Transformer:
Primary 0-230v
Secondary
0-900v
0-200v No noise on load
0-10v No noise on load
0-10v No noise on load
0-6.3v No noise on load
AC load without rectifier 6000 ohm resistor = 150mA noise on transformer 20db ( aceptable noise )
DC load with rectifier R-C1-L-C2 ( 50R - 180uf - 15H ( 90ohm) - 470uf ) 6000 ohm on C2 noise on transformer 40db ( too loud )
notes:
Three different transformer ( no potted ) EI124 5kg from two different factories with the same result
I have replace all diodes, cap,resistor and choke without changes
loaded DC with resistor or tubes the same noise ( 2 211 tubes )
Core is correctly lock ( shapes no noise ) noise from bobbin
AC 0-900v secondary line when are rectified has 0.150V DC ( maybe this is the problem )
Noise frequency go from 400hz to 1100hz
AC 230v normally has 0.040V DC
ideas or answer are welcome.
Primary 0-230v
Secondary
0-900v
0-200v No noise on load
0-10v No noise on load
0-10v No noise on load
0-6.3v No noise on load
AC load without rectifier 6000 ohm resistor = 150mA noise on transformer 20db ( aceptable noise )
DC load with rectifier R-C1-L-C2 ( 50R - 180uf - 15H ( 90ohm) - 470uf ) 6000 ohm on C2 noise on transformer 40db ( too loud )
notes:
Three different transformer ( no potted ) EI124 5kg from two different factories with the same result
I have replace all diodes, cap,resistor and choke without changes
loaded DC with resistor or tubes the same noise ( 2 211 tubes )
Core is correctly lock ( shapes no noise ) noise from bobbin
AC 0-900v secondary line when are rectified has 0.150V DC ( maybe this is the problem )
Noise frequency go from 400hz to 1100hz
AC 230v normally has 0.040V DC
ideas or answer are welcome.
No free lunch.
Large first capacitor==large charging pulses.
On the secondary about 300mA RMS current pulses (over 2s).
Large first capacitor==large charging pulses.
On the secondary about 300mA RMS current pulses (over 2s).
but on PSU Designer I do not see a big difference between 47uf to 180uf diode normally work over this values only when change the R1 by inductor or bigger size resistor, reduce the current under 300mA to that is correct ?
Reevaluate your ground wiring and make it more literally like the schematic.
Run the rectifier (-) directly to the input C (-), and also run the output C (-) directly to the input C (-).
Ground reference should be the output C (-).
Run the rectifier (-) directly to the input C (-), and also run the output C (-) directly to the input C (-).
Ground reference should be the output C (-).
I have made a plate test literally like schematic, is too instructive, where all negative are wired as schematic.
when I replace the R1 50ohm to 350ohm the secondary current drop to 250mA original is over 350mA as Euro21 said , now is over 22db.
Is very interesting to study the transformer noise, thanks for your comment
Hi Tony, is mechanical noise in the transformer bobbin, ripple noise is under parameters 8mV at plate
0-900v has many turns with small wire diameter is not easy keep tense, this type of transformer are not easy to do,
with input choke or >250ohm resistor after diode bridge keep the transformer quiet.
How do you measure that mechanical noise?
"20dB", "40dB" makes no sense.
If possible post a recording, showing first the lightly loaded "noiseless" background, then the "noisy" one under load.
"20dB", "40dB" makes no sense.
doesn´t make much sense either, unless you explain where do you get that from, how you measure it, etc.
If possible post a recording, showing first the lightly loaded "noiseless" background, then the "noisy" one under load.
20db 40db makes perfect sence if there are using an sound meter, no? I have a meter that measures sound in dB for testing HVAC. Starts at 30.1dB and goes up to about 110dB or so.
Agreed though, the noise frequency should be mains or mains X2, no?
Agreed though, the noise frequency should be mains or mains X2, no?
well, is easy with telephone app, I put the phone on the transformer, these two pictures are:
6000 ohm load without rectifier directly on transformer, less noise
6000 ohm load with rectifier as schematic , too noise
do not need the phone you listen clearly the difference.
When your put the transformer inside the metal case cover is very audible at 3 or 4 meters in a silent listen room, my target is no audible transformer noise, the easy solution is potted but is danger to the increment of the temperature inside and need to increase the size and weight, I prefer to work in the schematic loosing a few volts ( 50v in this case ) at B+ but keep silent the transformer. after the change of R1 50ohm to 275 ohm , the transformer noise sound much better and relaxing.
