As I type one question, yet another idea comes in.
This is a good example of what I was asking.
No, it's not a part of the amp design but I also dig these meters and had wondered what was involved in anyones design to add them.
Do we consider questions of this nature irrelevant to the subject and as newbees how are we to know what is relevant to eliminate confusion when asking questions?
Bluto
This is a good example of what I was asking.
No, it's not a part of the amp design but I also dig these meters and had wondered what was involved in anyones design to add them.
Do we consider questions of this nature irrelevant to the subject and as newbees how are we to know what is relevant to eliminate confusion when asking questions?
Bluto
Bluto said:
To help in eliminating confustion, I need to build a simplified sample and document it.
It is also designed to use whatever components are available in your local market. That means there's no need to purchase any botique parts.
The design is held as a prior art, but its provided $0 for non-profit use and for educational use--as that is the documented intention of the design. This clears away copyright/patent concerns to stay at $0. So, that should help too.
I've worked fairly hard to decrease the complexities of the design; however, any input in the form of questions is quite helpful to further refine the project.
I'll have a sample layout for you to critique soon.
Re: Power, Power, and yet more Power. 
Sir,
Thanks, I have learnt quite a lot from your "Transformer lingo". Can we bridge LM3886 to get a higher power output ? for example like 100W or 200W rms ?
if it is possible then, what will be the recommended voltage and ampere rating for a Center tap transformer. Also, it'll be great if you could provide a simple power supply diagram along with the ratings of the different parts needed.
Regards
Chuba
Sir,
Thanks, I have learnt quite a lot from your "Transformer lingo". Can we bridge LM3886 to get a higher power output ? for example like 100W or 200W rms ?
if it is possible then, what will be the recommended voltage and ampere rating for a Center tap transformer. Also, it'll be great if you could provide a simple power supply diagram along with the ratings of the different parts needed.
Regards
Chuba
danielwritesbac said:Let's cover transformer lingo for a moment.
EDIT2: Fun photo content in this post.
Our LM3875TF can use dc rails voltages from 25vdc to 37vdc.
For academic reference, its centerline tolerance is approximately 30vdc. EDIT: Figures given are at expected power output levels
Well, the transformer outputs AC.
This is changed into dc by the rectifier (4 diodes or one-piece block).
Transformer lingo.
48vct and 24v dual, are 24 + 24 volts ac (rectifies to 35 +35 dc)
40vct and 20v dual, are 20 + 20 volts ac (rectifies to 30 +30 dc)
36vct and 18v dual, are 18 + 18 volts ac (rectifies to 26 +26 dc)
Dual secondaries, indicates that there are 4 wires for output, and quite confusing. See the power supply forum for that tangle.
Center tap, vct, indicates that there are 3 convenient wires for output, as used with the easy one-piece bridge rectifier.
Applied:
How much power do we get out of LM3875TF. . .
With 8 ohm speakers:
For 30vdc, we get 42 watts
For 35vdc, we get 57 watts**
With 6 ohm speakers:
For 26vdc, we get 39 watts
For 30vdc, we get 52 watts
For 35vdc, we get 71 watts (use big heatsinks)
With 4 ohm speakers**:
For 26vdc, we get 50 watts
For 30vdc, we get 50 watts
**If using the 48vct (or 24+24) transformer, rectified to 35vdc rails, then please purchase generously large heatsinks for your amplifier, as that can provide a measure of safety in case of "accidental" connection to 4 ohm speakers. Builder's tip: If you favor the 4 ohm drivers, then use the cheap inductor for the woofer and "pad" the mid&tweet. Else, choose a lower voltage transformer.
Meanwhile, back to the power. . .
Click the picture below for an example of one super-easy method to connect an amplifier board with the pair of 470uF, pair of 0.1uF, and the LM3875TF to the power supply.
Hi Chub,
if your amp is designed to deliver a maximum of 60W into 4r0 then a pair of bridged can deliver a maximum of 120W into 8r0.
But to get 60W into 4r0 requires 21.9Vpk across the 4r0 resistive load.
That results in 5.5Apk into the resistive load. Well within the capacity of a well cooled 3886 with it's spike protection remaining inactive.
Now change the load to a reactive speaker.
The peak current, into a severe speaker, for that same 21.9Vpk can be up around 15Apk.
The 3886 cannot deliver that level of peak transient current.
The end result is that a bridged pair cannot deliver 120W into an 8ohm reactive load without risking triggering the protection and/or overheating the chipamp.
High power levels are better suited to discrete output stages.
if your amp is designed to deliver a maximum of 60W into 4r0 then a pair of bridged can deliver a maximum of 120W into 8r0.
