• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

DCR requirments for Chokes

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Those on this forum who obtain and use mercury vapour rectifiers worry me. Mercury is VERY nasty stuff.

Sooner or later a tube is going to be accidentally broken, and folk made ill by the mercury.

Those who leave their tubes on full display are really taking a risk. There is little real danger from some visiting little kid breaking a signal tube, even with HT on, and even with power tubes, he's likely to only get a severe shock. But with mercury vapour tubes he's going to end up in hospital, or perhaps dead. Shock risk (AC in the tube) and poisoning.

Mercury vapour rectifiers were never intended for home equipment. They are dangerous, and put out ultraviolet light too.

This guy disagrees on your view of mercury :p YouTube
 
Kodabmx,

That Cody guy on Youtube is an idiot. Or, malicious. How do we know his iron ball is actually iron? It seems to float much too well - mercury is only about 30% more dense than iron.

However, at least he has his mecury in a plastic vessel and not in thin glass under high vacuum. He's not going to have glass fragments and mercury flying thru the air if he breaks the vessel.

The larger mecury vapor rectifiers, three phase, were known as "Mekons". If you've ever seen one, and read the 1930's to 1950's English comic "Eagle", you'll understand the term "Mekon". Mekons ranged from 10 inches or so across to 6 foot or more. They were always installed in protective enclosures with interlocks on the door. Through the glass, you could see the pool of mercury in the bottom.

A group of old radio nuts operated a museum here on a volunteer basis. Somehow they acquired an old mekon rectifier. They had it on display behind glass. However, after a while the authorities found out and made them get rid of it. Too much risk some accident would occur and break the glass. Due to the size and being under substantial vacuum, glass and mercury would go everywhere.
 
Last edited:
The sims by jhs look nice, but where did the 24H value of speaker field inductance come from, apart for the obvious thing that a simple calculation shows that 24H is required to resonate with 0.07 uF at 120 Hz?

I hadn't responded before because I had been looking for the typical inductance of a speaker field coil, without much success.

24 Henries seems very high, given that:
a) it is way greater than is needed. The reactance at 120 Hz is 18 kohm. The sim bears that out, the ripple on the output is given as about 600mV even without the cap. Since the signal level on the output transformer primary will be about 150V at ordinary listening volume, the hum is 50 dB down - audible, but only just. In practice, the output of a pentode is essentially constant current, so what ripple goes into the output transformer primary is considerably suppressed by the pentode. So you are not going to hear any hum from this source.
b) it is not very believable, given that since the pole pieces are operated in saturation, the speaker field is effectively a choke with an airgap measured around an inch,
c) Chokes used in this application were typically 5 to 10 Henry.

This guy disagrees on your view of mercury :p YouTube

I had never heard of Cody. I watched a couple more of his videos. Definitely he is a mad idiot. Loves doing pointless experiments with mercury, or nitric acid, or both at the same time.
 
The operating inductance of the field coil winding could be noticeably lower than 24H (the 2H resonant frequency) based on the relatively low level of higher harmonics from the rectifier - ie. the attenuation notch doesn't have to be at 2H resonance in practice.

I haven't measured a field coil inductance. I'd also be cautious of any reported measurements unless it was confirmed to be at the nominal operating dc current, and using an appropriate measurement technique.
 
The operating inductance of the field coil winding could be noticeably lower than 24H (the 2H resonant frequency) based on the relatively low level of higher harmonics from the rectifier - ie. the attenuation notch doesn't have to be at 2H resonance in practice.

Huh? You are right about the harmonics - at the output of the rectifier, the second harmonic (120Hz) is by far the largest amplitude frequency present. And the higher harmonics drop off further due to the electros.

Ripple would be pretty much inaudible in this set anyway, even though 120Hz is well in-band in even cheap speakers, and this radio has pretensions of quality seen in several aspects of the circuit. See my earlier post on this.

This means that if you are going to put a notch in the filter response, it had better be at 120Hz. No point in putting it at any higher frequency.
 
Trobbins,

You seem to be a little confused. Reducing c would give negligible cost saving. Reducing L may do, if its a choke, but you need to reduce it by a significant amount, not just 10% or so. So reducing L is not a cost saving strategy in this case.

In any case, the primary purpose of the field coil is to energise the magnetic field through the voice coil. Once you design the whole speaker to do that adequately at minimum cost, the field coil inductance is merely what you happen get. If that is sufficient, and it usually was, in conjunction with a couple of electrolytics of the sizes available back then, to suppress ripple sufficiently, well and good. If not, you do something else, like add a stage of RC filtering or perhaps adding a C to resonate the field coil at 120 Hz.

Resonating chokes at 120Hz was sometimes done. Not very often. But it was done, because you could reduce the size of the choke SIGNIFICANTLY.

Resonating the field coil of speakers used as chokes is quite unlikely to let you choose a cheaper speaker. With speakers, the field coil inductance is what you get - you can't specify a lower inductance as then the voice coil will not be properly energized.

As I said in a previous post, I suspect the capacitor shunting the field coil in the Majestic is not there to suppress ripple, which is going to be inaudible anyway due to other circuit features. I think it was added during prototype testing to cure some other subtle problem - instability perhaps. It's in an odd position to do that, but perhaps the test engineer discovered that it worked, and its cheap.
 
