Hello Everybody,
I'm not an expert in electronics. Although I've studied computer engineering, my knowledge in electronic is mostly theoretical. So I need the help of some of you expert guys.
I own a tube preamplifier, and I'm experiencing a little hum coming from both speakers which is independent from the volume and from the selected source. The hearing of this hum is probably helped by the high sensitivity of my Class D power amp monoblocks.
Cabling is RCA, no XLR in my system.
I've already excluded any possibility of ground loops by making all the usual tests (tube rolling, cheater plugs, isolation transformers, removing all sources, etc...).
So I bought myself an oscilloscope and a mains' isolation transformer to measure things directly inside the preamp.
Now, in this preamp heaters are DC powered. here is an extract of the power supply schema:
First thing I noticed: although in the schematics the capacitor "3C10" is described as a 10000uF/16V, the one soldered in my preamp is 8800uF 25V. I assume this is good as well for its purpose, but not 100% sure.
Now, I've measured the rectified voltage between +Uf and -Uf with a multimeter and it stays around 11.80V. That's already a little bit strange, as it should be at least 12V. I assume that the 2940 here are the 12V ones, I cannot read it from the soldered one as they are too close to each other and cannot read the specs on them.
I also assume the designer has put three 2940 in parallel to obtain more juice (current) for the heaters (my preamp has a total of 8 tubes, n.4 12AX7 for the phono stage and n.4 12AT7 for the line stage).
With the scope, bypassing the DC, I've measured this strange signal (still between +Uf and -Uf):
It is around 20mV wide and the frequency is around 100Hz.
What do you experienced guys think? Is such level of noise normal with 2940 regulators? Or could this be the reason of the hum I'm hearing?
P.S.
I've also measured B+ and it has not such noise at all.
Thanks in advance
Kind regards
Marco
I'm not an expert in electronics. Although I've studied computer engineering, my knowledge in electronic is mostly theoretical. So I need the help of some of you expert guys.
I own a tube preamplifier, and I'm experiencing a little hum coming from both speakers which is independent from the volume and from the selected source. The hearing of this hum is probably helped by the high sensitivity of my Class D power amp monoblocks.
Cabling is RCA, no XLR in my system.
I've already excluded any possibility of ground loops by making all the usual tests (tube rolling, cheater plugs, isolation transformers, removing all sources, etc...).
So I bought myself an oscilloscope and a mains' isolation transformer to measure things directly inside the preamp.
Now, in this preamp heaters are DC powered. here is an extract of the power supply schema:
First thing I noticed: although in the schematics the capacitor "3C10" is described as a 10000uF/16V, the one soldered in my preamp is 8800uF 25V. I assume this is good as well for its purpose, but not 100% sure.
Now, I've measured the rectified voltage between +Uf and -Uf with a multimeter and it stays around 11.80V. That's already a little bit strange, as it should be at least 12V. I assume that the 2940 here are the 12V ones, I cannot read it from the soldered one as they are too close to each other and cannot read the specs on them.
I also assume the designer has put three 2940 in parallel to obtain more juice (current) for the heaters (my preamp has a total of 8 tubes, n.4 12AX7 for the phono stage and n.4 12AT7 for the line stage).
With the scope, bypassing the DC, I've measured this strange signal (still between +Uf and -Uf):
It is around 20mV wide and the frequency is around 100Hz.
What do you experienced guys think? Is such level of noise normal with 2940 regulators? Or could this be the reason of the hum I'm hearing?
P.S.
I've also measured B+ and it has not such noise at all.
Thanks in advance
Kind regards
Marco
Last edited:
10mF at the output of the regulator is a crazy thing. Use 100uF and no much more.
Tubes can perform quite well with AC for heating, so 20mV of ripple seems to be negligibly small to be the origin.
Suggest to take off tube by tube to see where the hum is introduced into the amplification chain.
Tubes can perform quite well with AC for heating, so 20mV of ripple seems to be negligibly small to be the origin.
Suggest to take off tube by tube to see where the hum is introduced into the amplification chain.
@Osvaldo de Banfield, thank you for your answer and your suggestions.
This is not a DIY preamp, it is a commercial product. So, unless you say the 10mF capacitor is a design error, I'll tend to keep things like they were designed.
I'm rather in search of a bad component, like a faulty capacitor / regulator.
But if you say that this 20mV ripple is not the problem of the hum I hear, I'll investigate other possibilities.
I've already tried removing the tubes. Without them (all) there is no hum. I'll try removing the tubes one by one like you suggested.
This is not a DIY preamp, it is a commercial product. So, unless you say the 10mF capacitor is a design error, I'll tend to keep things like they were designed.
I'm rather in search of a bad component, like a faulty capacitor / regulator.
But if you say that this 20mV ripple is not the problem of the hum I hear, I'll investigate other possibilities.
I've already tried removing the tubes. Without them (all) there is no hum. I'll try removing the tubes one by one like you suggested.
