Caiman power supply mod

[robyonekenoby]

Hello, currently I'm running my caiman on a good quality linear PSU (ripple <0.3mV) at 13.2 V

I know the optimum voltage should be 15, but since I've done the passive mod, this is not an issue because the 7812 is not used, and I can say it is working beautifully

But I've a couple of question

1) is the 7805 feeded by the 7812 or from the main supply? I can read 9 volts on the 7805 input pin; may I eventually feed the 7805 with the main 13.2V?

2) Let's suppose I want to bypass the 7812 and eventually feed the analog board (the stock 4562 or a gator) directly with the main at 13.2; I'm going to burn something or is it ok? I've already done this mod on my previous 7510

thank you for your help

[Reid Malenfant]

Hello, currently I'm running my caiman on a good quality linear PSU (ripple <0.3mV) at 13.2 V

1) is the 7805 feeded by the 7812 or from the main supply? I can read 9 volts on the 7805 input pin; may I eventually feed the 7805 with the main 13.2V?

2) Let's suppose I want to bypass the 7812 and eventually feed the analog board (the stock 4562 or a gator) directly with the main at 13.2; I'm going to burn something or is it ok? I've already done this mod on my previous 7510

1. I'm not sure if the 7805 is fed from the 7812. However there is a resistor connected in series with the 7805 input pin, i'd suggest you measure the voltage on the other side of the resistor to find out. If you are feeding the whole unit 13.2V then the 7812 isn't going to have enough voltage input to regulate properly & i think you'll find the output may well be lower than 12V. Yes you could feed the 7805 with 13.2V but you may need to increase the size of the heatsink as it'll be dissipating more power ;)

2. 1.2V will burn nothing & increasing the voltage to the analogue PCB may well have the effect of very very slightly reducing the distortion as the op amps will be running on a little more voltage & thus be a tad slightly higher biased. I'm sure the op amps are capable of operation on 30V supplies (most are) or +/-15V DC.

[robyonekenoby]

Thanks for your reply, I just checked the voltage before the resistor and I can read 13.2 V

I know opamp has no problem with such increasing in voltage, I'm just concerned about burning something else on the pcb actually fed by 7812

And I can confirm what you said about lower opamp distortion: when I swapped the 4558 for the AD8620 on my old 7510, I noticed a definite improvement in clarity, definition and smoothness when I've bypassed the 7810 and fed the opamp with 13.2V, -6.6 +6.6 instead of -5 +5

I'm wondering if it is possible to lower further the PSU ripple by adding an inductor in series on the psu exit, but I'm scared about oscillation
The inductor should not affect transient load because the caiman can always rely to the local 10000uF tank

[Reid Malenfant]

I don't think you'd have a problem introducing an inductor :) Of course the best thing to do would be modify the power supply itself to reduce the ripple ;)

Give the inductor a try, the output of the linear PSU must have a capacitor of some sort on the output ot that'd possibly oscillate. You'll effectively be making a Pi filter & these work pretty well on valve amplifiers :eyebrows:

[robyonekenoby]

The PSU is a DIY project built with a toroidal transformer, a rail with 5 panasonic FM 2200 uF of the lowest available ESR, then the LM something (don't remember the regulator code), and another rail of 5 identical panasonic FM 2200uf plus a small film cap

My plan is to put another film cap on the exit, a mundorf metallized poly 2.2uF plus an inductor in the attempt to furhter reduce ripple

[Reid Malenfant]

The limiting factor will be the Hitano 10,000uf in the Caiman itself ;)

I just bought one of these recently myself, not plugged it in & tried it yet as i'm a happy listener as it is :)

What i intend to do is put the whole lot in a new case & build a totally discrete single ended class A audio output stage running off of +/- supplies & obviously change regulators as they aren't very good tbh :eyebrows:

I'll have to change one of the S/PDIF inputs to AES/EBU balanced digital input & both left & right audio outputs will need to be balanced to :scratch:

It'll be interesting to see what it can do, guess i'd better give it a listen before i disect the poor thing :lol:

[Surreydocker]

a totally discrete single ended class A audio output stage running off of +/- supplies

& both left & right audio outputs will need to be balanced to



That's exactly what I need - I'd be very interested to see what you do...........

Ian

[robyonekenoby]

UPDATE

inserted ferrite ring inductor on the psu exit, the ripple now is below the threshold of my oscilloscope, it shows a flat line without any ripple ;)

[Reid Malenfant]

Good news, i thought it'd at least help rather than hinder things :)