6000 ohm load without rectifier directly on transformer, less noise
6000 ohm load with rectifier as schematic , too noise
do not need the phone you listen clearly the difference.
When your put the transformer inside the metal case cover is very audible at 3 or 4 meters in a silent listen room, my target is no audible transformer noise, the easy solution is potted but is danger to the increment of the temperature inside and need to increase the size and weight, I prefer to work in the schematic loosing a few volts ( 50v in this case ) at B+ but keep silent the transformer. after the change of R1 50ohm to 275 ohm , the transformer noise sound much better and relaxing.
Attachments
I have a Hammond that buzzes like that. I eventually abandoned EI and went toroid. No damned buzz anymore.
Thanks.
Sorry but you are stating it wrong.
Your unqualified
means, at least in the Acoustics World, an SPL value of 20dB or 40dB ... almost impossible to achieve.
For a real world reference about background noise present in various environments:
"Acceptable" value you quote is leaves rustling in the middle of nowhere, not even birds chirping, and "unbearable" 40 dB is quiet library/bedroom at night level or an isolated (and obviously empty) Studio.
In any case, for some reason your app does not give straight SPL values, but negative Db values relative to some unknown reference level.
I suggest you instead measure and post here "SPL level at 1 meter away from transformer", in both conditions, also a "transformer OFF" background noise value, for us to be able to grasp your problem as well as we can.
Dear JMFahey, I have measured many speakers with REW and understand its meaning, the data from the transformer is correct if you have a reference as background ( BG ) noise -100db more or less, but this info is only to see the different between them.
maybe the SPL at 1m in my work room is not possible to measure, too noise room. 🙂
maybe the SPL at 1m in my work room is not possible to measure, too noise room. 🙂
Sounds about right. I can hear a battery powered quartz clock ticking on the wall 10 feet away and it drives me nuts when I'm trying to test a phono preamp! Of course I only hear it if the fridge isn't running, the AC is off, everything else with a fan is off, night so no outside construction noise. Oh, and no one can use an elevator either. If a transformer is louder than the clock on the wall, it's a POS and unuseable in that job (PSU for audio) IMHO. Hence why I went toroid.
Never mind how loud it is (or why the numbers are negative). Who has the stock text about soaking a transformer in floor-varnish?
My target is how reduce the mechanical noise in normal transformer with >900v, normally this type of transformer are noisier than 300b transformer, I can affirm that with resistor o inductor after diode bridge reduce clearly the noise as 300b transformer, because the current drop clearly in the secondary of the transformer working much more relax, now speak about why background is negative -100db on a cheap App is not comprensive, is only a reference as Celsius, Farenheit or Kelvin degree.
follow with the test , another important thing is the turns volt ratio of the transformer I have checked three different transformer to know the turn volt ratio I wind 5 turn around the bobbin that give a voltage to know the ratio, the noisier transformer has 1,4 turn / volt since the silent and cooler has 1,8 turn / volt , only the problem is when more turn per volt ratio resistance of the wire increment to keep all in the same core, so is not too clear where stop, go to 2 turn per volt or decide that 1.8 is ok, to keep 900v and not increment the core need to reduce 0,32mm wire to 0,25mm but the resistance up from 70 to 110 ohm this value is not a problem looses are too small, who has played with this parameter ?
two factories who made the transformer in my country not have too clear the optimum ratio, they work in many field with many types the transformer. One of them prefer low ratio since the other prefer higher ratio.
two factories who made the transformer in my country not have too clear the optimum ratio, they work in many field with many types the transformer. One of them prefer low ratio since the other prefer higher ratio.
Last edited:
turns per volt is inversely proportional to core area, higher flux density of core, the higher the flux the lower the turns per volt..
frequency too, more turns per volt at 50 hz than say 400 hz...
up to the designer of the traffo, what he has in terms of core quality
so there has to be a complete set of design variables under consideration to come up with a design...
frequency too, more turns per volt at 50 hz than say 400 hz...
up to the designer of the traffo, what he has in terms of core quality
so there has to be a complete set of design variables under consideration to come up with a design...
mechanical noise is a construction issue, not an electrical one...
this is why potting in tar is the method of choice for many manufacturers, diy'ers are not keen on this...