Nuuk said:Perhaps in the interests of clarity, the mods could edit out the irrelelvent posts?
That would give another discussion...
What to edit, what to cut?
Let it be as it is Nuuk.
I'm sorry, but when someone comes here to talk about things he doesn't know, he didn't try, he doesn't sell, I have to defend myself, for the sake of sanity, because I know what's better when I listen to it.
Guys, don't let this "theorists" fool you.
I'm glad to receive anyone asking questions, trying and later giving his/her impressions about it.
Neal gave it a try, followed my instructions, we exchanged some posts and he reported his oppinion.
This is a positive attitude, he didn't have an oppinion before testing.
I only wish everyone here had this attitude...
Nuuk said:Perhaps in the interests of clarity, the mods could edit out the irrelelvent posts?
Where are the relevant posts? I hardly can find any.
We all only know what we've tried and *nobody* has tried it all. And even that which we've tried is still limited by our own systems, room and ears.
Is it at all surprising that there is disagreement about what is best? Why waste time debating the point? Even people listening to the same system in the same room at the same time will often disagree. How can agreement be reached over the net where we can barely communicate?
Is it at all surprising that there is disagreement about what is best? Why waste time debating the point? Even people listening to the same system in the same room at the same time will often disagree. How can agreement be reached over the net where we can barely communicate?
peranders said:Where are the relevant posts? I hardly can find any.
This thread has a link on the first post to another one, where I report my experiences with this PSU, listening tests, fine-tuning.
This thread has two schematics.
What are you trying to say, P-A?
Search again.
jeff mai said:
I don't want agreement I just want that people who post here have tried and listened to this , or else put the crap talking elsewhere.😡
Is this hard to understand?
Two guys listening to the same system, one likes it, one doesn't.
But they are both LISTENING .
Of course audio is subjective.
Deam, am I clear now?
Some guys have oppinion without knowing.
Others, have oppinion based on what they've read on a magazine.
carlosfm said:
You can't stop people talking. They will say what they like.
Certainly people are taking notice of regulated supplies for chip amps. Eventually they'll make up their own minds in their own time.
Relax and enjoy the tunes and the football.
Thanks for sharing with everyone what you've learned about chip amps (that goes for all the pioneers here on the forum.) It will be a little while, but trying something similar to what you've built is on my list of projects.
Fine, I'm relaxed
Nuuk?
How's your PSU?
Make your tests today.
Don't leave it for tomorrow after the match, something can go wrong. 😀
Nuuk?
How's your PSU?
Make your tests today.
Don't leave it for tomorrow after the match, something can go wrong. 😀
Well if its of any help here is my condensed building tips.
Most of this is repeated elswhere but what the heck, I though I would try and condense some of it.
OK, tips for successful implementation of a regulated GC supply based on LM338 and Pedja's schematic 😀 All notes from my experience and advise from Carlos. YMMV!
Buy a ‘scope!! 😀 Only kidding!
The regulator needs an input voltage 3v higher than output to maintain regulation, I chose a 25-0 0-25v transformer to give ~35v or so before the reg underload. I wanted a 26v output from the reg. I was looking for 35W into a 6Ohm load so 26v seems a good choice for the 3875 when driving a 6Ohm load according to the datasheet.
Select the voltage sense resistor to pass 5mA minimum to keep reg output stable under no load, 120R to 240R. Select the adjust resistor to give your desired o/p voltage by ((R2/R1)+1) * 1.25 IE ((4700/240)+1) * 1.25 = 25.73v
Use 33uF to 100uF 63v low ESR caps on the 3875/3886 directly. By pass these with .1uF 63v box ‘d WIMA polyprop caps. If direct wiring (no PCB 😉 ) couple the cap earthy sides together and return back to PSU star earth with 6amp multistrand wire. One wire per channel.
Use the metal can TO3 LM338, it has better thermal characteristics AFAIK, use a heatsink with insulation washers!