But to get 60W into 4r0 requires 21.9Vpk across the 4r0 resistive load.
That results in 5.5Apk into the resistive load. Well within the capacity of a well cooled 3886 with it's spike protection remaining inactive.
Now change the load to a reactive speaker.
The peak current, into a severe speaker, for that same 21.9Vpk can be up around 15Apk.
The 3886 cannot deliver that level of peak transient current.
The end result is that a bridged pair cannot deliver 120W into an 8ohm reactive load without risking triggering the protection and/or overheating the chipamp.
High power levels are better suited to discrete output stages.
Here's the "math-free" version:
Bridged can increase the demand on the chip amp. There's a built in limiter that doesn't change. Hit that limit with an authentic National Semiconductor chip and you get "ker-thunk" a moment of silence. Hit the limit with a "clone" chip and you get smoke.
Parallel doesn't necessarily, and doesn't often, increase the output.
So, if you want to double your power with a chipamp, purchase speakers drivers that are 3db more efficient than your current speakers.
What? That's right, double the power is only good for a tiny little 3db boost, and that's much easier to do with speakers.
Remember, speakers are the output, not the amplifier.
Plan for beginners:
Shop the prosound market for efficient speakers or shop for efficient speaker drivers and build your own.
For every 3db greater speaker efficiency--that's the same as if you doubled your amplifier power.
Have a look at my comment on turning the Pioneer B20 into a Harbeth Monitor 40, by simply adding a highly efficient prosound woofer.
Also, look at this: http://www.partsexpress.com/projectshowcase/indexn.cfm?project=MagnaCumLaude
Loud can be pretty when it doesn't have the crossover point at the ear's most sensitive area. See Fletcher-Munson chart attached, and don't put your crossover at the ear's sensitivity peak.
Loud can also harm your ears. Whenever I hear a request for more loudness or more compression, or more, more, more of something, then I think that the system has a distraction that hinders transparency. This distraction could be either a distortion or a frequency response issue. And, frequency response is most probable, given compoenent variety, ear variety, and at least 12db of tomfooerly from differences in room response.
LOL!! "12db of tomfoolery"!! That cracks me up! Well, that's what rooms can do to otherwise perfect audio equipment.
Hey, look at this chart and don't stick crossovers on the peak. . . don't allow "distractions" or "distortions" out of your audio system. . . and then you don't "need" your audio to be so loud.
Edit: After that, if you still want it loud, check out the concert-worthy speakers with high efficiency.
Bridged can increase the demand on the chip amp. There's a built in limiter that doesn't change. Hit that limit with an authentic National Semiconductor chip and you get "ker-thunk" a moment of silence. Hit the limit with a "clone" chip and you get smoke.
Parallel doesn't necessarily, and doesn't often, increase the output.
So, if you want to double your power with a chipamp, purchase speakers drivers that are 3db more efficient than your current speakers.
What? That's right, double the power is only good for a tiny little 3db boost, and that's much easier to do with speakers.
Remember, speakers are the output, not the amplifier.
Plan for beginners:
Shop the prosound market for efficient speakers or shop for efficient speaker drivers and build your own.
For every 3db greater speaker efficiency--that's the same as if you doubled your amplifier power.
Have a look at my comment on turning the Pioneer B20 into a Harbeth Monitor 40, by simply adding a highly efficient prosound woofer.
Also, look at this: http://www.partsexpress.com/projectshowcase/indexn.cfm?project=MagnaCumLaude
Loud can be pretty when it doesn't have the crossover point at the ear's most sensitive area. See Fletcher-Munson chart attached, and don't put your crossover at the ear's sensitivity peak.
Loud can also harm your ears. Whenever I hear a request for more loudness or more compression, or more, more, more of something, then I think that the system has a distraction that hinders transparency. This distraction could be either a distortion or a frequency response issue. And, frequency response is most probable, given compoenent variety, ear variety, and at least 12db of tomfooerly from differences in room response.
LOL!! "12db of tomfoolery"!! That cracks me up!
Well, that's what rooms can do to otherwise perfect audio equipment. Hey, look at this chart and don't stick crossovers on the peak. . . don't allow "distractions" or "distortions" out of your audio system. . . and then you don't "need" your audio to be so loud.
Edit: After that, if you still want it loud, check out the concert-worthy speakers with high efficiency.
Attachments
I forgot to mention, albeit with so much typing it seems unlikely that I left something out. . .
Here it is:
With the LM3875 producing roughly 50 watts, this brings most 8" and 10" drivers near or to xmax, while having already overcome the limits on most 6" drivers.
In any case, this is approximately the fidelity limit of most speakers today, so more power isn't appreciable.
