Can be strong 1A because its CCS (does not let time to rest) and full bridge rectifier cuts Tx AC current rating by 40%. Feels like 1.4A AC to the winding.
Does not look like loud mechanical noise. And I don't see 50Hz hum frequency peak.
That's a good idea can be snuberized using Quasimodo but I believe do nothing against the high buzz.
Less current will help the Tx go more quiet I meant. By the way, if R-Core has fallen and hit or had strong transportation shake it can become loose. Especially if already mounted on a chassis. I had a 120VA R-Core it once came loose because of bad courier handling. I saw it had base frame metal teeth to keep it tight. Those went out of shape from the shaking and I bent them back to grip the core.
If the bobbin on R-Core loosed, the excessive load possible shake the bobbin.
Try to wedge out the bobbin.
Have you tried with thermal padding in between? It might bring the temps down a bit further.
I tested UBiB 2 days ago quickly without heatsink.
100 Ohm resistor at each outputs to relief stress from output transistors.
UBiB worked and output voltage set.
Later I found cute small heatsinks that I used for Neg Pos M1 and different one for M2.
Pos M2 resistor became anomaly hot. I even burn a finger a little.
IR thermometer showed temp rised fast to 70 80 90 100! Celsius.
I turns out heatsink was close to pcb, it touched output pad of R1 and metallic part of M1 transistor. I didn’t use thermal pad.
Well.. nothing have died and UBiB still works.
I reset output voltage after 2 hours.
Pos is pretty stable at 18V.
Neg could drift from air to +/- 0.04 mV and then be back to 18V.
P.S.: I forgot to tell that my 5W 100 Ohm resistor that I used for testing was like about 115 Celsius hot. 10W with long leads might be more appropriate.
P.S.S.: Heatsinks for M1 was flipped upside down. So problem with heat sink touches near pads was solved.
100 Ohm resistor at each outputs to relief stress from output transistors.
UBiB worked and output voltage set.
Later I found cute small heatsinks that I used for Neg Pos M1 and different one for M2.
Pos M2 resistor became anomaly hot. I even burn a finger a little.
IR thermometer showed temp rised fast to 70 80 90 100! Celsius.
I turns out heatsink was close to pcb, it touched output pad of R1 and metallic part of M1 transistor. I didn’t use thermal pad.
Well.. nothing have died and UBiB still works.
I reset output voltage after 2 hours.
Pos is pretty stable at 18V.
Neg could drift from air to +/- 0.04 mV and then be back to 18V.
P.S.: I forgot to tell that my 5W 100 Ohm resistor that I used for testing was like about 115 Celsius hot. 10W with long leads might be more appropriate.
P.S.S.: Heatsinks for M1 was flipped upside down. So problem with heat sink touches near pads was solved.
Last edited:
I simulate this mistake in ltspice just now and it seems M2 dissipated 42W. 😱
So glad output tranistors is durable and I quickly turned power down.
2.32 Ohm.
I seems if R1 would be 0.1 Ohm then M2 would die instantly. It would dissipate 500W.. 🙂
So yes, current limiting saved the day.
I seems if R1 would be 0.1 Ohm then M2 would die instantly. It would dissipate 500W.. 🙂
So yes, current limiting saved the day.
Q1Vbe=VdropR1. Our CCS formula divides that by R1 i.e. 0.6V/0.1=6A
Assuming 18V output, times 6A would be 108W shunted through M2
Assuming 18V output, times 6A would be 108W shunted through M2