Use the protection diodes, 1N4002, output back to input and Adj to output, note orientation.
I use a 10uF 63v ‘lytic cap to bypass the adjust resistor to gnd, Carlos uses more. The spec sheet says that going over 22uF offers little PSRR improvement than 10uF, it’s your choice.
Wire directly across the 338 input to gnd a poly-something bypass cap, .47uF. Place before this your 4700uF to 10,000uF supply resiviour caps, keep them as close as possible to the regulator, again use 6amp cable, 8-10cm max. wire, less is best.
Common both regulators gnd connections (Adj pin on positive regulator and output pin on negative regulator) and take a single 6amp wire back to PSU star earth. Actually this may not be the best way, two separate wire runs might be best but it works OK for me.
Wire across the regulator output a .47uf to .047uF (47nF) quality film ‘n’ foil cap. Test with a ‘scope and sig gen for stability………..oh sorry I forgot … 😀 😀 Adjust this cap value and type to suit your listening taste.
Connect regulators to chip amps with no more than 10cm of 6amp wire, the closer the better.
Listen to Carlos suggestions 😀 😀 😀
Most of this is repeated elswhere but what the heck, I though I would try and condense some of it.
OK, tips for successful implementation of a regulated GC supply based on LM338 and Pedja's schematic 😀 All notes from my experience and advise from Carlos. YMMV!
Buy a ‘scope!! 😀 Only kidding!
The regulator needs an input voltage 3v higher than output to maintain regulation, I chose a 25-0 0-25v transformer to give ~35v or so before the reg underload. I wanted a 26v output from the reg. I was looking for 35W into a 6Ohm load so 26v seems a good choice for the 3875 when driving a 6Ohm load according to the datasheet.
Select the voltage sense resistor to pass 5mA minimum to keep reg output stable under no load, 120R to 240R. Select the adjust resistor to give your desired o/p voltage by ((R2/R1)+1) * 1.25 IE ((4700/240)+1) * 1.25 = 25.73v
Use 33uF to 100uF 63v low ESR caps on the 3875/3886 directly. By pass these with .1uF 63v box ‘d WIMA polyprop caps. If direct wiring (no PCB 😉 ) couple the cap earthy sides together and return back to PSU star earth with 6amp multistrand wire. One wire per channel.
Use the metal can TO3 LM338, it has better thermal characteristics AFAIK, use a heatsink with insulation washers!
Use the protection diodes, 1N4002, output back to input and Adj to output, note orientation.
I use a 10uF 63v ‘lytic cap to bypass the adjust resistor to gnd, Carlos uses more. The spec sheet says that going over 22uF offers little PSRR improvement than 10uF, it’s your choice.
Wire directly across the 338 input to gnd a poly-something bypass cap, .47uF. Place before this your 4700uF to 10,000uF supply resiviour caps, keep them as close as possible to the regulator, again use 6amp cable, 8-10cm max. wire, less is best.
Common both regulators gnd connections (Adj pin on positive regulator and output pin on negative regulator) and take a single 6amp wire back to PSU star earth. Actually this may not be the best way, two separate wire runs might be best but it works OK for me.
Wire across the regulator output a .47uf to .047uF (47nF) quality film ‘n’ foil cap. Test with a ‘scope and sig gen for stability………..oh sorry I forgot … 😀 😀 Adjust this cap value and type to suit your listening taste.
Connect regulators to chip amps with no more than 10cm of 6amp wire, the closer the better.
Listen to Carlos suggestions 😀 😀 😀
Good news and bad news Carlos.
The faulty section now works and I have soldered in another 47 uF cap.
This is testing the module out of the amp. I notice on the big 10,000 uF caps that they have about five lugs and I am wondering if one of the wires was accidentally touching an extra lug. Why do they put all those extra lugs on the caps when in theory we only need two?
The bad news is that the other section on that channel now only puts out 6 volts! Could I have somehow damaged the LM338?
