Hotrodding the UCD modules

Status
Not open for further replies.
Jump to Latest
#1,625
A question of regulation!

Hi all,
I have been busy educating myself so I can vote intelligently in the up coming election, I would recommend all do the same. Remember the internet is still mostly free! The stakes are very high! Enough politics, sick of it!

I would recommend doing the regulator with an LM431 replacing the zener as it is adjustable and very quiet. Possibly adding the extra parts on a small piece of perf board with short leads to the UcD. Pay close attention to the 431 app notes! The regulator input cap is a no brainer and doesn’t need to be very large to have a very worthwhile effect. Also follow Chris’s suggestions on bypassing the power pins on the opamp. This is particularly important with fast chips and has a great bearing on the overall sound.

On the cap clamp question, I wouldn’t recommend it. The disadvantages would far outweigh any advantages. It would act like an antenna and this could increase the EMI significantly. It would also require more distance between adjacent channels to prevent heterodyning/whistling.
Roger
 
#1,626
Hi T.

Anybody try those LM4562 in the UCD's yet?
Click to expand...

Look what our colleagues at PFM are doing (Andy is there too).
http://www.pinkfishmedia.net/forum/showthread.php?t=26205&page=14

Look at Ron's post #618.

I have mine on the way (I will practice SMD soldering). Appart that, J511 as adviced here and there...and...I found Rubycon ZL 470uF/100VDC. I hope they are because I already bought some:

http://www.referenceaudiomods.com/M...Screen=CTGY&Store_Code=RAM&Category_Code=RUBY

My Autograph enclosures deserve more than stock UCD's 😎

Cheers,
M
 
#1,627
t. said:
Anybody try those LM4562 in the UCD's yet?
Click to expand...

I just replaced the AD8620s I was using for the mid and tweeter channels in my UcD triamped system. In every respect I find the LM4562s to sound better. Like most others who have tried this chip, I find there is more detail yet the sound is generally "more relaxed". :up::up:
 
#1,628
LM4562 question

Downloaded the datasheet from National, these chips have fantastic specifications, should be getting close to a "wire with gain".

I have a question regarding the CRD mod and the use of a separate power supply for the op amp. As these things have >100db PSSR even at 10Khz (significantly more at lower frequencies), and the PSSR is similar for both rails, I assume that the CRD mod and having separate power supplies wont make any difference with a LM4562 fitted?

Anyone care to comment on the theory vs implementation?

Regards,
Dean
 
#1,630
mac said:


I just replaced the AD8620s I was using for the mid and tweeter channels in my UcD triamped system. In every respect I find the LM4562s to sound better. Like most others who have tried this chip, I find there is more detail yet the sound is generally "more relaxed". :up::up:
Click to expand...

Thanks Mac, I managed to get some and fit them too🙂 Just giving them a good burn in
 
#1,635
t. said:


Yes, you need to remove the CRD if its on the -12v supply though
Click to expand...


When I look at the datasheet of the LM4562, it looks like the PSRR is best with a +- 15V supply. At that voltage it is the same for + and - rail and below 100dB, at 12V and 17V it is not as good for the + rail. So I guess, for people who want to really get the best of out it need to operate it with a +-15V supply, I assume the rest of the UcD circuit can handle that +-15V as the early UcD versions used a +-15V supply for the opamps. The specs of the lm4562 look truly amazing, what do they cost and where can you get them (online)?

Best regards

Gertjan
 
#1,636
ghemink said:



When I look at the datasheet of the LM4562, it looks like the PSRR is best with a +- 15V supply. At that voltage it is the same for + and - rail and below 100dB, at 12V and 17V it is not as good for the + rail. So I guess, for people who want to really get the best of out it need to operate it with a +-15V supply, I assume the rest of the UcD circuit can handle that +-15V as the early UcD versions used a +-15V supply for the opamps. The specs of the lm4562 look truly amazing, what do they cost and where can you get them (online)?

Best regards

Gertjan
Click to expand...

Hi Gertjan,
I think it should be ok to replace the 12v zeners to 15v on the ucd pcb or just run separate supplys

The LM4562's could be had as samples only from National semicondutor, I don't know if they are available to buy yet.
Be interestig to know what you guys think, I got sick of waiting so managed to get a couple to try myself😀
 
#1,637
t. said:


Hi Gertjan,
I think it should be ok to replace the 12v zeners to 15v on the ucd pcb or just run separate supplys

The LM4562's could be had as samples only from National semicondutor, I don't know if they are available to buy yet.
Be interestig to know what you guys think, I got sick of waiting so managed to get a couple to try myself😀
Click to expand...

Yes, was thinking of replacing the zeners or using an external supply. Just need to find a place to get them, guess have to wait a bit until places like digikey have them.

Thanks

Gertjan
 
#1,638
ghemink said:


for people who want to really get "the best of out it"
Best regards

Gertjan
Click to expand...


Gertjan all other circuitry supplied with + and -12 Volt is CMOS. In principle it can withstand 15 Volt. A far better solution is to disconnect the original supply from the op amp (as we do in our kits) and supply only the analog input circuit(s) with an external supply. In the latter case the voltage could be + and -18 Volt which gives some extra headroom.
 
#1,639
johnrtd said:
Gertjan all other circuitry supplied with + and -12 Volt is CMOS. In principle it can withstand 15 Volt. A far better solution is to disconnect the original supply from the op amp (as we do in our kits) and supply only the analog input circuit(s) with an external supply. In the latter case the voltage could be + and -18 Volt which gives some extra headroom.
Click to expand...
18V is the absolutemaximum rating, which you would want to avoid. It's specified as having a power supply voltage range of +-2.5V to +-17V. Correct me if I'm wrong, but I believe it's wise to steer away from maximum ratings by a fair margin. Apart from that (safemargin), I don't see a reason to go that high since the 8620 with it's 12V supply already has more than enough headroom in this particular application. A more common +-15V seperate psu would be just fine here and leave the option open to try a lot of other opamps as well if you wish.
 
#1,640
ghemink said:
I assume the rest of the UcD circuit can handle that +-15V as the early UcD versions used a +-15V supply for the opamps.
Click to expand...

I failed to mention that my modules are early versions (like yours) and use the 15v zeners without the CRD.

ghemink said:
The specs of the lm4562 look truly amazing, what do they cost and where can you get them (online)?
Click to expand...

When I called Digikey a few weeks back they had absolutely zero knowledge of this chip. As others have mentioned, if you want them now you have to be resourceful. 😉
 
Status
Not open for further replies.
Top Bottom