Tuba SMPS filter: two linear regulators inside VFET/Theseus chassis; incl thump kill

Member
Joined 2011
Paid Member
I was looking for parts and found that the LD1084 is a max 30V input voltage device.
I understand this regulator is meant to work at 40VDC. What am I missing?

Probably you haven't yet read explanations of how to get >150V from the LM317 family of adjustable regulator ICs. (LD1084 is pin and function compatible with LM317, plus higher current and lower dropout). The standard introduction to the topic is National Semiconductor's Linear Brief 47, attached below. Read its first paragraph very carefully. Go ahead and read the entire 4 page article (of course), but the answer to the question above appears in the first paragraph.

Then ponder the impact of component "ZD1" in the circuit schematic of Tuba (post #1 of this thread).
 

Attachments

  • Maida_LB-47_Linear_Brief_High_Voltage_Supply.pdf
    108.6 KB · Views: 221
Probably you haven't yet read explanations of how to get >150V from the LM317 family of adjustable regulator ICs. (LD1084 is pin and function compatible with LM317, plus higher current and lower dropout). The standard introduction to the topic is National Semiconductor's Linear Brief 47, attached below. Read its first paragraph very carefully. Go ahead and read the entire 4 page article (of course), but the answer to the question above appears in the first paragraph.

Then ponder the impact of component "ZD1" in the circuit schematic of Tuba (post #1 of this thread).
thanks Mark, that helps. I didn't read that.:)
 
I got Qty 4. 511-LD1084V at Mouser. I ordered Dec 2nd.
Assuming my board fires up OK without releasing magic smoke from the first 2, I'll offer up the 2 spares here.

Stock looks pretty sparse.
https://octopart.com/ld1084v-stmicroelectronics-526908
Maybe the elves are getting them ready to be stocking stuffers? The LD1084 is the hottest gift this year for the cool kids.
Yeah, those are long gone from Mouser. As soon as Mark posts a new design the components fly off the shelf. :) Keep me in mind if you have extras.
 
My mistake, I meant the LM1085. This and the LM1084 are functional and pin compatible replacements for the 317, with lower dropout voltage.
I generally like to go with the higher figure part number to get my money worth ;) but the LM1085 seems to be a 3A and SMD part + they're all out.

What do you guys use for headers? There is a myriad of headers except the 2 pin header with jumper. I
 
Member
Joined 2011
Paid Member
Maybe some of the 2 pin headers made by Samtec, might fit and might work. Jumpers sold separately.

samtec.png

_
 

Attachments

  • samtec.png
    samtec.png
    4.8 KB · Views: 70
Member
Joined 2018
Paid Member
I generally like to go with the higher figure part number to get my money worth ;) but the LM1085 seems to be a 3A and SMD part + they're all out.

What do you guys use for headers? There is a myriad of headers except the 2 pin header with jumper. I

Here a few few Digkey p/n's

2.54mm jumpers: 1849-09200-71-BBGB00-ND
Alternate idea... I think I got a few hundred (supply for many lifetimes) for a few bucks on Amazon or Ebay a few years ago when I was building SissySIT or Iron Pre or something.

2.54mm jumper header - 2057-PH1-15-UA-ND
Buy longer strip and cut down what you need.
Again, you could get a bulk generic pack on Ebay or Amazon and have enough for your next 50 projects using jumpers.
 
I have my first TUBA board stuffed and ready to setup/test. Hoping for some thoughts on 3 questions.

Question 1: is the target 2V across the 1084 reg a "hard" target, or can it be reduced to any value between 2V and the max 1.5V dropout voltage spec'ed in the datasheet?

I ask b/c I'd really like to test the TUBA with my upcoming Big MoFo build - which will require 24V @ 2.5A per channel. Reducing the voltage dropped across the reg to 1.7V would reduce the heat that needs to be dissipated by 0,75W.

Question 2: can the TUBA otherwise handle 5A output, 2.5A per channel? I followed Mark's advice and had the boards made with 2 oz copper - so I'm hoping the traces can handle the current and won't overheat the board. The choke is rated for 5.2A - so 5A is pushing things, and I believe inductance varies inversely with current - so maybe I'd be compromising the hash filter.

Question 3: Assuming I make no construction mistakes, is there any obstacle to adding a second identical hash filter, so that each channel has its own and can be powered by its own separate SMPS? I figure I could add a second identical hash filter, cut the trace between C2 and C3 and connect the second hash filter by exposing the trace near TP1. I'd lose TP1 as a test point for the other channel, but there's room to connect a grabber probe to the output leg of the original choke.

many thanks in advance, Derek