The choke was buzzing slightly before, now it's perfectly quiet, but it was mainly electrical interference that was the problem. I bought it along with some other electronic odds and ends at an estate sale. I found the spec from a number on the choke, but can't find it, I'll need to go look it up again. It's rated 10.5Hy at 115ma, and is on maybe a 3.5"x3.5" square core. Pretty stout.
"made for choke input" - is there a specific spec to look for? A rule of thumb?
I have used surplus chokes spec'd for much higher current, or told the winder that is was for choke input use when ordering custom. I've had no issues, knock on wood...
Some applications and specs are hidden by some manufacturers.
Is is because they don't have the time, don't want you to know it is limited, etc?
I can not answer that.
But suppose the inductor says it is good for 500VDC, but only for 25VAC. That will tell you it is intended for a Capacitor input power supply. You have to get the ripple down to 25VAC in the first cap.
This is Not a swinging choke.
The inductors I use say reactor. All inductors are reactors. But using them for a choke input does require a choke that is either purposefully designed to do it, or just has such good insulation and enough core, and low enough loss core to do all that properly.
I could mention the choke manufacturer's name of what I always use and that always works for me, But I guess we try and not endorse or advertise, so I would rather not.
Is is because they don't have the time, don't want you to know it is limited, etc?
I can not answer that.
But suppose the inductor says it is good for 500VDC, but only for 25VAC. That will tell you it is intended for a Capacitor input power supply. You have to get the ripple down to 25VAC in the first cap.
This is Not a swinging choke.
The inductors I use say reactor. All inductors are reactors. But using them for a choke input does require a choke that is either purposefully designed to do it, or just has such good insulation and enough core, and low enough loss core to do all that properly.
I could mention the choke manufacturer's name of what I always use and that always works for me, But I guess we try and not endorse or advertise, so I would rather not.
I am experiencing a similar issue:
I get an annoying buzzing out of my speaker. When I introduce a 1 uF before the choke, its is almost completely gone. I will now experiement with different values.
what is annoying me though: I am using choke-input everywhere: My driver stage for the power amp. The separate PSU of the 300B, even my raw-DC-supply for Rod's DC-Regs: All choke input and all of the 100% quiet. I do not yet get why it is this time different.
THE PSU looks like this:
- Double C-Core from AE Europe 350-0-350
- Tried both, AZ1/RGN1064-Tube Rectification and Hexfreds. Result is the same, Buzzing is a bit louder with the Hexfreds.
- LCLC with 9H(35ohm)-50uF (ASC Oil)-13H(35ohm)-100uF(ASC Oil)
- 60mA DC
I raise the 60Ohm secondary winding resistance of Will's Transformer from 60ohm to 200 Ohm for the tube rectifier, but result was the same.
As the materials of the chokes are the same as withinthe rest of my system where quietness is king...I am a bit annoyed not to find the real origin....the cap-trick seems to me like a work-around...
I get an annoying buzzing out of my speaker. When I introduce a 1 uF before the choke, its is almost completely gone. I will now experiement with different values.
what is annoying me though: I am using choke-input everywhere: My driver stage for the power amp. The separate PSU of the 300B, even my raw-DC-supply for Rod's DC-Regs: All choke input and all of the 100% quiet. I do not yet get why it is this time different.
THE PSU looks like this:
- Double C-Core from AE Europe 350-0-350
- Tried both, AZ1/RGN1064-Tube Rectification and Hexfreds. Result is the same, Buzzing is a bit louder with the Hexfreds.
- LCLC with 9H(35ohm)-50uF (ASC Oil)-13H(35ohm)-100uF(ASC Oil)
- 60mA DC
I raise the 60Ohm secondary winding resistance of Will's Transformer from 60ohm to 200 Ohm for the tube rectifier, but result was the same.
As the materials of the chokes are the same as withinthe rest of my system where quietness is king...I am a bit annoyed not to find the real origin....the cap-trick seems to me like a work-around...
Blitz,
. . . Back to the threads earlier discussion . . . How much load current versus the choke current rating? Steel chassis? Is the choke made for 50/60Hz, or for 400Hz? Is the Manufacturer making consistently good quality chokes?
For a choke input supply, the choke has 500V peak to peak across the choke windings for a 350-0-350vrms transformer secondary). Not all chokes are made for that.
For cap input supply, a 1uf cap has 1,590 Ohms of capacitive reactance at 100Hz full wave rectification, and with a 60ma load on that 1 uF cap there is 95.5V peak to peak ripple. In that case, the 9H choke only has 95.5V peak to peak across the same windings (5 times less than with the choke input supply).
By the way, 1uF and 9H is resonant at 53Hz (its good that it is not resonant at 100Hz, and that you are not using half wave (50Hz) rectification.
There has to be a real cause, and therefore a real solution to your problem. Good luck finding the cause.
. . . Back to the threads earlier discussion . . . How much load current versus the choke current rating? Steel chassis? Is the choke made for 50/60Hz, or for 400Hz? Is the Manufacturer making consistently good quality chokes?
