"I have already built a normal 3886 amp, and want to try the new design. Is the circuit I attached before correct?"
Couldn't possibly work. Download the LM3886 notes from National Semiconductor and copy the circuit for a non-inverting amp. Add 1R0 resistors to the supply pins. After you get that working properly then add the booster transistors.
"I built this some months ago at work,my memory has played tricks on me!,it was into a 2R load of a 100 volt line transformer via bipolar coupling caps."
You would have been much better off with a single LM1875 with booster transistors in the supply pins and running off a single supply voltage with a single polarized output cap.
Been there, done that.
Couldn't possibly work. Download the LM3886 notes from National Semiconductor and copy the circuit for a non-inverting amp. Add 1R0 resistors to the supply pins. After you get that working properly then add the booster transistors.
"I built this some months ago at work,my memory has played tricks on me!,it was into a 2R load of a 100 volt line transformer via bipolar coupling caps."
You would have been much better off with a single LM1875 with booster transistors in the supply pins and running off a single supply voltage with a single polarized output cap.
Been there, done that.
soundNerd,
You really should learn to use the information given in the data sheets, and in application notes.
For the LM1875, power supply voltages of +-24 V are great for an 8 Ohm load. According to the Power Output vs. Supply Voltage curve, you should get about 25 W using +-25 V supplies. However, using +- 25 V supplies with a 4 Ohm load creates a lot of excess heat which may result in thermal shutdown (see Power Dissipation vs. Power Output and Notes 3 and 4). A 2 ohm load is simply out of the question. You can't do it, especially not with +- 24 V supplies.
So you have a 4 ohm speaker, you want to use two LM1875's, and according to the above, you should have each LM1875 drive an 8 ohm load. So build a parallel amp, not a bridged amp. See National's AN1192 application note. Since each of the amps will output about 25 W, into 8 ohms, you will get 50 W into 4 ohms. Building a parallel amp is actually quite easy; you basically build two identical amps, then connect a 0.1 ohm resistor between the output of each and and the speaker (and yes, the resistors are absolutely necessary). The two amps can be either inverting or non-inverting; just make sure that they are the same.
You really should learn to use the information given in the data sheets, and in application notes.
For the LM1875, power supply voltages of +-24 V are great for an 8 Ohm load. According to the Power Output vs. Supply Voltage curve, you should get about 25 W using +-25 V supplies. However, using +- 25 V supplies with a 4 Ohm load creates a lot of excess heat which may result in thermal shutdown (see Power Dissipation vs. Power Output and Notes 3 and 4). A 2 ohm load is simply out of the question. You can't do it, especially not with +- 24 V supplies.
So you have a 4 ohm speaker, you want to use two LM1875's, and according to the above, you should have each LM1875 drive an 8 ohm load. So build a parallel amp, not a bridged amp. See National's AN1192 application note. Since each of the amps will output about 25 W, into 8 ohms, you will get 50 W into 4 ohms. Building a parallel amp is actually quite easy; you basically build two identical amps, then connect a 0.1 ohm resistor between the output of each and and the speaker (and yes, the resistors are absolutely necessary). The two amps can be either inverting or non-inverting; just make sure that they are the same.
The application for the LM1875 which I built at work had to produce a minimum of 60 W RMS into a 100 volt line system from (single supply of) 24volts supplied by a lead acid accumulator, these requirements were rigid. The company has large stocks of LM1875 from other product lines hence I used four of them in parrallel bridge mode to give the required power.Up to 10% distortion was ok for this application (not my choice !).
It was whilst "messing about" with it that I discovered just how much power and low THD it was capable of.
I hope this explains things to the likeing of certain contributers to this thread...
I could explain how I got it stable and reduced distortion by a slightly unusual means but I'd probably be told It won't work so I wont bother.
Jez.
It was whilst "messing about" with it that I discovered just how much power and low THD it was capable of.
I hope this explains things to the likeing of certain contributers to this thread...
I could explain how I got it stable and reduced distortion by a slightly unusual means but I'd probably be told It won't work so I wont bother.
Jez.
O.K Greg here goes. (why do I think I'm about to be told "we all knew that already", remember I've only been on this site for a few days!).
I connected the power to each of the four chips' + & - psu inputs via 0.1R 3W resistors without any decoupling after the resistors and omitted the output 0.1R resistors, the result was distortion approx halved and it became stable enough to power up via 4 metre leads with only 220uF decoupling at the end of the leads, it needed short leads and thousands of uF before this.Remember though I had to use output caps in this application (company practice) and have not tried this for hi-fi use.
Hope this is of some use and if it is original.... remember where you heard it first
.
Jez.
I connected the power to each of the four chips' + & - psu inputs via 0.1R 3W resistors without any decoupling after the resistors and omitted the output 0.1R resistors, the result was distortion approx halved and it became stable enough to power up via 4 metre leads with only 220uF decoupling at the end of the leads, it needed short leads and thousands of uF before this.Remember though I had to use output caps in this application (company practice) and have not tried this for hi-fi use.
