Salas SSLV1.3 UltraBiB shunt regulator

Jump to Latest
#2,983
I did a simple measurement of the Id while changing the Rs value on 2sk117 to keep constant Id = 1.5mA, Vds = 10V

The plot is made of 8 points:

2sk117-f(Idss)=Rs.png

The Id is pretty constant in the saturation region while changing Vds, which means - if I'm not mistaken - that the lambda is small, so it's not a "tubey" JFET but it's a good one for CCS 🙂 It also drifts with temp as all JFEts.

It looks like a straight line. So having this:

point 1: Idss=8.2mA, Rs=273 Ohm
point 2: Idss=3.65mA Rs=98 Ohm

we can find the equation (used Least Squares from 8 points):

Rs = f(Idss) = 38.35*Idss - 40.05 (Ohm) - Idss value in mA

Idss (mA) Rs
3.65 98
3.76 102.6
4.03 114
4.36 130
5.03 153.4
5.09 155
5.63 178
8.2 273
 
Last edited:
#2,984
Its not the steadiest but it has small Vp and low noise also so I use it in SSHV2 for ages. Very steady ones of high Gos are usually noisier with higher Vp. I look at Vp also because I nest it under a transistor's Vbe in the particular circuit.

Finding a degeneration resistor for each K117 giving 1.5mA in circuit when for Ubib well done.
 
#2,985
Thank you Salas.

SSHV2... soon I am going to build a prototype of a preamp with 100year old DHT tubes and need around 200V DC PSU... are these PCBs for sale? 🙂

I have already parts and PCBs for the capacitance multiplier but I could try both 🙂

EDIT: forgotten to mention, it can draw up to 25mA per channel, 3mA the tube, about 20mA the follower
 
Last edited:
#2,987
Hi everyone,

I have a quick question on my build. I threw together a positive psu about a year ago but stuck it in a drawer , never used or tested. Now I have a project that I want to use it on. It is set up to deliver 24 vdc and the load will be ~ 150 mA, I set R1 to 2R to give a CCS of 300 mA. I powered it up quickly today with no load or heat sinking just to see if it worked. The LED lit and I was able to adjust the outlet to 24 vdc but M2 got hot very very quickly, needed to kill the power after 10 sec or so to keep from burning M2. In the guide it alludes to M2 getting hot with no load. I plan on installing a beefy heat sink before I do anymore testing, but I'm wondering if M2 getting so hot so quickly is normal.

thanks
 
#2,988
It has to burn all 300mA at 24V i.e. 7.2W if not shared with a load. M2 shunts undelivered power to ground. That's why its named shunt reg. It controls in parallel so it shares or does it all by itself when unloaded. Test with a 150 Ohm power resistor (5-10W) as dummy load. That will pull 160mA which is close. So you will be able to predetermine heat sink size for M2 when with the real load you have in mind.
 
#2,989
Salas said:
It has to burn all 300mA at 24V i.e. 7.2W if not shared with a load. M2 shunts undelivered power to ground. That's why its named shunt reg. It controls in parallel so it shares or does it all by itself when unloaded. Test with a 150 Ohm power resistor (5-10W) as dummy load. That will pull 160mA which is close. So you will be able to predetermine heat sink size for M2 when with the real load you have in mind.
Click to expand...
Will do, if still too hot for my taste after heat sinking and dummy load I'll change R1 to reduce current to M2
 
#2,993
hello salas,
with 21vac trafo i have 27vdc input, and when turn on uBib always shows 14vdc
turning trimpot will not rise the voltage above 14v, it's only can drop voltage below 14v
and the M1 and M2 not hot at all
but when i turn trimpot until i have below 10v the M1 M2 start working (its hot)
it's look like the M1 M2 gate won't open if voltage above 10v

why is this happen?

thanks.
Peter
 
Last edited:
#2,995
wengtech said:
hello salas,
with 21vac trafo i have 27vdc input, and when turn on uBib always shows 14vdc
turning trimpot will not rise the voltage above 14v, it's only can drop voltage below 14v
and the M1 and M2 not hot at all
but when i turn trimpot until i have below 10v the M1 M2 start working (its hot)
it's look like the M1 M2 gate won't open if voltage above 10v

why is this happen?

thanks.
Peter
Click to expand...
Hi,

Maybe Q2 is bad. Possibly Q3 too. Measure Vbe for both. If not between 0.5V-0.6V replace. If good, we will think of another checking. What value trimmer you use?
 
#2,996
TioFrancotirador said:
Salas,
Is there any difference in performance between using: dual positive ultrabibs connected in way to output -/0/+, versus using positve and negative ultrabibs version to output the same?
Click to expand...
I don't believe there is a significant performance difference. Dual positive builds are most alike between them due to same gender semiconductors. Grounding schemes befitting each approach i.e. virtual ground vs actual ground maybe play more of role regarding the main audio circuit itself.
 
#2,997
Salas said:
Hi,

Maybe Q2 is bad. Possibly Q3 too. Measure Vbe for both. If not between 0.5V-0.6V replace. If good, we will think of another checking. What value trimmer you use?
Click to expand...
hello salas,
R5 R6 shows 0.6v (240R)
R4,7,8 shows 0.5v (220R)

I'm using 20k trimpot

problem is, what ever resistor values between 20k to 8k, voltage always shows 14v, input voltage is 27v steady, and I need at least 20v

below 8k voltage start to drop according resistor value, from 14v i can get 8v by trimming into 2k2.

thanks.
 
#2,998
Those Vbe and Jfet bias figures on their associated resistors are pretty normal. Could be you have little CCS with chosen R1 for a particular test load and the system current limits when you push beyond some output voltage threshold. How much Vgs on M1 and M2 when pulling? Q1 Q4 have normal Vbe?
 
#2,999
Salas said:
Those Vbe and Jfet bias figures on their associated resistors are pretty normal. Could be you have little CCS with chosen R1 for a particular test load and the system current limits when you push beyond some output voltage threshold. How much Vgs on M1 and M2 when pulling? Q1 Q4 have normal Vbe?
Click to expand...
Hi salas,
i changed the trafo tap into 24vac (before 21vac)
same result, always shows 14-15vdc at output with trimmer 8k to 20k or above.
then voltage start to going down after trim below 8k.

i've measured
pin 1,2,3

Q1 = 24.8 28.5 29
Q4 = 0.0 3.6 4.24

Q3 = 3.6 14.2 3.7
Q2 = 3.7 14.2 15.3

M1 = 25.2 15.3 28.8
M2 = 4.22 15.3 0.00

LEDS = 3.5V

R1 = 0.580 (3R3) = 175mA

thanks.
 
You must log in or register to reply here.
Top Bottom