For a choke input supply, the choke has 500V peak to peak across the choke windings for a 350-0-350vrms transformer secondary). Not all chokes are made for that.
For cap input supply, a 1uf cap has 1,590 Ohms of capacitive reactance at 100Hz full wave rectification, and with a 60ma load on that 1 uF cap there is 95.5V peak to peak ripple. In that case, the 9H choke only has 95.5V peak to peak across the same windings (5 times less than with the choke input supply).
By the way, 1uF and 9H is resonant at 53Hz (its good that it is not resonant at 100Hz, and that you are not using half wave (50Hz) rectification.
There has to be a real cause, and therefore a real solution to your problem. Good luck finding the cause.
Last edited:
Hi
Choke is like SE trammy vs Idc and Vac.
Gap is calculated for Ia (0.6-0.9 Tesla for example) AND Iac surimpresed (variation of Ia with modulation) for 0.4-0.5Tesla.
In this case Idc + Iac and Idc -Iac let the induction always in the linear zone
0.6 - 0.4 = 0.2 tesla
0.6 + 0.4 = 1.0 tesla.
Yan24
Choke is like SE trammy vs Idc and Vac.
Gap is calculated for Ia (0.6-0.9 Tesla for example) AND Iac surimpresed (variation of Ia with modulation) for 0.4-0.5Tesla.
In this case Idc + Iac and Idc -Iac let the induction always in the linear zone
0.6 - 0.4 = 0.2 tesla
0.6 + 0.4 = 1.0 tesla.
Yan24
Thanks for the answer and you confirm my suspicion: The must be a real cause which I now only cure with the cap...
Some more data:
- The chokes are big guys, AMCC160 amorphous core. I am using ten of these all together in my system and all are completely quiet, even at much higher voltages and currents. I believe this one is rated for 200mA.
- i am living in Germany, so we have 50Hz.
- My impression is, that the buzzing sound did not come from the chokes, but from the Double-C-Transformers AE Europe did for me...which is surprising as those are reqlly huge and very well made. So I do not believe that they are the cause, but showing the symptom.
What can be, I need to try: Bringing much more distance between the chokes and the rest of the amplifier. Currently they are quiet near the actual tube stage/opt of the line stage and there is no steel chassis, it is more breadboard design currently and the chokes are as well open frame.
Some more data:
- The chokes are big guys, AMCC160 amorphous core. I am using ten of these all together in my system and all are completely quiet, even at much higher voltages and currents. I believe this one is rated for 200mA.
- i am living in Germany, so we have 50Hz.
- My impression is, that the buzzing sound did not come from the chokes, but from the Double-C-Transformers AE Europe did for me...which is surprising as those are reqlly huge and very well made. So I do not believe that they are the cause, but showing the symptom.
What can be, I need to try: Bringing much more distance between the chokes and the rest of the amplifier. Currently they are quiet near the actual tube stage/opt of the line stage and there is no steel chassis, it is more breadboard design currently and the chokes are as well open frame.
Yes, there is a real cause...
If you design a choke input supply without a capacitor right after the rectifier, and apply a dc load, consider what happens:
- there will be an instant during each mains cycle where the voltage at the node between the rectifier output and the choke's input-side reduces to zero. At this exact instant, the rectifiers are non-conducting, and the input side of the choke is floating. The duration of floating-time depends on the rectifier construction. (I am neglecting reverse-recovery pulses for the sake of simplicity in telling the tale...). This happens every 10ms in a 50Hz mains network.
- At the zero-input voltage instant, the current in the choke may be non-zero, sustained by the field in the choke.
- Now the choke's field begins to collapse. As usual, this action forces an abrupt change in the voltage at the floating end of the choke, which rises sharply.
- Next, the rectifiers conduct again, as the mains waveform rises. Current through the rectifiers crashes into the ragged waveform from the choke. It all then wriggles around for a while, according to the various LC resonances from the choke, power transformer, parasitics &c, and then settles down again.
- Adding a suitable cap (about 680nF 1500V stacked-construction MKP is a good choice) keeps the voltage from getting all the way to zero, & prevents the « disconnect » activity that causes the problems.
Always add a cap right at the rectifier, and using a suitable HV probe, check that the voltage reliably keeps above zero.
OTOH, don't overdo the cap value, or the regulation of the dc output will suffer, and the rectifier current-pulses will deepen.
If there is sufficient DC current draw then in the ideal situation there will always be a rectifier which is conducting. The choke can send its terminal voltage below zero so all that happens is that at the AC zero crossing the flow commutates from one rectifier to the other. Circuit parasitics can complicate this, of course.
If there is insufficient DC current draw then rectifiers will switch off, but not necessarily at the zero crossing. To get zero crossing turn off you need to draw exactly the critical current, which is unlikely by accident.
If there is insufficient DC current draw then rectifiers will switch off, but not necessarily at the zero crossing. To get zero crossing turn off you need to draw exactly the critical current, which is unlikely by accident.
