Finished an amp using lm3875. Happy. One chip overheated. Considering options for replacement- so looking at this debate. Again!
as far as i can see -
3886 is capable of higher current so better if using 4ohm speakers. Lower distortion (according to datasheet). Allegedly fuller/more bass??
3875 has no mute, less connections, so simpler to build. Lower noise (according to datasheet). Allegedly more ppl prefer the sound.
In the end i think i will go for one that's cheapest!!
3875TF - £7.54
3875T - not stocked
3886 - £5.60
then i find the LM3876 - £4.48 i better check out that one now
as far as i can see -
3886 is capable of higher current so better if using 4ohm speakers. Lower distortion (according to datasheet). Allegedly fuller/more bass??
3875 has no mute, less connections, so simpler to build. Lower noise (according to datasheet). Allegedly more ppl prefer the sound.
In the end i think i will go for one that's cheapest!!
3875TF - £7.54
3875T - not stocked
3886 - £5.60
then i find the LM3876 - £4.48 i better check out that one now
Okay take your favorite lm3886 circuit and your favorite lm3875 circuit and just swap the chips and then report which one sounds the best original or just the chips swapped.
And make sure it is a blind test too boot!
So far nearly every schematic I have seen uses different component values for the same circuit configuration.
So which one sounds the best?
If you can actually tell the difference that is more than subtle.
Also make sure that each test done is at the same power level as the other one.
And maybe a set of tests at several different power levels as well. jer
And make sure it is a blind test too boot!
So far nearly every schematic I have seen uses different component values for the same circuit configuration.
So which one sounds the best?
If you can actually tell the difference that is more than subtle.
Also make sure that each test done is at the same power level as the other one.
And maybe a set of tests at several different power levels as well. jer
Last edited:
Honestly I don't have one yet.
But I do have some lm4780's that I am going to build either two BPA800's or two PA800's to power my esl's with.
Considering the nature of an esl I'm thinking paralleled is the way togo.
But since the transformers I have are 35v with no center tap I will have togo with a BPA configuration for now or else build a regulator to drop the extra 7v. jer
But I do have some lm4780's that I am going to build either two BPA800's or two PA800's to power my esl's with.
Considering the nature of an esl I'm thinking paralleled is the way togo.
But since the transformers I have are 35v with no center tap I will have togo with a BPA configuration for now or else build a regulator to drop the extra 7v. jer
That does not compute. If you use a 35 V transformer for single supply that corresponds to 48 V which is the same as ±24 V without load. That is not enough to get 68 W from each IC.
Not that it makes much difference whether you have 400 W or 544 W to destroy your ears and your speakers.
Yes ,it comes to 49.35v exactly which would be fine for a bridged amp.
But for parallel amp, it requires a bipolar supply and it would be to much voltage.
It would have to be regulated down by 8 volts.
I have been working with and recording rock and roll bands for a number of years and I do like it very loud occasionaly.
But as far as the high power goes, my esl's present a load as low as 1 to 2 ohms and some musical transients can be as high as +20db (100 times) over the average music power.
By monitoring the signal with a scope I have found this to be very true.
So it is not just a matter of rms power and how loud it goes (somewhat).
I have 10 of these 200 watt transformers and I got them for $20 a piece and it is what I have as I don't have the funds to by any larger ones at the moment (+$400).
I already have every thing I need minus a few more chips to finnish the project. jer
But for parallel amp, it requires a bipolar supply and it would be to much voltage.
It would have to be regulated down by 8 volts.
I have been working with and recording rock and roll bands for a number of years and I do like it very loud occasionaly.
But as far as the high power goes, my esl's present a load as low as 1 to 2 ohms and some musical transients can be as high as +20db (100 times) over the average music power.
By monitoring the signal with a scope I have found this to be very true.
So it is not just a matter of rms power and how loud it goes (somewhat).
I have 10 of these 200 watt transformers and I got them for $20 a piece and it is what I have as I don't have the funds to by any larger ones at the moment (+$400).
