Hi!
I am in the process of building my first PSU for a hybrid tuba amp. My requirements are +24Vdc, 1A output.
I have a a Hammond 166K35 transformer that I got relatively cheap, brand new. The primary will be current limited at startup by two parallel 150R resistors (120V/75R = 1.6A primary, 5.5A secondary) The resisitors will be bypassed by a relay on a timer. The design comes from Rod Elliotts project 39, found here I am also planning to use a linear voltage regulator.
Here is the schematic of my design.
The transformer has 35V secondary with a center tap. I am not planning to use the center tap, so that wire will be unused. The bridge rectifier is 400V 15A.
The primary current limiting resistors are 50W, and should be able to handle the inrush current.
R1, R2, L1 and C3, C4 make a RLC filter.
R1 and R2 are 50W 2R resistors, and should be able to handle the inrush current, and a constant 1A just fine.
All caps are rated at 50V.
Simulation gives good results - 29V before the linear regulator, and less than 30mV ripple.
My concerns are with the capacitors, and the inrush current they will see at turn on. I have got a BOMs already t go from digikey. C1 and C2 have ripple current of 3.13A, and C3, C4 have ripple current of 4.5A. With the primary side current limiting, C1 and C2 will see almost 4A ripple, and C3 and C4 will see the max of 5.5A for well over 100ms. (if my calculations are right, which they probably aren't - Im new to transformers)
Besides my concerns about the capacitors, can anyone find any mistakes with my design?
Thanks, Hengy
I am in the process of building my first PSU for a hybrid tuba amp. My requirements are +24Vdc, 1A output.
I have a a Hammond 166K35 transformer that I got relatively cheap, brand new. The primary will be current limited at startup by two parallel 150R resistors (120V/75R = 1.6A primary, 5.5A secondary) The resisitors will be bypassed by a relay on a timer. The design comes from Rod Elliotts project 39, found here I am also planning to use a linear voltage regulator.
Here is the schematic of my design.
The transformer has 35V secondary with a center tap. I am not planning to use the center tap, so that wire will be unused. The bridge rectifier is 400V 15A.
The primary current limiting resistors are 50W, and should be able to handle the inrush current.
R1, R2, L1 and C3, C4 make a RLC filter.
R1 and R2 are 50W 2R resistors, and should be able to handle the inrush current, and a constant 1A just fine.
All caps are rated at 50V.
Simulation gives good results - 29V before the linear regulator, and less than 30mV ripple.
My concerns are with the capacitors, and the inrush current they will see at turn on. I have got a BOMs already t go from digikey. C1 and C2 have ripple current of 3.13A, and C3, C4 have ripple current of 4.5A. With the primary side current limiting, C1 and C2 will see almost 4A ripple, and C3 and C4 will see the max of 5.5A for well over 100ms. (if my calculations are right, which they probably aren't - Im new to transformers)
Besides my concerns about the capacitors, can anyone find any mistakes with my design?
Thanks, Hengy
All capacitor input PSUs have a large charging current at first start.
Use a current limiter to reduce this current pulse/s
CRC and CLC PSU have a continuous heavy ripple current on the first stage smoothing capacitor, not just during starting.
Use a very expensive, very large volume, very low ESR capacitor to survive this continuous ripple duty.
or
use a set of paralleled normal grade capacitors that have a total ripple current rating that comfortably exceeds the continuous ripple current.
Your 330uF+330uF is the second solution.
Use a current limiter to reduce this current pulse/s
CRC and CLC PSU have a continuous heavy ripple current on the first stage smoothing capacitor, not just during starting.
Use a very expensive, very large volume, very low ESR capacitor to survive this continuous ripple duty.
or
use a set of paralleled normal grade capacitors that have a total ripple current rating that comfortably exceeds the continuous ripple current.
Your 330uF+330uF is the second solution.
Thsi is only a 1A power supply, which is trivial. The capacitors you quoted will have no problem handling a few amps of inrush current, and the steady-state ripple current will not be more than 2Arms (and it will be shared between your caps anyway). I wouldn't even bother with C1 and C2 -just put C3 on the 'input' side of R1/R2.
Off the top of my head, two refinements you could make are to add an RC snubber across the transformer secondary (10nF+1k is a good start), and to put another choke in the ground side of the supply. This will form a common-mode filter, which is a powerful way to keep mains-borne hash out of the supply (unless the rest of your circuit doesn't allow anything in this negative path, I don't know).
Last edited:
Thanks for the responses! I will consider your suggestion of a RC snubber and relocating L1. I guess I have a bit more research to do!
I did notice that adding the two 330uf caps made more of a difference than adding another 10000uf after the inductor. (In my simulations). The whole reason I tried adding caps there was because LTspice showed bad oscillation.
Thanks,
Hengy
I did notice that adding the two 330uf caps made more of a difference than adding another 10000uf after the inductor. (In my simulations). The whole reason I tried adding caps there was because LTspice showed bad oscillation.
Thanks,
Hengy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.