~You would have to change the caps on the audio PCB to 25V or more. The headphone section is directly driven from the external power supply instead of from the 12V regulator PSU.
Printable View
~You would have to change the caps on the audio PCB to 25V or more. The headphone section is directly driven from the external power supply instead of from the 12V regulator PSU.
I know the superteddy is a good PSU and better than stock, obviously as you have bought one!Quote:
Originally Posted by MartinT
Is it WORTH it Martin, in bang for buck terms?
If it is..im going to do it right.
To my ears, yes. It isn't cheap to put the whole PSU together with a transformer, rectifiers, smoothing caps etc. in a case. I probably spent £100 on it in total. There's a picture story of it hereabouts.
The best sound I have achieved by far is by building two adjustable linear PSU's, one to deliver 12v and one to deliver 5v. You can then scrap the internal regulators and feed the supplies directly to the reg points.
The twin linear PSU can be built for around £120.00 if you shop around a bit and it can be as simple or as complex as you like.
I posted how to do this a while back with some photos but unfortunately can't find the thread.
I run the Caiman with this and the passive output mod through the variable outputs straight to a power amp - fantastic!
Regards
Chris
This http://theartofsound.net/forum/showthread.php?t=9144Quote:
Originally Posted by chrism
[PSU+GENERAL]
Finally, I bought two Belleson SPJ Super Regulators 12V and 5V respectively...
+ More on this later... :steam:
I have also swapped all tantalum caps to Elna Silmic II and the main reservoir cap to two Elna Silmic II (2x4700uF) (Because of new 16v PSU).
This was the best upgrade I made so far. Imaging improvement is INCREDIBLE
[HEADPHONES]
Also, I swapped the op-amp for 2xLT1028ACN. Simply the best. Stunning!! (Checked against: AD797BR, OPA627, OPA827, AD8066, AD826, AD8022, ADA4627-1B,OPA2107, AD8620,OPA2107,etc.)
If you want to improve the headphones output, just throw away the AD826 for one of these: LT1028ACN, AD797BR, ADA4627-1B period.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
[REGULATORS AND NOISE]
So, when I bought the Bellesons I was very happy because I thought I was going to end forever the background noise story... (110$ pair) -140dB noise spectrum...
I soldered everything up, turned on, and... NOOOOO! The noise was still there!!! I got really upset!!
What was going on? Is it the 12V regulator?? I took the scope and started measuring again:
[BELLESON SPJ 12V]
- Fixed output: 10mVp-p background noise. Same as Murata reg. Too noisy for me
- Variable output: Unmeasurable noise till full volume: 2mVp-p then. PERFECT
- Headphones: measured 100mVp-p at the headphone output at 2-3 o'clock position of volume pot.
This translates to real world in audible background hiss from 10 o'clock position using headphones.
* Measured Dynamic range at headphone output: 90dB
So I decided to conduct more test: I took an LM340T 12V reg and soldered it a Tantalum 22uF cap Input-ground and a 0.1uF cap Ground-Output.
When I replaced the Belleson 12V regulator (59$) with a LM340T (0.76$ regulator) i repeated the test.
[LM340T]
- Fixed output: 7-8mVp-p background noise. Better but not perfect.
- Variable output: Unmeasurable noise till full volume 1mVp-p then. PERFECT AGAIN.
- Headphones: measured 18mVp-p at the headphone output at 2-3 o'clock position of volume pot. WOWWWWW!! What an improvement!!
* Measured Dynamic range at headphone output: >=100dB.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
So! What's going on?? If you use a expensive regulator Caiman+GATOR performance is MUCH WORSE (noise-wise).
Dynamic ranges:
- Murata(18$):80dB
- Belleson SR SPJ(59$):90dB
- LM340T(0.76$): >100dB
Why??? It souldn't be this way ... :scratch: I measured more and more things, but couldn't find a reason! The only repeating pattern I found was that Murata and Belleson output is >12V (12.10V) and LM340T and 7812CV is just below 12V.
Noise measured at the output of the regulatos (Murata, Belleson) is 3-1mVp-p! Where is this noise coming from??
Any ideas???
Depends on what the internal voltage reference of those regulators is like. If it's zener-like, it will regulate well but be noisy. Have you tried a Dexa in the 12V position? I don't get noise at any sensible volume level at all.
Belleson is resistor based and should be clearly 'better' than Dexa! But... it is not with the Caiman!! (http://www.belleson.com/comparenewclassd.php). I do not recommend it to Caiman users
I test noise using a special noise track I made with all spectrum freq. at or below -140dB cause under normal operation relays won't switch unless you play music so you can't test for noise.
Hi nox, seems that you've gone even further with trobuleshooting this noise issue with other than default regulators, I'm really interested once this is sorted out wat is the root cause and fix it for my Muratas :)Quote:
Originally Posted by mcebrian
*EDIT* Hmm, I've just got an idea - what about putting some resistor after the 12v regulator's Vout to trim the output voltage to 12.0 ?
I notice that the much higher performance regulators are causing you problems :scratch:Quote:
Originally Posted by mcebrian
Have you possibly considered that it's not the lower noise spec of the better regulators but something to do with their much wider bandwidth that may be problematic?
Any circuitry connected to these regulators will only have a certain amount of PSRR (power supply rejection ratio) & this decreases with increasing frequency. As the super regs have a much wider bandwidth it could be possible for them to upset any circuits being powered by them at ultrasonic frequencies? Just thinking off the cuff here, it might need looking into.
What you need to remember is that the Caiman & Gator were designed to run & run well on industry standard regulators. You'd need to ask Stan if he ever substituted them for either of those you mention & if so did he test for stability?
I'd get a 100MHz oscilloscope on the rails & take a look & see what's going on if i was you ;)