UnixMan said:
given that such caps seems to be very good and are not so expensive, I would use the panasonic FC for all the electrolytic caps... unless Mauro as any objection about it, of course!
Hey UnixMan,
Yes, Brian and I discussed this a few days back, and we both agreed unless there was some reason not to do it (not likely) then using good low ESR caps all the way around was a great idea.
I fully agree, in fact all of the EL caps in my prototype except those for the speaker protection circuit are low ESR. (Panasonic, Elna, and Nichicon).Cheers!
Russ
I have just done a few mods - (more in favour of personal taste) replaced the 220pf with 330pf. and the feedback cap bypassed with a 100nf cap. i can now say that it matched the type of sound produced by my JLH. little bit more layed back and richer in the mids. (i am not saying this i the most linear but it suits my taste)
One experiment still to do is to create a bipolar arrangement with the feedback cap. I have found very favourable results with this arrangement in my Class A amps. there is a possiblility that it might work well for this setup. Mauro i would like to hear your thoughts on this kind of arrangement. two 470uf connected + to + to create a 235uf feedback cap or if you can find a 220uf bipolar it solves the problem. (not so easy to find here in SA)
regards
Rudi
One experiment still to do is to create a bipolar arrangement with the feedback cap. I have found very favourable results with this arrangement in my Class A amps. there is a possiblility that it might work well for this setup. Mauro i would like to hear your thoughts on this kind of arrangement. two 470uf connected + to + to create a 235uf feedback cap or if you can find a 220uf bipolar it solves the problem. (not so easy to find here in SA)
regards
Rudi
rudi said:I have just done a few mods - (more in favour of personal taste) replaced the 220pf with 330pf. and the feedback cap bypassed with a 100nf cap. i can now say that it matched the type of sound produced by my JLH. little bit more layed back and richer in the mids. (i am not saying this i the most linear but it suits my taste)
One experiment still to do is to create a bipolar arrangement with the feedback cap. I have found very favourable results with this arrangement in my Class A amps. there is a possiblility that it might work well for this setup. Mauro i would like to hear your thoughts on this kind of arrangement. two 470uf connected + to + to create a 235uf feedback cap or if you can find a 220uf bipolar it solves the problem. (not so easy to find here in SA)
regards
Rudi
Hi Rudi,
I used a bi-polar 220uf for my NFB cap when I built my prototype, and I used a 100nf multilayer ceramic for bypass this morning as well as an MKT for comparison. I think with the 100nf bypass it sounds just as good or better than it did, but I would have to build two Rev Cs and A/B test to describe just how. It is not a large change in any regard. One thing is that it did change my DC offset, now it is less than 1mv.
Cheers!
Russ
What would the ultimate configuration be?
Hi Mauro
if you had to put together the ultimate configuration of your amplifier together and forget about the economical compromise aspect.
how would you configure the power supply (transformer/s and then the bridge rectifier) I would love to hear you views on this. and obviously others aswell
I will give a quick rundown of my idea (not saying it is correct but it is crazy enough)
- 2 transformers with two secondaries
- each amp with two bridge rectifiers
- each transformer suppplying just the negative or just the possitve rail i.e. one seconday going to amp1 possitive and then the other secondary going to amp2 possitive and then the same for the other transformer supplying negative.
this is overkill but sometimes we like to go overboard with our ideas and pays of maybe another 2%
best regards
Rudi
Hi Mauro
if you had to put together the ultimate configuration of your amplifier together and forget about the economical compromise aspect.
how would you configure the power supply (transformer/s and then the bridge rectifier) I would love to hear you views on this. and obviously others aswell
I will give a quick rundown of my idea (not saying it is correct but it is crazy enough)
- 2 transformers with two secondaries
- each amp with two bridge rectifiers
- each transformer suppplying just the negative or just the possitve rail i.e. one seconday going to amp1 possitive and then the other secondary going to amp2 possitive and then the same for the other transformer supplying negative.
this is overkill but sometimes we like to go overboard with our ideas and pays of maybe another 2%
best regards
Rudi
Russ White said:
couple of questions here...
Russ: does your PCB foresee the space+pads for the 100nF bypass capacitor? would be nice to have the option in place...
Brian: as far as I have understood from your posted BOM, the 220uF capacitor in your kit is a polarized one... do you foresee to replace it with a non-polar one?
Mauro: what do you think about these?
(polarized cap or not, and what about bypassing... :-? )
UnixMan said:
I have added a spot for a 5mm bypass cap for the NFB 220uf cap.
