Building a symmetrical PSU B1 buffer

Jump to Latest
#601
Salas said:
Its a good rig the one you used so to check Vf at enough mA, as they will work on the real boards.

Just use 4 of them 1.95V and in a Led position put 2 little diodes (1N4004) in series instead then. 1.95X4=7.8+DiodeDropX2+Vbe=about 10V. Those diodes are checked once by a member and measured very low noise.
Just tie the 2 diodes on air and insert them as a single component in one of the led positions. There are solutions so not to wait and wait.
See the picture, and don't forget, in the negative reg cathodes are facing the other way around.
Click to expand...


There's a string of 3 and 5 leds - where does this apply?

Thanks,
JG
 
#606
I added in the new transformer last night, 15V secondaries. So it powers up fine and I get basically the same voltages out of the shunt. So all is well. The other thing I did was do a bit of fiddling with the 68R//68R. I followed AndrewTs advice and started by adding a 3rd 68R in parallel to drop the resistance in stages. Anyway, right now I have 68R//68R//10R in parallel (just didn't want to take out the 2 68R yet until I was sure there was no heat issues). So anyways, I haven;t really got to listen yet, but the 140/9140 still barely get warm, I'd guess 30degC, no more (I'm used to the F5 !!). So next step now is to swap them out for the 10R on its own.

BTW, I also did the mod using the coax instead of the regular silver in teflon and it made absolutely no difference at all. I was really surprised that there wasn't some change! I know the runs are short (about 50mm-80mm), but I expected something!!

So the plan tonight is to listen, swap the shunt resistors, listen, and then maybe put in the lightspeed if time allows.


Fran
 
#607
Measure the resistance of the combination you got right now for Rset and the voltage across so we know you really got high current. 30c sounds fishy. No difference with coax, means you had no real noise picked up by the open weave. Checked.

P.S. The regs outputs are supposed not to change, its just that sufficient input margin lets them perform as intended.
 
#610
Hi,

I got 100x 2sk170BL in from China the other day. I wont be needing them all. (Unless I can think of something else to build.)

I initially placed an order with Mouser - but they were on backorder for months.

So, if anyone wants 50 pieces email me. I'll set the price to 1/2 of what I paid for them.

$20 USD (free shipping too.)

Thanks,
JG
 
#613
68//68//10 should only give a total resistance of 7R. Something is wrong here, there should be more heat!!!

I'll do some checks later and report back. If I check the resistance as it is set up right now and the Vdrop across the paralleled resistors, that should tell me the current being dropped.

Fran
 
#615
woodturner-fran said:
68//68//10 should only give a total resistance of 7R. Something is wrong here, there should be more heat!!!

I'll do some checks later and report back. If I check the resistance as it is set up right now and the Vdrop across the paralleled resistors, that should tell me the current being dropped.

Fran
Click to expand...

Around 300mA +/- 50mA due to particular Vgs and Leds standards must be achieved with 7R27.
 
#620
woodturner-fran said:
I have to go back to this later tonight, but the resistors in parallel do measure ~7R and I get 1.6V and 1.9V drop across them....


The heatsink is fairly big so maybe its not as bad as I thought!

Fran
Click to expand...

The 1.6V side gives 230mA when the 1.9V side gives 270mA. Different Vgs to leds proportions between the particular P&N Mosfets. You can trim the 1.6V side even less for current set resistor value so it can go around 270mA too. Maybe 6.8R and 5.6R final resistors for 1.9V and 1.6V side respectively. For 280 & 285mA possibly. Your sinks seem like taking it very well.
 
You must log in or register to reply here.
Top Bottom