5Z4 inrush current too high

[Craig405]

Hello,

Slowly progressing through my Aikido build at moment, hit a minor snag.
Im using a 5Z4 rectifier in the power supply followed by:

____________________-L-C---------channel 1
rectifier-----C-L-C-L-C
____________________-L-C---------channel 2

with values:

_______________________-7H-1000uF
20uF-20H-20uF-20H-1000uF
_______________________-7H-1000uF

Each channel has a branch off of 1000uf and 7H... is what i have tried to illustrate! how successful i have been im not sure..

The peak current at turn on exceeds my rectifiers specification (0.88A peak).

Without removing my beloved 1000uf caps whats the best solution?

I already have some chunky octal relays i was thinking of using to either switch in a current limiting resistor after the rectifier, or switching in the first 1000uF cap in a few seconds after the PSU is powered up.

Is there a better solution, or any issues i may have with my solutions?

[Colt45]

Do you really need that many bloody chokes? That's bound to get expensive.

or that much capacitance either...

[Craig405]

Haha, not really but i have them already and my case fits them nicely so why not? Ill just have about 1mv of ripple

[Robert McLean]

The inrush current will not be affected significantly by anything after the first capacitor. Changing the 1000 uF to another value, or switching it in or out will make no difference. The only way to reduce it is to add series resistance, or a choke before the first capacitor. ( or somehow change the voltage differential between the rectifier output and the capacitor , but I have no idea how to do that.)

[pedroskova]

Is this measured... or simmed thru PSUD2? Is it at turn on, or a constant draw? Dropping the size of the input cap down would probably gain you some leeway. If it's measured current that you are referring to, and it isn't just a turn on spike, adding series resistance would help (maybe not with the sound).

[Craig405]

Thanks for replies, i reduced the first cap after the rectifier to 1uf in the simulation (yes it is all simulation of initial transient) and it works within the rectifiers safe values now and still gives a high enough HT!

Guess ill have to put my octal relays away for another day (shame i like their heavy clunk noise).

Cheers

[PeteN]

Would it not be possible to use a suitably chunky inrush limiter thermistor in series with the power transformer primary (or would secondary, before the rectifier, be better?) to stop this?

If you have separate PTs, another solution could perhaps be - which any Morgan Jones fan will recognise - relays which latch closed once the AC voltage passes 0 (so that the initial current is 0, and increases 'gently'). *mutters to self* I believe they require a LT supply to run their internal circuitry so would need either a smoothed tap from the filament supply or a PP3 cell and a 'turn on' switch I guess. I could be talking rubbish here, pure speculation : )
(Incidentally, I have looked around for these but can't seem to find them anywhere. Actually, I found some on eBay, but they looked like the sort of thing you'd use on a 600kW motor than a valve amp. I think it could just be a case of not knowing what they're called.)

[Sheldon]

Originally posted by Craig405
Thanks for replies, i reduced the first cap after the rectifier to 1uf in the simulation (yes it is all simulation of initial transient) and it works within the rectifiers safe values now and still gives a high enough HT!

Guess ill have to put my octal relays away for another day (shame i like their heavy clunk noise).

Cheers

Go to the options menu in PSUD and select soft start. That will simulate actual start up conditions for your rectifier. You will find that you are not exeeding the safe values. The only exception would be a hot start with the initial cap drained - unlikely with your supply design.

Sheldon

[Craig405]

PeteN, i did think about this inrush limiter idea. Although i dont think the current is high enough under normal use of the amp to heat them up so i may as well stick a low value resistor in there.
Im just speculating though i never tested them.. but i ruled them out on those grounds.

Interesting idea, never new these type of relays existed. If i had one id surely give it a try! Although on PSUD the peak of the transient is at about 1/2 seconds after turn on so maybe it wouldnt make any difference starting at zero volts with 50hz.

I checked my options menu in PSUD and soft start was ticked for my simulations.
When i get this amp PSU running i think ill set upon it with a fast digital storage scope and a current probe to see exactly whats going on and compare with PSUD.. that should be interesting.

[Dave Cigna]

A question: did you include the resistance of the transformer in your sim?

A comment: the inrush at turn on will be controlled by the heating of the 5Z4. If it's cold when you apply power, then it will not conduct instantly; it will 'come up' gradually as the filament heats up.

Personally, I wouldn't worry about it in actual use. It should be able to handle 20uF input cap without trouble.