| ZenV4 PSU, Nelson & Kristijan combination? - Click HERE for Original Thread |
| Tazzy |
I've been collecting parts for 2 Zen V4 mono's.
Since the parts for the PSU are complete I'm looking into the layout and setup of the PSU. I've mixed Nelson's gebruik. and Kristijan's design and came up with:
The transformator is a 2x 35V 300VA amplimo ring that's used in a 1x 35V 600VA setup.
Any tips or hints? |
|
|
| wuffwaff |
Hello Tazzy,
why do you use such small caps for C2 (4x1200microF). I think itŽs better to use the same values for C1 and C2.
C2 must be able to supply the current peaks that will occur with the Aleph current source.
william |
|
|
| Tazzy |
I thought that big caps where slow on discharge, smaller ones are fast?
Lose the 1200's and go for 6800 - 2,2 mH - 6800 - 6800 - 1? |
|
|
| Nelson Pass |
This seems to be a common perception. Assuming that
this is the case, then paralleling large "slow" capacitors with
small "fast discharge" capacitors ought to solve the
problem. |
|
|
| wuffwaff |
Tazzy,
the first cap will determine the voltage youŽll get after the choke. Small --> smaller voltage, big ---> higher voltage (max. determined by transformer).
IŽm not shure if 6800-2,2-6800-6800-1 will do the job. This depends a bit on the bias setting you go for but seems rather small to me (I tend to overdo this cappy thing).
I would just try this setup and ad more capacity if the voltage is too low (C1) or if you get a significant ripple at C2 under full power.
william
p.s. try installing PSUDesigner, works great to calculate the voltages you will get in advance. |
|
|
| Tazzy |
I ran the first design with PSU2, doesn't look bad at all??? |
|
|
| HDTVman |
Did you check the current in the transformer or the diodes in the first few ms, no reporting delay? Now set the current to a value that the amp will be biased to using the constant current feature. A delay is OK here. Next set C1 resistance to .1 ohms not 2 ohms and C2 to .33 ohms not 2 ohms.
Now that the real world and the simulation are a little closer I think it will look a little different.
Later
Bruce:geezer: |
|
|
| wuffwaff |
Tazzy,
Bruce is right. You also could look after a delay of 1000ms and then just for 100ms. YouŽll see the ripple for c1 and c2 a lot better.
Try the step function for the current going from bias to 2x bias current and look at the effect on Vc2 (only for one period)
william |
|
|
|
|