I already have every thing I need minus a few more chips to finnish the project. jer
Yes ,I see that they have very good prices and would rather get a couple,but that is impossibile until my economy gets better.
However I am thinking of feeding them with a SMPS circuit to boost the power handling.
Adding another secocdary winding or two or three to handle the current demand would be very easy to do for a VA rating of 400watts to 600 watts or so as 800watts may be pushing the primary current to far for these transformers .
Just another idea and that is some thing I will try at a much later time.
But could be done very simply since there is two seperate 120v windings.
Right now I just need to get some thing going, because my current amp is just not cutting the mustard at all by any means.
I know could go with a discrete amp design but I already have the chips and I am very curious to see how well it would work.
Sorry didn't mean to hijack this thread. jer
However I am thinking of feeding them with a SMPS circuit to boost the power handling.
Adding another secocdary winding or two or three to handle the current demand would be very easy to do for a VA rating of 400watts to 600 watts or so as 800watts may be pushing the primary current to far for these transformers .
Just another idea and that is some thing I will try at a much later time.
But could be done very simply since there is two seperate 120v windings.
Right now I just need to get some thing going, because my current amp is just not cutting the mustard at all by any means.
I know could go with a discrete amp design but I already have the chips and I am very curious to see how well it would work.
Sorry didn't mean to hijack this thread. jer
You can use single supply with any configuration.
If the load is 1 Ohm a BPA should ideally have 8 parallel channels. For a 2 Ohm load 4 channels would be sufficient.
You have no money for a transformer, but you do have for a 500+ VA audio grade SMPS?
yes,I know how to setup an opamp for single supply operation.
However this would require rather large output capacitors therefore quite costly and would take up alot of space.
yes,I have had plans try a BPA with a total of 16 devices.
It maybe overkill and a disceret design might be a better choice for over 600watts.
Although I am very curious to find out if and how well something like that would work.
No,but I do have lots of dead computer supplies laying around and a switcher to drive one of my transformers cores could be made very easly.
jer
However this would require rather large output capacitors therefore quite costly and would take up alot of space.
yes,I have had plans try a BPA with a total of 16 devices.
It maybe overkill and a disceret design might be a better choice for over 600watts.
Although I am very curious to find out if and how well something like that would work.
No,but I do have lots of dead computer supplies laying around and a switcher to drive one of my transformers cores could be made very easly.
jer
Some stomach feeling tells me that normal power transformers are not the optimal choice for a switching supply. To transform a square wave at a switching frequency in the tens to hundreds of kHz range requires at least a ferrite core. Many people even use powder cores, though others say that ferrite cores are the best choice.
The torriodal power transformers that I have ,have shown some results with square waves .
The best I have ever seen before compared to an E-I type.
With just my whimpy amp driving a 16v winding with a 60v peak to peak 10khz square wave I get a good 400v peak to peak square wave on a 120v winding.
After 10khz then slewing distortion from the amplifier sets in.
But what goes in ,comes out as far as waveshape.
I have measured the response of this transformer to be good to 1mhz with two peaks at around 310khz and 620khz or so using just the signal generator.
1mhz is the limit of my generator using using a sine or square wave.
The square wave test was good to at least 300khz but I remember that 50khz to 100khz was almost perfect,this was what sparked the idea.
I can easly retest and post some pictures if you like. jer
The best I have ever seen before compared to an E-I type.
With just my whimpy amp driving a 16v winding with a 60v peak to peak 10khz square wave I get a good 400v peak to peak square wave on a 120v winding.
After 10khz then slewing distortion from the amplifier sets in.
But what goes in ,comes out as far as waveshape.
I have measured the response of this transformer to be good to 1mhz with two peaks at around 310khz and 620khz or so using just the signal generator.
1mhz is the limit of my generator using using a sine or square wave.
The square wave test was good to at least 300khz but I remember that 50khz to 100khz was almost perfect,this was what sparked the idea.
I can easly retest and post some pictures if you like. jer
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.