OK, i have an old Cambridge Audio amp with one bust channel. It has a pretty hefy transformer, but its not a toroidal. Think its just a std E/I. Ive heard bad things about using them for gainclones. But if it works for that amp and i pay attention to the grounding surely it should work for a couple of LM3875s or whatever. Oh yeh, and im assuming the already insulated version of chips is best to use and just connect to a suitable heatsink/large lump metal?
Hi,
there´s no reason to not use the transformer you have on hand.
The bigger it is the better😉 , but watch out for your voltage rails you get after the rectifier+caps. You shouldn´t exceed the +-42V.
Cheers
Jens
there´s no reason to not use the transformer you have on hand.
The bigger it is the better😉 , but watch out for your voltage rails you get after the rectifier+caps. You shouldn´t exceed the +-42V.
Cheers
Jens
????
you shouldn't exceed 42 volts? you mean after the bridges?
i have 44 volts before and the thinghy just works fine, with each 3875 TF on a heatsink from apexjr.com 8 by 4 by 1/2
they get about 63 volts in.
J-P
you shouldn't exceed 42 volts? you mean after the bridges?
i have 44 volts before and the thinghy just works fine, with each 3875 TF on a heatsink from apexjr.com 8 by 4 by 1/2
they get about 63 volts in.
J-P
Uvodee, please note that this was a "recommendation" from me
and is in fact the absolute maximum rating of the device given from National Semiconductor.
As a fearless DIY´er you can do what you want but I certainly won´t advise anybody to exceed those ratings in the first place.
Exploding IC´s are not nice by the way.
In addition there is a thread here on the forums where somebody experienced disgusting sound due to voltage rails over the given specs so....
Rotellian, do you know the secondary AC voltage of your transformer?
Cheers
Jens
and is in fact the absolute maximum rating of the device given from National Semiconductor.
As a fearless DIY´er you can do what you want but I certainly won´t advise anybody to exceed those ratings in the first place.
Exploding IC´s are not nice by the way.
In addition there is a thread here on the forums where somebody experienced disgusting sound due to voltage rails over the given specs so....
Rotellian, do you know the secondary AC voltage of your transformer?
Cheers
Jens
Dont know yet because ive not had a chance to measure. I will do so. I would think it likely that its within the range needed
I know but does
Joensd knows that?
Thank you for your moral support though!
The little thingies have been operating (without power down) for about 2 and 1/2 weeks now and no sign of overheating or any other malfunction.
Still looking for nice enclosures.
J-P
Joensd knows that?
Thank you for your moral support though!
The little thingies have been operating (without power down) for about 2 and 1/2 weeks now and no sign of overheating or any other malfunction.
Still looking for nice enclosures.
J-P
+42-(-42)=84 BTW 😉
Come on the circuitry of that chip can be simple enough to skip that output cap.
Cheers
Jens
Thanks for the replies guys. I do have a question tho. Reading a lot of information leaves me a little confused as to somethings. I understadn that i need 2 gnds 1 for signal and 1 for power. and that these should be conected. The IC must be electrically insulated from the heatsink.
My question is about the conenction (in non-inverted config) to ground from pin 2 (LM3875) thru a resistor and cap. (1K and 0.22uf) Which gnd is this connected to and is this conencted to the heatsink? (ive seen references to this somewhere) Apologies for the stupid questions, im just trying to make sure ive got everything sorted before i proceed.
My question is about the conenction (in non-inverted config) to ground from pin 2 (LM3875) thru a resistor and cap. (1K and 0.22uf) Which gnd is this connected to and is this conencted to the heatsink? (ive seen references to this somewhere) Apologies for the stupid questions, im just trying to make sure ive got everything sorted before i proceed.
oh drat. There goes my over enthusiasm getting the better of me again. 🙂 Looking at the datasheet, i beleive i mean pin 8.
In that case, try just the resistor connected between pin 8 and the signal star ground. (One less cap = more 'on the cheap' kudos 😉 )
You might wanna have a look here in the forums and search for "star ground".
High current ground leads definitely have to be connected directly to the star ground (like the speaker ground).
Another recommendation would be to look for others that finished their gainclones (for example the never-ending "this is not just another gainclone"-thread) cause there´s not only one approach to it but a lot of suggestions.
Quite a few people used remote power supplies with "not much" of filtering in remote power supply but 2*1000uF capacitors in the amp enclosure etc. so...
Good luck and loads of fun
Jens
High current ground leads definitely have to be connected directly to the star ground (like the speaker ground).
Another recommendation would be to look for others that finished their gainclones (for example the never-ending "this is not just another gainclone"-thread) cause there´s not only one approach to it but a lot of suggestions.
Quite a few people used remote power supplies with "not much" of filtering in remote power supply but 2*1000uF capacitors in the amp enclosure etc. so...
Good luck and loads of fun
Jens
Thanks for the help guys - much appreciated. 🙂 Yeh, not that im a cheapskate or anything, just call me scrooge. Yeh ive got quite a few parts lying about (panny FC 1000uf etc) - mostly just need some carbon resistors. one final thing, im assuming the heatsink should be grounded? and to the signal gnd?
If this scrooge-clone works i might consider building something a bit nicer.
If this scrooge-clone works i might consider building something a bit nicer.
The Panasonic FC caps will work just fine. That's a popular choice for value builders. I used them for mine.
Also, if you use the insulated 3875's, you shouldn't have to worry about grounding the heatsinks. But it wouldn't hurt anything to have them at ground and would be a little safer, IMO.
Also, if you use the insulated 3875's, you shouldn't have to worry about grounding the heatsinks. But it wouldn't hurt anything to have them at ground and would be a little safer, IMO.
Yeh thats what i thought - but i cant find the insulated package anywhere in the UK. Perhaps i should just order some samples. Interestingly i think i might be able to use the existing heatsink and mountings in the cambridge amp. The whole thing might just work.
Rotellian:
I couldn't find the TF (insulated) version of the chip in the UK either. Just use a small insulating thermal pad (from Farnell) between the chip and h/sink. Also, don't forget (as I did) to isolate the bolt with a plastic bush (also Farnell).
Piero
PS: Very tricky to get NS samples in the UK!
I couldn't find the TF (insulated) version of the chip in the UK either. Just use a small insulating thermal pad (from Farnell) between the chip and h/sink. Also, don't forget (as I did) to isolate the bolt with a plastic bush (also Farnell).
Piero
PS: Very tricky to get NS samples in the UK!
yeh thats what i was intending to do (although some of the existing mounting gear in te amp may prove useful) I only took a quick look earlier but the T0-220 thermal pad looked too small for the LM3875. Which ones did u use?
Cheers
Cheers
Part no 681090 is just about big enough. Otherwise you can buy a whole sheet to cut up, but that will cost you!
- Status
- Not open for further replies.