That is not noise but rectification ripple current going through some resistance.
How much ripple voltage is present at the first capacitor? It is possible that a ground loop
has been designed into the circuit. Does the (-) lead of the first capacitor go directly back to the rectifiers,
and then a separate trace goes from the (-) first capacitor lead to the regulators?
Yes, a 22uF or larger capacitor MUST be connected directly at the output pin of each regulator.
It's not just a good idea, it's the law. Also there must be > 3VDC across each regulator at all times,
including any input ripple minimum voltage sag.
The output capacitor ESR is somewhat critical for this regulator, needing > 30 mOhms and < 3 Ohms.
See figure 20 in the data sheet. How old is this preamp, the capacitors may be out of spec.
How much ripple voltage is present at the first capacitor? It is possible that a ground loop
has been designed into the circuit. Does the (-) lead of the first capacitor go directly back to the rectifiers,
and then a separate trace goes from the (-) first capacitor lead to the regulators?
Yes, a 22uF or larger capacitor MUST be connected directly at the output pin of each regulator.
It's not just a good idea, it's the law. Also there must be > 3VDC across each regulator at all times,
including any input ripple minimum voltage sag.
The output capacitor ESR is somewhat critical for this regulator, needing > 30 mOhms and < 3 Ohms.
See figure 20 in the data sheet. How old is this preamp, the capacitors may be out of spec.
Last edited:
Theoretically lm2941 is a low drop reg and 0.5...1v drop looks legit, but they never tell how good low drops are when the input output voltage is low...and you find a lot with older low drop chips that 3v in-out is still a lot better than 1V if you want good performance.They were good enough with regulating digital lines, not low noise analogue ...lm317 is still a lot better for that.Newer low drop designs retain and have even better performance at 0.5 v in out drop but they are usually too low power for a tube heater.
Aggreed - 20 mV of hum on the heater supply is nothing. Is either side of the heater supply grounded, or is it totally floating? Capacitance from primary to secondary can pass noise, which is common mode - regulators and filters don't stop it. If it's not grounded, ground -Uf at the input stage.
But it could be inductive pickup from the heater supply - the peak current will be 6A or more to get 1.2A of heater current. And there's full mains voltage somewhere in that box... hope it's far away from inputs!
But it could be inductive pickup from the heater supply - the peak current will be 6A or more to get 1.2A of heater current. And there's full mains voltage somewhere in that box... hope it's far away from inputs!
I think the +Uf is elevated, and goes to ground through a 62Kohm resistor.
This is a more extended extract of the schema:
Yes, inputs are far away and isolated by an aluminium shield.
Last edited:
You could try using Schottky diodes to get a slightly higher voltage for 3C11 (before the regulators). This would give the regulators a bit more headroom. The only disadvantage could be a slightly higher dissipation in your regulators.
You could also try to connect the heater elevation voltage to -UF instead of +UF.
Regards, Gerrit
You could also try to connect the heater elevation voltage to -UF instead of +UF.
Regards, Gerrit
Low voltage drop considering ripple free input, but if the source has ripple whose negative peak is under the regulator's drop plus output voltage, the reg'or can't do anything to alleviate output ripple. There must be always sufficient room to proper job.
Hi all, thanks for all the useful advices you gave to me.
In order to rule out the heater supply as a cause of the little hum I'm hearing, I think that I'll try with an external 12V regulated supply.
Changing the transformer windings is not a choice for now. So, the idea of using Schottky diodes in place of the usual ones in the rectifier is a good one. I'll try it if the result of the above mentioned experiment will rule that the heater supply is indeed the root cause of the hum.
In order to rule out the heater supply as a cause of the little hum I'm hearing, I think that I'll try with an external 12V regulated supply.
Changing the transformer windings is not a choice for now. So, the idea of using Schottky diodes in place of the usual ones in the rectifier is a good one. I'll try it if the result of the above mentioned experiment will rule that the heater supply is indeed the root cause of the hum.
What is the frequency of the hum you're hearing? 50/60hz would mean interference. 100/120hz means you're hearing ripple from not enough filtering.
It's not that simple, it seams modulated, not a simple 50Hz or 100Hz one. When measured with a "poor" iPhone spectrum analyser directly from the speaker, the peak seems to stay around 100Hz. But by ear it does not seem like a pure 100Hz buzz (I've heard some 100Hz example just as reference).
Aluminium as shield is totally ineffective at audio frequencies. It might be good at RF.
You only need to rule out the fillament of the first tube in the amplification chain.i'd put 100uF at the output of the Ic supplying it.Then the fillament potential lifting is better done with two resistors of 220k...1Megohm leaving that divider node and touching both ends of the first tube fillament.Also better to replace 3r21=10 ohms with 10nF capacitor
I must correct myself. When measured with the "poor" iPhone spectrum analyser app, it seems to peak at 150Hz. But to the ear it seems lower, around 50Hz.