To get more power handling so you can use a high power amplifier, then that takes a concert style speaker, and here's the really ironic part:
Those concert style speakers are so much more efficient that you don't need to replace the amplifier--the LM3875 still works just fine. Remember, speakers are the output.
I wish that AndrewT would come back and tell us how much amplifier power it takes to double the actual loudness. Its certainly an amazing amount.
Here it is:
With the LM3875 producing roughly 50 watts, this brings most 8" and 10" drivers near or to xmax, while having already overcome the limits on most 6" drivers.
In any case, this is approximately the fidelity limit of most speakers today, so more power isn't appreciable.
To get more power handling so you can use a high power amplifier, then that takes a concert style speaker, and here's the really ironic part:
Those concert style speakers are so much more efficient that you don't need to replace the amplifier--the LM3875 still works just fine. Remember, speakers are the output.
I wish that AndrewT would come back and tell us how much amplifier power it takes to double the actual loudness. Its certainly an amazing amount.
AndrewT said:
Yes, much learning. I did finally take your advice about Ci. You said that I might find the results interesting.
The results have been very interesting indeed.
On the amplifier power thing, I was thinking of one of the more "awesome" examples where it takes a breaker blowing amount of power to do much less than expected. I just can't remember the example.
AndrewT said:
Ah, no it was a more simple example, something like requring 10x the power for every. . . doubling of loudness? After all, 3db isn't twice as loud to the ear. Its barely noticable.
The example (that I'm trying to remember), could blow one of our little 10 amper mains breakers here in the U.S. For instance, 500 watts per channel stereo will probably knock out the lights.
Edit: The point is, if you want loud, do it with speaker efficiency, or it might get dark before it gets loud.
Might?
darn! the picture... i'm taking weeks to get what the hell is that.
Thts why i turn down the mid tonal level ( 3band tonal ) when play the music louder.
thats why we take more hundreds wattage to awake the lower hz ( subwoofer ). thats why some friends make a little delay for their mid and high.
sorry daniel, i don't get it last time.
you better tell 'the other indonesian' in his thread about this.
Brgds.
eketehe said:
Bass guitarists rarely play from underneath sofas.
Modern speaker stands remove space that was previously available for woofers.
The traditional large 3-way actually consumes less square inches of real-estate (because there's no seperate subwoofer), there's less wires to run, and, of course, the number of woofers is twice as much as a single subwoofer.
Double the number of speakers is one way to double the output.
So, if after our 50w per channel LM3875TF project (a 100w stereo amp)--if after that, you'd like more impact, then have a look at your speakers, which are the output.
Hi,
I'm testing the first channel. I am copying this one: http://dogbreath.de/Chipamps/GainCardCopy/GainCardCopy.html
Without connecting any source I put a 10 ohms resistor as a bridge on the speaker terminals and put the multimeter in DC reading (V=) on 200m and I am getting a result of: 87.4 . I suppose these are milli volts.
I have read that it should not be more than 30mv .
Could the test be wrong or I need to check all the connections again?
Many thanks
Antonio
I'm testing the first channel. I am copying this one: http://dogbreath.de/Chipamps/GainCardCopy/GainCardCopy.html
Without connecting any source I put a 10 ohms resistor as a bridge on the speaker terminals and put the multimeter in DC reading (V=) on 200m and I am getting a result of: 87.4 . I suppose these are milli volts.
I have read that it should not be more than 30mv .
Could the test be wrong or I need to check all the connections again?
Many thanks
Antonio
A Sanchez said:
As we discussed earlier, but now in reference to http://dogbreath.de/Chipamps/GainCardCopy/GainCardCopy.html
Change RG to 15k
In english, change your input impedance to 15k.
Reason why:
The dc blocker has isolated the potentiometer so that it cannot contribute to the load--so you need a stronger load.
Reference the quote and response here, at post#131 on this same thread: http://www.diyaudio.com/forums/showthread.php?postid=1498746#post1498746
The Dogbreath is a mixed AC & DC coupled amplifier.
It makes no attempt to minimise input offset voltage with the result that output offset voltage tends to be higher than necessary.
Ri//Rf~=Rg, but this is silly to try to obtain in a mixed coupled amplifier.
Add a cap in series with Ri to convert the amplifier to AC coupled.
Then you match Rf~=Rg.
It makes no attempt to minimise input offset voltage with the result that output offset voltage tends to be higher than necessary.
Ri//Rf~=Rg, but this is silly to try to obtain in a mixed coupled amplifier.
Add a cap in series with Ri to convert the amplifier to AC coupled.
Then you match Rf~=Rg.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.