good quality 220u caps. Also use the smallest size amplifier boards that you can find, so as to get the 220uF caps as close as possible to the chip! That layout issue will dramatically affect the stability. You'll be looking for the extremely small double-sided boards.
Andrew T is correct about the small signal filtering, and that's easy to prove too. I haven't mentioned anything about it, because he already said it.
However, a small signal problem could exist in this application. Digital means (and Active crossovers) does the eq job with such excellent quality and is able to decrease and many times remove the need of expensive inductors. But, the very worst thing to do with it is try to replace speaker dampening crossover components with what is basically a maxed-out eq. That's abusive and increases distortion horrendously. The good news is that alleviating the situation is accomplished by a few cheap resistors and caps. You'll want a padding resistor for the midrange to take out the shout, which is a job that eq means can't do thoroughly. You'll also want a padding resistor for the tweeter teamed up with both a tiny bypass cap and an RC as well, so as to prevent treble droop from the padding resistor. Since these caps merely assist the uppermost tiny signal to cross the padding resistor, they can be inexpensive, such as the very high resolution Nichicon ES series. . . which in that locale is not going to bear the full load anyway. Unfortunately for me, I had to spend an extra 60 cents because I needed 0.68u, so the cap that has bypassed my tweeter padding resistor is polyester (the 0.47u ES wasn't quite big enough). I almost spend an extra $20 for a high end 8uF cap to use for the RC that bypasses my tweeter padding resistor, but at the last minute, discovered that parallel 4.7u Nichicon ES could do the job on 12 cents and at considerably higher resolution. Anyway, don't use eq'ing means as a substitute for speaker driver electronic dampening modification. The amount of distortion is of great concern; so, plan on a primary approach of making the speaker work better.
AndrewT's small signal filtering advice and my large signal filter advice both serve to reduce reliance, distortion, and workload of the active crossover components, so another way to look at it is to reduce the reliance on active crossovers. Eventually, you might want to price-compare the difference between the active crossovers versus the large signal inductors, which is the only parts they can reasonably replace if compared at similar quality. I have indeed seen cases where the active components could win that comparison (albeit with a little assistance, previously mentioned). Active crossovers also make fairly good test equipment since any excessive/tall settings do serve to pinpoint the need of adding passive parts to make the speaker work better at less distortion.
Last edited:
Or, to make a long story short, all variables affect stability; and, that's why there's a dial on the schematic to adjust it to your situation.
The top variables that affect stability with these amplifiers, are voltage (the lower the better) and layout (how close the decoupler caps are to the predrive power pins--the closer the better). So, you want a tiny board and not max volts.
Full Thiele-Small output parts is also a good plan.
By starting out using these several other means to get better stability, you'll be able to, at the last step, dial the gain lower without causing bad sound. That approach makes available the high-end imaging/realism, if you also under-volt it.
Transformers for that approach
Tda7294, TDA7295, TDA7296,
12+12vac
13.5+13.5vac
Both TDA7294 and TDA7293,
14+14vac
15+15vac
And TDA7293 can also use these,
18+18vac
20+20vac
The top variables that affect stability with these amplifiers, are voltage (the lower the better) and layout (how close the decoupler caps are to the predrive power pins--the closer the better). So, you want a tiny board and not max volts.
Full Thiele-Small output parts is also a good plan.
By starting out using these several other means to get better stability, you'll be able to, at the last step, dial the gain lower without causing bad sound. That approach makes available the high-end imaging/realism, if you also under-volt it.
Transformers for that approach
Tda7294, TDA7295, TDA7296,
12+12vac
13.5+13.5vac
Both TDA7294 and TDA7293,
14+14vac
15+15vac
And TDA7293 can also use these,
18+18vac
20+20vac
Load impedance for filter design?
Thank you, I appreciate the advice. You guys have me reading every word at least 3 times before I even begin to internalize this. In other words, it is taking me a while to digest all the information provided so far. I am in the process of getting a better understanding of how RF filters work before posting a proposed configuration for my mids and treble amps for discussion. The journey is thrilling. I hope the end result is as good as I anticipate.
One question I have not yet figured out is determining what the device input impedance or "load impedance should be for calculating the F-3db value for the 3 or 4 RC filters recommended so far: DC, NFB and signal Input (+ and -?). When there is a parallel or series resistor working together with a parallel or series capacitor, I'm thinking I should use that resistor's value for the RC filter (High Pass, Low Pass or Band Pass), but when there is just a capacitor in series or in parallel, (such as is shown on most schematics for the DC input) should I use the proposed driver impedance of 8 ohms to calculate the appropriate cap value? Does the TDA7293 itself have additional impedance? Is it negligible? Do the other filters, chip and driver all add-up cumulatively and significantly to the total impedance?
Andrew used 8-Ohms in one of his responses to me, which is why I'm thinking that the driver impedance comes into play. So does that mean that when a series or parallel resistor is not present, the load (speaker driver?) impedance rules?
TIA - Sixto
Thank you, I appreciate the advice. You guys have me reading every word at least 3 times before I even begin to internalize this. In other words, it is taking me a while to digest all the information provided so far. I am in the process of getting a better understanding of how RF filters work before posting a proposed configuration for my mids and treble amps for discussion. The journey is thrilling. I hope the end result is as good as I anticipate.
One question I have not yet figured out is determining what the device input impedance or "load impedance should be for calculating the F-3db value for the 3 or 4 RC filters recommended so far: DC, NFB and signal Input (+ and -?). When there is a parallel or series resistor working together with a parallel or series capacitor, I'm thinking I should use that resistor's value for the RC filter (High Pass, Low Pass or Band Pass), but when there is just a capacitor in series or in parallel, (such as is shown on most schematics for the DC input) should I use the proposed driver impedance of 8 ohms to calculate the appropriate cap value? Does the TDA7293 itself have additional impedance? Is it negligible? Do the other filters, chip and driver all add-up cumulatively and significantly to the total impedance?
Andrew used 8-Ohms in one of his responses to me, which is why I'm thinking that the driver impedance comes into play. So does that mean that when a series or parallel resistor is not present, the load (speaker driver?) impedance rules?
TIA - Sixto
Last edited:
I found this page when I did a search for "Amplifier input impedance": Input Impedance of an Amplifier and How to Calculate it I don't think it answers all of my questions, but I hope it can help point me in the right direction...
Also, Daniel I am not sure I understand the comments below: are you recommending using a combination of RC and IR filters to stabilize the amp at lower voltages? (the well-documented fix?) I must admit I've only studied a handful of TDA7293 schematics (no discrete amps yet) so, I'm not familiar with this approach. Can you point me to an example?
Thanks - Six.
Also, Daniel I am not sure I understand the comments below: are you recommending using a combination of RC and IR filters to stabilize the amp at lower voltages? (the well-documented fix?) I must admit I've only studied a handful of TDA7293 schematics (no discrete amps yet) so, I'm not familiar with this approach. Can you point me to an example?
Thanks - Six.
Oh, man - as if I didn't have enough rabbit holes to crawl into yet... Good information - especially if I want to avoid oscillation on the tweeter/mid amps when I use a cap to block low frequency thumps!
Here are a few threads I found on the forums:
http://www.diyaudio.com/forums/solid-state/125679-function-output-inductor-37.html
http://www.diyaudio.com/forums/solid-state/132811-amp-stability-techniques-zobel-et-al.html
http://www.diyaudio.com/forums/solid-state/120748-zobel-no-zobel-2.html
I've subscribed and marked them for future analysis. Thank you. But first, I'd like to redirect your attentions to some of my earlier questions that I've not been able to figure out yet:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.