Hope this is of some use and if it is original.... remember where you heard it first
.Jez.
jez,
Thanks for sharing your personal findings with the crowd.
I am an absolute newbie in Gainclones, however had a lot of prior soldering and upgrading experience.
Can you maybe picture your thoughts in a diagram, say "a proposed schematic of a paralleled four LM1875 chip gainclone".
I personally need the thing to power my 2 12'' 4Ohm subs (i can build 2 gainclones for each sub). Already got two nice heatsinks/chassis, and about to get other parts (only found LM1875 in a local store, so thats what i am going to use)
I expect each GC to put out 80-100W in 4Ohms.
If an adequate schematic had been proposed already, please post the link, to make things easier.
Thanks
Alex
Thanks for sharing your personal findings with the crowd.
I am an absolute newbie in Gainclones, however had a lot of prior soldering and upgrading experience.
Can you maybe picture your thoughts in a diagram, say "a proposed schematic of a paralleled four LM1875 chip gainclone".
I personally need the thing to power my 2 12'' 4Ohm subs (i can build 2 gainclones for each sub). Already got two nice heatsinks/chassis, and about to get other parts (only found LM1875 in a local store, so thats what i am going to use)
I expect each GC to put out 80-100W in 4Ohms.
If an adequate schematic had been proposed already, please post the link, to make things easier.
Thanks
Alex
jez said:O.K Greg here goes. (why do I think I'm about to be told "we all knew that already", remember I've only been on this site for a few days!).
I connected the power to each of the four chips' + & - psu inputs via 0.1R 3W resistors without any decoupling after the resistors and omitted the output 0.1R resistors, the result was distortion approx halved and it became stable enough to power up via 4 metre leads with only 220uF decoupling at the end of the leads, it needed short leads and thousands of uF before this.Remember though I had to use output caps in this application (company practice) and have not tried this for hi-fi use.
Hope this is of some use and if it is original.... remember where you heard it first
Jez.
Thanks Jez.
Interesting. It's good to know. And I didn't know that.
The interesting/fun part of all that is that it makes me think "why" (I don't question the end results).
Can't wait
It would be real nice if everyone can combine their knowledge/experiences with GC to make a close to ultimate design. Like, for example, there was a thread about different grounding schemes and people reporting their findings. This is all good, but IMO needs to be combined together in a single schematic (i am probably dreaming), so us noobs won't have to feel dumb when the people in-the-know talk about the "rocket science" of lowering THD, for instance. Schematics with later different variations, as there would always be differences in theoretical of subjective opinions in implementing the circuit.
Thanks
Alex
I have already decided on the LM3886, and the LM1875 is already out of question, mostly because National doesn't offer an insulated version of it.
Actually, I learned something new. I thought that parallel didn't increase wattage, just current. But I am still using the LM3886 not the 1875.
Why are the power supply resistors so important? I didn't use them on my first 3886 amp and it works fine.
-Mike
Actually, I learned something new. I thought that parallel didn't increase wattage, just current. But I am still using the LM3886 not the 1875.
Why are the power supply resistors so important? I didn't use them on my first 3886 amp and it works fine.
-Mike
I know this is not strictly on thread but is there any one with better IT skills than myself who could assist in trying to send the diagram for this(*&^%$£ amplifier!?.
I've been trying for two hours now.the problem is that any file format I try to use that is of sufficient resolution is too big to post on the site. pdf would be easy but is not allowed , also, I need to scan the image first to remove commercially sensitive information via editing software.I keep ending up with files of 0.8-1.5 Mb and the site allows a max of 102k....
As people will have realised by now computers are'nt my strong point
Jez.
I've been trying for two hours now.the problem is that any file format I try to use that is of sufficient resolution is too big to post on the site. pdf would be easy but is not allowed , also, I need to scan the image first to remove commercially sensitive information via editing software.I keep ending up with files of 0.8-1.5 Mb and the site allows a max of 102k....
As people will have realised by now computers are'nt my strong point
Jez.
jez said:I know this is not strictly on thread but is there any one with better IT skills than myself who could assist in trying to send the diagram for this(*&^%$£ amplifier!?.
I've been trying for two hours now.the problem is that any file format I try to use that is of sufficient resolution is too big to post on the site. pdf would be easy but is not allowed , also, I need to scan the image first to remove commercially sensitive information via editing software.I keep ending up with files of 0.8-1.5 Mb and the site allows a max of 102k....
As people will have realised by now computers are'nt my strong point
Jez.
When you scan it use black/white setting not photo quality. It'll give you much smaller size. And use a jpg or gif file format too (more compression).
Good luck
GregGC is right!
alternatively, if you are fead up with fiddling with it, zip the file up and send it to me as an attachment to: ababkin_NOSPAM_@yahoo.com (remove the _NOSPAM_ prat from the address), and i will take care of it and find way to post it on the forum
Alex
alternatively, if you are fead up with fiddling with it, zip the file up and send it to me as an attachment to: ababkin_NOSPAM_@yahoo.com (remove the _NOSPAM_ prat from the address), and i will take care of it and find way to post it on the forum
Alex
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.