Meanwhile I am building the rest of the amp as the weather here is more like mid December than mid June and I have little desire to go out on my bike! 🙁
I hope that you are sleeping well and not having too many Rooney nightmares! 😉
The faulty section now works and I have soldered in another 47 uF cap.
This is testing the module out of the amp. I notice on the big 10,000 uF caps that they have about five lugs and I am wondering if one of the wires was accidentally touching an extra lug. Why do they put all those extra lugs on the caps when in theory we only need two?
The bad news is that the other section on that channel now only puts out 6 volts! Could I have somehow damaged the LM338?
Meanwhile I am building the rest of the amp as the weather here is more like mid December than mid June and I have little desire to go out on my bike! 🙁
I hope that you are sleeping well and not having too many Rooney nightmares! 😉
NealG said:
Thanks for summarizing this, Neal, as I think that we've discussed this on another thread that had nothing to do with regulators.😀
Nuuk said:
Pay attention to what wires are you joining to make the ground.
If you separate them, how much voltage do you have on each reg?
Nuuk said:
No, just sweet Ronaldo, Deco and Figo dreams.😀
Nuuk, how are you measuring it?
Is it the negative regulator?
If you measure from ground to the IN pin you should see something like 10v depending on what o/p voltage you set. From Gnd to the bottom of the Adj res should give you your full o/p voltage. Otherwise the 338 id fried
Is it the negative regulator?
If you measure from ground to the IN pin you should see something like 10v depending on what o/p voltage you set. From Gnd to the bottom of the Adj res should give you your full o/p voltage. Otherwise the 338 id fried
I hadn't changed anything Carlos!
I'm measuring between the output pin and the zero volt rail (on the positive regulator side). I've got to pop out and help a friend with a PC problem so will measure again when I come back.
Basically, I have three wires coming of the regulator boards, red = positive rail, blue = zero volt rail as in the circuit diagram and black which is negative voltage rail.
NealG said:Nuuk, I fear it may be friedthen............
I'm now suspecting that too...
Nuuk, change that reg before the match, with steady hands.
Nuuk said:I will have to order another one first Carlos - when I can make up a 20 UKP order for Farnell!
You mean, there's no place near you where you can buy some parts?
Oh, 20 UKP?
You get there with just a couple of LT1083 regs.😱 😀
That's it! It's difficult to buy even a speaker plug where I live. It's a 100 kilometre round trip to buy even a resistor!
Now Maplin are getting out of components supply and Farnell have this silly 20 UKP minimum order, it's getting harder to buy the parts than to build the equipment! 😡
Four off, LT1083? That would be my entire hi-fi budget for the year!
May be it's time for another hobby!
Nuuk said:
That's it! It's difficult to buy even a speaker plug where I live. It's a 100 kilometre round trip to buy even a resistor!
Now Maplin are getting out of components supply and Farnell have this silly 20 UKP minimum order, it's getting harder to buy the parts than to build the equipment! 😡
Four off, LT1083? That would be my entire hi-fi budget for the year!
May be it's time for another hobby!
You are not alone! I have to order every electronic parts online too! Not a single shop around to buy anything.
There are probably not to many people with this hobby around.🙁
We have the headquater of RS-components Norway in town, but that doesn`t help much when they send everything from England and won`t sell anything to private persons!
Hopefully a firend of mine will build an amp in the near future, then I`ll order parts for regulated supply and thoose 3886 chips and we can share the shipping costs.
Tor Martin
There are at least 5 pages of posts i haven't read yet, but...
What about the Physics of microphony? All components are microphonic to some extent...
Malcolm Hawksford wrote a nice paper with all sorts of proper math & physics that gives a good grounding.
dave
geewhizbang said:
What about the Physics of microphony? All components are microphonic to some extent...
Malcolm Hawksford wrote a nice paper with all sorts of proper math & physics that gives a good grounding.
dave
- Status
- Not open for further replies.