OLD THREAD DAC End by Andrea Ciuffoli

[quanghao]

Ok, i just dropped in another 27k into r13 and now its at 5.22v max.
Isn't that still low?

if you change R33 = 22R, the V-out is still that, you can change trimmer 200R bay 500R or 1K!

Quanghao

 

[adamus]

its not the trimmer, its is not at the end... something else is happening with the shunt reg.

at a certain point vgs of the second mosfet goes to zero

 

[Agi]

From the source to the gate of the second fet, i get the same voltages as the outputs. Surely this is incorrect?
The design by salas says it should be 4.07v across them...
http://www.audiodesignguide.com/DACend2/Sala-shunt-for-DAC.jpg

Which in turn leads me to think that something is wrong with that drivers part...
What v drop should we be getting across the led's?

Apellizo if you have your voltages about right ~5.5v. Could you possibly check the voltages you get around that part of the board.
Thanks.

 

[Salas]

If you use a 470R as test dummy load, then 10mA will be running to the dummy, rest to the shunt mos element. If it works like that with measurable Vgs circa 3-4V, there is nothing malfunctioning in the reg itself. The thing here is that a shunt reg feeds many little shunt regs at about the same set voltage. See what is the consumption by using a dvm in series current mode when feeding the DAC board directly from the main filter cap with crock clips also. Is the running current you set enough so the shunt mosfet has at least +20% to work with so not to cut off? Maybe you have consumption differences between builds and some are asking beyond your typical CCS settings? I am afraid that the TLs rob the main CCS current and set the voltage themselves, starving the main mosfet. Vdrop across the Leds should be their number X individual Vf.

 

[adamus]

Salas, i think you are right. my build showed around an 85ma pull from the dac board without chips in, so we should be lookign at 115ma+ which is what.... 2.2/r = 0.120, R= 18ohms minimum max?

I have a couple of 33r in parallel so 16r, and its still close...

but as i said, if you adjust to say 5.15 so vgs comes up, the tl431 still do the job (albeit adding there own noise), they dont want much overhead. I'll try going downt 10r or 8r to put a bit more current through the shunt mos.

 

[Agi]

Thanks for your input Salas. I just pulled out my wire from the shunt to the dac board, i popped my multimeter in place of it and got 64ma across the digi 5v without the chips in.
Hmmm... So if thats the case, or like adamus, can you think of any explanation of why the vgs of the second fet is 0 (when only a single 27r is used, considering i always get 3.5+v across the second fet on the +5v analog, considering they're identical?
Is it just as the tl431's are using such a high current draw?
Is it ok to run the fets as low as 5.15v? if they have values above 3.5v for vgs.

Also sorry what is CCS?

Sorry, noob learning. 😱

 

[adamus]

constant current source or sink,

 

[Salas]

I know nothing of the system concept but from what you measured, I would set the current source at 100mA first, so 36mA remains in the main reg. But verify the draw with the chips loaded.

 

[Salas]

Salas, i think you are right. my build showed around an 85ma pull from the dac board without chips in, so we should be lookign at 115ma+ which is what.... 2.2/r = 0.120, R= 18ohms minimum max?

I have a couple of 33r in parallel so 16r, and its still close...

but as i said, if you adjust to say 5.15 so vgs comes up, the tl431 still do the job (albeit adding there own noise), they dont want much overhead. I'll try going downt 10r or 8r to put a bit more current through the shunt mos.

2.2/r is ballpark, measure your individual vdrop across r and divide by r. Use same vdrop for projecting new r. Watch to be conservative with heat.

 

[Agi]

Hmmmm... after running my PS up for my I/V end I got 250v after about 2 mins.
Is that normal?
It's still increasing but even slower than upto 250v.
I have a 300v input and the fet gets warm, quite warm after only a few mins, however it is well mounted to a sink so the sink gets warm, yet the fet stay relatively cool.

Sound about right?

 

[adamus]

sounds about right to me.

i have 260vac so roughly ~ 366vdc input, and get around 280 out regulated. The fet gets hot, but not too hot.

whats the tranformer spec'd at?

 

[Agi]

Outputs 2x
Power 100VA
Output Voltage Current
300V .15A
6.3V 3A

I'll leave it on a few mins then and see just how high it does go to. I assume it will stabilise about 270v.
Thanks.

 

[Agi]

I've let it run for about 5 mins, no tubes in, no connections to the 9-12v side either. So just the I/V board and the 300v input.

It climbed all the way upto 292v within 2 mins then sunk down to 289v then stuck solid.

I guess 289v is too high?

 

[adamus]

its fine as long as its regulating, look at the curves on the datasheet

 

[Agi]

I meant for the valves, not for the reg.

 

[adamus]

yep fine, look at the datasheet,

 

[Agi]

Cool. Thanks for your help.
I'm happy with everything now, just waiting for the trim for the heater supply and the new cs8414 then i'm ready to go.

 

[Agi]

My resistors are smelling like paint on my I/V board. They read 11.64K (4 x 47k paralleled)
So it's not too far off ~11.75.

Is it bad that they smell, they don't smell a huge amount, just a lil wiff of paint after a few mins of operation. No 6.3v connected, nor valves in.
As soon as i smelt it off it went. I've never worked on anything from scratch before, so do resistors having some considerable power through them smell to start off with? It's not like the burnt resistor smell, which i've smelt before...

I guess my voltage out of my supply is too high, right?

 

[adamus]

what rating are they?

15ma, 190v drop = 2.85watts

4 resistors, so under a watt each. 2 watt rating is fine.

 

[Agi]

Yes they are 2 watts each and they are dropping about 200v each, so still within limits.