F5 Turbo Builders Thread

[AndrewT]

Have you tested whether the Power Thermistor "alone" can survive a mains short to PE, blow the mains fuse and allow time for the arc to extinguish?

If not, then you are guessing !

And after all that can you test whether the Power Thermistor can be reused, or must be binned after the test?

If you search my name you will find my report on the testing of the "Disconnecting Network". Not only did it survive, the tiny 100nF cap and the 600mW resistor were both undamaged.

 

[aft]

-if possible , input Jfets need to be of same Idss ; if going to multi output pairs route , all N need to be matched , and all P need to be matched
-if you buy Metal Oxide ones , they're already in enough low tolerance ; if using white ceramic ones (wirewounds) , buy more and measure them (with ccs , measure voltage drop)
-preferably , but not so important

see what's h_a is having to sell

Dear Zen Mod,
What cascoding input Jfets' how matching should look like between 2SK170(Q1)//2SK170(Q1.1)/2SJ74(Q2)//2SJ74(Q2.1)
Thanks. Raoul

 

[Zen Mod]

I'm not getting your question

what "Q" means ?

 

[BobEllis]

ZM - That's the part numbering scheme on the store F5T boards.

Aft, you want all four as close as possible. I''m not sure what a good Idss match is, but I suspect .2mA or closer should be decent. Minor variations between the SK and SJ pairs can be trimmed out with P3, so get the SKs to match each other and SJs to match each other more closely then work on SK to SJ.

 

[Mallard]

PSU CRC filter dimension

I have a question regarding the resistors in my psu crc filter.
How so I select the right value?

My psu looks like: 2 x 800VA (24V sec.) feeding in total ~350kuF using Tea-Bag boards.

Can someone explain how I calculate the correc value!?

Thanks,
Mallard

 

[Mallard]

ahh, forgot to mention my bias will be in the range of 2A each side.

I have a question regarding the resistors in my psu crc filter.
How so I select the right value?

My psu looks like: 2 x 800VA (24V sec.) feeding in total ~350kuF using Tea-Bag boards.

Can someone explain how I calculate the correc value!?

Thanks,
Mallard

 

[Zen Mod]

that's the compromise between ripple taming and ,say, speed of spoon-feeding last caps

say that's nothing wrong with 0R1

 

[Mallard]

ZM, thanks for your reply! Is there any rule of thumb how to calculate it?
The only thing I found was: the best way to find correct valus is to simulate it. But I don't have such simulator.

that's the compromise between ripple taming and ,say, speed of spoon-feeding last caps

say that's nothing wrong with 0R1

 

[Zen Mod]

xformer is 400VA , presuming that you are living in Vaterland , with 230Vac mains , that means that inclusion of 10R NTC in mains side of xformer , will allow use of (400VA/230V) 2A to 2A5 slow blow fuse

175mF per PSU leg is on limit ( for 400VA Donut ) , judging only by gut , without any calc

make it CRC (with 0R1 as R ) with 1/3 of C as first and 2/3 of C as final cell

 

[Mallard]

I need to correct you, I'm using 800VA donuts, these are more sweeter

Furthermore I'm using Tea-Bags PSU pcb so it is CCRCC.

xformer is 400VA , presuming that you are living in Vaterland , with 230Vac mains , that means that inclusion of 10R NTC in mains side of xformer , will allow use of (400VA/230V) 2A to 2A5 slow blow fuse

175mF per PSU leg is on limit ( for 400VA Donut ) , judging only by gut , without any calc

make it CRC (with 0R1 as R ) with 1/3 of C as first and 2/3 of C as final cell

 

[Zen Mod]

sorry , my mistake

then , with 800VA Donut , that amount of capacity is fine

that change fuse value - needing something in range of 800/230 , so 4A seems fine

 

[Mallard]

And actually I planned to use CL30 NTC which has 2.5R and 8A. (like Mr.Pass recommended it)

But wat does it mean for the R ind the CCRCC configuration? Still .1R ok?

sorry , my mistake

then , with 800VA Donut , that amount of capacity is fine

that change fuse value - needing something in range of 800/230 , so 4A seems fine

 

[Zen Mod]

NP is using CL60 , in 115Vac mains area

my experience is that 10R NTC is better for this side of Big Splash , juicing 230Vac from wall

regarding CRC - no matter what size of cap bank is in question , I can't stand more than 0R1 in line

if Donut/Graetz size permits , I'm not using it at all

 

[AndrewT]

Pass uses one CL60 for each 115Vac Primary.
In a 220/240Vac supply you still use one CL60 in each 110/120Vac Primary.

i.e. use two CL60 in series to feed the two series connected primaries in a 220/240Vac supply.

 

[Mallard]

In the NPs F5T article is stated:

CLC? Go for it. The thermistor in series with the primary is familiar, and I
recommend the CL30 type. Of course this shows the 120V AC version. For
240V the primary coils are wired in series and the fuse value is halved.

So is there anything what speakes against CL30 from your perspective?

 

[AndrewT]

The cold resistance is too low.
I would guess that 4 in series would be required for effective current limiting during the start up. But then I don't have the money to squander on 4 CL30 for each big transformer.

 

[Mallard]

4 CL30 or one CL60?

The cold resistance is too low.
I would guess that 4 in series would be required for effective current limiting during the start up. But then I don't have the money to squander on 4 CL30 for each big transformer.

 

[AndrewT]

Pass uses one CL60 for each 115Vac Primary

4 CL30 or 2 CL60

 

[james watt]

ZM, thanks for your reply! Is there any rule of thumb how to calculate it?
The only thing I found was: the best way to find correct valus is to simulate it. But I don't have such simulator.

This PSU Simulator is for only one positive rail. But one can play with different configurations and learn :
http://www.duncanamps.com/psud2/

 

[dhenryp]

Where to mount PS thermal switch?

I assume the 75 decg C Thermostat Switch on the power into the xformer should be mounted on one of the heatsink. Is that correct?