That won't be in the beta(since it is already ordered) but it is quite simple to add the bypass cap by soldering it to the leads of the 220uf cap on the bottom of the board.
One question I have is what voltage rating is really necessary for that 220uf cap, I used 50V since I had some, but those are big, the 25V and 16V Nichicon Muse radials fit much better.
Cheers!
Russ
Hi all,
220uF cap:
to depart my personal experience ( neutral on this argues ), the excellent Cyril Bateman's studies on all the types of audio caps ( El. World 2002-2003) says that the verifiable THD on 1 ELcap polar = 2 ser polar = 1 no-polar elcap = high.
To reduce the effect is has to it move cut-off frequence in subsonic zones, use a good lowesr cap and, if want, (for high frequency) a bypass with film cap...
An other alternative it is uses a solid electrolyte ( no tantalium ) very rare and expensive, or only film caps...
The measures that I have done I on the original circuit it says that not there are particular problems to use a "good polar elcap ".
The general criterion to choose a (audio) elcap is able be:
- great dimensions( better precision of construction ), no "miniature" versions
- low esr
- high term grade
- High DC voltage (to parity of assemblage space)
Ciao
Mauro
220uF cap:
to depart my personal experience ( neutral on this argues ), the excellent Cyril Bateman's studies on all the types of audio caps ( El. World 2002-2003) says that the verifiable THD on 1 ELcap polar = 2 ser polar = 1 no-polar elcap = high.
To reduce the effect is has to it move cut-off frequence in subsonic zones, use a good lowesr cap and, if want, (for high frequency) a bypass with film cap...
An other alternative it is uses a solid electrolyte ( no tantalium ) very rare and expensive, or only film caps...
The measures that I have done I on the original circuit it says that not there are particular problems to use a "good polar elcap ".
The general criterion to choose a (audio) elcap is able be:
- great dimensions( better precision of construction ), no "miniature" versions
- low esr
- high term grade
- High DC voltage (to parity of assemblage space)
Ciao
Mauro
Regulated supplies on the LM318
I've read a couple of posts on this HUGE topic and had a look at the data sheet for the LM318.
Given that the 318 has relatively low power supply noise rejection ratios at 10k and above (around 60dB and below), would it not make sense to REGULATE the +- 12V to 15 for the 318 op-amp? Easy enough to try for those that have built the amp.
Perhaps this has been addressed before, but my searches have not brought up anything.
I've read a couple of posts on this HUGE topic and had a look at the data sheet for the LM318.
Given that the 318 has relatively low power supply noise rejection ratios at 10k and above (around 60dB and below), would it not make sense to REGULATE the +- 12V to 15 for the 318 op-amp? Easy enough to try for those that have built the amp.
Perhaps this has been addressed before, but my searches have not brought up anything.
Re: Regulated supplies on the LM318
I have built a version of the REV C amp with LM317/337 regulation at 15V. There is no appreciable benefit.
The zener regulator to 12V is more than adequate and actually does a great job of regulation, it is just not very efficient.
Also because of the PS design there is actually very little PS noise to reject... so its not really an issue at all. this is one of the quietest least noisy amps you could ever build.
Cheers!
Russ
Dr.H said:
I have built a version of the REV C amp with LM317/337 regulation at 15V. There is no appreciable benefit.
The zener regulator to 12V is more than adequate and actually does a great job of regulation, it is just not very efficient.
Also because of the PS design there is actually very little PS noise to reject... so its not really an issue at all.
this is one of the quietest least noisy amps you could ever build.Cheers!
Russ
needlz said:
It is critical. You must provide some sort of heatsink, even if it is just a slab of aluminum or something. Otherwise only bad things can happen. the chip will quickly hit the SPIKE protection, and your chips life will be likely shortened.
If you have an old CPU heatsink or a bar of aluminum or something you could use those for a quick temporary heatsink, but watch it close, when(if) it gets too hot to leave your hand on comfortably switch it off.
Cool. I used the bottom cover of an aluminum hammond enclosure as my temporary heatsink. I also built the powersupply in a seperate case so I can easily check out different amp versions. I don't want to run it too hard until I get a proper heatsink but it sounds pretty damn good right now.
Russ/Brian/Mauro Thanks again for a stellar job in putting this project together. It's as close to a pro kit that a DIYer could do. I got this completely together in less than 2 hours.
cheers,
erik
Russ/Brian/Mauro Thanks again for a stellar job in putting this project together. It's as close to a pro kit that a DIYer could do. I got this completely together in less than 2 hours.
cheers,
erik