Rod,
I always appreciate your thoughts and deep knowledge. My logic was very simple:
I have LCLC supplies all over the place in my system - without tis first cap. And they are dead quiet.
Sometimes they draw 120mA (two 300B) at 450V, sometimes only 20mA at 200V (like e182cc transformer coupled), this time its for 300V at 60mA. Soooo...you cant really say that you have to have this small cap ALWAYS with choke input as the rest of my system is really quiet.
I never had the issue before, so I try to understand the delta, why this time when all the other times I came away with it...maybe I should change some chokes to see any difference in their behavior.
I always appreciate your thoughts and deep knowledge. My logic was very simple:
I have LCLC supplies all over the place in my system - without tis first cap. And they are dead quiet.
Sometimes they draw 120mA (two 300B) at 450V, sometimes only 20mA at 200V (like e182cc transformer coupled), this time its for 300V at 60mA. Soooo...you cant really say that you have to have this small cap ALWAYS with choke input as the rest of my system is really quiet.
I never had the issue before, so I try to understand the delta, why this time when all the other times I came away with it...maybe I should change some chokes to see any difference in their behavior.
For 300V supply at 60 mA draw, as stated above, the critical inductance is 5.3H. So the inductance is properly sized, unless the choke is badly made....or its specs are not true!
Hmmm...if I got it right:
Lc = Rc/940
Lc=300/60=5kohm, correct ?
F=50Hz, so 5000/950=5,3H ....so 9H should be good enough.
The idea to check grounding is worthwhile...I guess I will build a complete separate external PSU to seaparte the dirty AC /rectifing division from DC...as well for stray purpose...
Lc = Rc/940
Lc=300/60=5kohm, correct ?
F=50Hz, so 5000/950=5,3H ....so 9H should be good enough.
The idea to check grounding is worthwhile...I guess I will build a complete separate external PSU to seaparte the dirty AC /rectifing division from DC...as well for stray purpose...
My bad.
Looking at my own examples again, the cap actually addresses the situation when there is less than the critical load for the circumstances. In this case, the current in the inductor falls to zero, and the diode turns OFF. This turn-OFF is likely to result in some ringing, since there is always some level of parasitic capacitance.
For Blitz: maybe you could measure around the choke and rectifier to see what kind of waveform you have.
You don't need to attach the scope-probe for this kind of measurement; just holding it near to the components will sense the waveform by coupling.
If the inductor current falls to zero (every mains cycle) you will see a ringing waveform. If you have that, the chances are that the inductance is not what you expect - or that the current is lower, perhaps. It's always worth measuring twice, to be sure.
Oh darn,
Sorry, I missed a few things:
1. I stand corrected, it is not 500V across the choke, it is 0.42 x 350 = 147V peak to peak across the choke winding. That is still different than using a 1uF input cap, 60mA DC load, and 100Hz full wave, resulting in 95.5V peak to peak. That is a 3.7dB difference.
2. I misread Blitz's original statement, it was buzz in the speaker (not the choke). That points to a different cause of the problem.
A. Magnetic coupling from the Choke to the output transformers (or to interstage transformers if they are employed)?
B. Magnetic coupling from the Power Transformer to the output transformers (or to interstage transformers if they are employed)?
How about distance and magnetic field orientations of A. and B.?
C. Choke input with the first cap returning to a central ground point, and not returning directly (separately) to the B+ secondary Center Tap. (ground loop) Versus: Then adding a 1 uF (cap input, not choke input) and returning the 1uF cap directly (separately) to the B+ secondary Center Tap. (no ground loop)
3. I have seen chokes that are only rated for quite small voltages. They would not work for use in choke input B+ power supplies. With respect to George Orwell: All chokes are created equal, but some chokes are more equal than others.
Sorry, I missed a few things:
1. I stand corrected, it is not 500V across the choke, it is 0.42 x 350 = 147V peak to peak across the choke winding. That is still different than using a 1uF input cap, 60mA DC load, and 100Hz full wave, resulting in 95.5V peak to peak. That is a 3.7dB difference.
2. I misread Blitz's original statement, it was buzz in the speaker (not the choke). That points to a different cause of the problem.
A. Magnetic coupling from the Choke to the output transformers (or to interstage transformers if they are employed)?
B. Magnetic coupling from the Power Transformer to the output transformers (or to interstage transformers if they are employed)?
How about distance and magnetic field orientations of A. and B.?
C. Choke input with the first cap returning to a central ground point, and not returning directly (separately) to the B+ secondary Center Tap. (ground loop) Versus: Then adding a 1 uF (cap input, not choke input) and returning the 1uF cap directly (separately) to the B+ secondary Center Tap. (no ground loop)
3. I have seen chokes that are only rated for quite small voltages. They would not work for use in choke input B+ power supplies. With respect to George Orwell: All chokes are created equal, but some chokes are more equal than others.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.