Last edited:
As you say, the speaker and field coil is what it is. Imho they would have chosen 70nF to suit the situation. What I was getting at is that the resonance freq of those two parts does not have to be pedantically at twice mains frequency - it can be significantly higher (say 20%) and still likely provide noticeable reduction of twice mains hum.

I think it is pretty obvious that twice mains hum reduction was the reason for adding in the 70nF cap. No other uses of that part placement have shown up in any references from that era that I know of. To propose that it wasn't for hum alleviation would I suggest need some more support than hypothetical reasons, and that probably is best done by tests on the actual equipment.
 
As you say, the speaker and field coil is what it is. Imho they would have chosen 70nF to suit the situation. What I was getting at is that the resonance freq of those two parts does not have to be pedantically at twice mains frequency - it can be significantly higher (say 20%) and still likely provide noticeable reduction of twice mains hum.

But there is no point in being 20% out. It gives no performance advantage, and no cost advantage. One sub-microfarad cap is the same cost as another. Caps back then were available no better than 20% tolerance. 20% cap error is only 10% frequency error - there's a square root in the math. More reason to get it right.

You are dead right that only testing on the actual equipment can really establish what it's there for. And sometimes even testing won't reveal the reason. For instance, remove the grid-stopper from a circuit. 9 times out of 10 it won't make any difference. In production, one must ensure All units will work, so the stopper is added.
 
Another Resonant Filter

Long before I got myself a copy of Beitmans schematics of 1930s radios I'd tried a couple of resonant PS filter systems. It was simply a matter of connecting caps in parallel with the choke & watching the scope. This is one of those.

I'd also tried using a heater transformer secondary as a choke for a low voltage, higher current supply, the primary resonated with a cap. The cct is here somewhere, I'll try simulating that later.

There are several examples of both in Beitmans 1930s receiver ccts.

The amp I pulled off the shelf is a lot heavier than I recall, must be caused by age, mine! This amp has not been run in at least 40yrs. The caps are most likely shot.:eek:

50W Amp Pre1960 with Resonant Choke Filter PP 6550s or KT88s, Pentode Connected, 50W+
Hammond 1700 Series OPT 5K P-P
Separate Hammond Transformers for Plate & Heater Supplies
Surplus No Name Choke Regulated Screen Supply,
7AU7 both sections paralleled_6AU6 error amp_991/NE16 ref 6AU6, then 6SN7 Both Sections
Paralleled Phase Splitter AC & DC Balance Controls & Fixed Bias Adjust

The HV PT & Choke Resistance measured today. HV PT Primary R is an estimate. The actual measured HV at the first cap was 520V at the time of build. Needs some numbers for the filter cap ESRs for the simulation to have better accuracy. Estimate for the 5AR4 rp are from RDH4 p1171
 

Attachments

  • 50W Amp Using Resonant PS Filter A.jpg
    50W Amp Using Resonant PS Filter A.jpg
    97.3 KB · Views: 106
  • 50W Amp Using Resonant PS Filter B.jpg
    50W Amp Using Resonant PS Filter B.jpg
    152.3 KB · Views: 107
  • 50W Amp Using Resonant PS Filter C.jpg
    50W Amp Using Resonant PS Filter C.jpg
    98.2 KB · Views: 108
  • 50W Amp Using Resonant PS Filter D.jpg
    50W Amp Using Resonant PS Filter D.jpg
    145.5 KB · Views: 106
  • IMG_4353 50W Amp LHS E.jpg
    IMG_4353 50W Amp LHS E.jpg
    312.4 KB · Views: 99
  • IMG_4354 50W Amp Underside E.jpg
    IMG_4354 50W Amp Underside E.jpg
    644.3 KB · Views: 63
Last edited:
I'd also tried using a heater transformer secondary as a choke for a low voltage, higher current supply, the primary resonated with a cap.

That would not be efficient use of components as a heater transformer would have no air gap and so the core would saturate at quite low current. And of course the DCR is twice what it needs to be for the window size, as there are two equal volume windings instead of only one.

You may get nearly the same ripple suppression for a given degree of regulation with a resistor instead of the transformer.
 
Yet Another Resonant Power Supply Filter

This one from a 1930s Brandes Radio. With very little in electrolytics manages a very respectable reduction in PS ripple.
 

Attachments

  • Brandes B15 Receiver Using Resonant Filter A.jpg
    Brandes B15 Receiver Using Resonant Filter A.jpg
    153.9 KB · Views: 37
  • Brandes B15 Receiver Using Resonant Filter B.jpg
    Brandes B15 Receiver Using Resonant Filter B.jpg
    72.8 KB · Views: 43
That would not be efficient use of components as a heater transformer would have no air gap and so the core would saturate at quite low current. And of course the DCR is twice what it needs to be for the window size, as there are two equal volume windings instead of only one.

You may get nearly the same ripple suppression for a given degree of regulation with a resistor instead of the transformer.
All too true. But I guess the core saturation was not near total, still got a cap to resonate the cct. I've used many Hammond 125 PP Series over 60 years in SE ccts with no problems. The LF performance is not great but for ordinary use it is OK. There were gapped SE OPT in the Hammond line back in those daze, don't recall ever buying one.

I'm not a critical listener, I don't try to analyze whether the Flugel Horn in the Orchestra is off key. For me the challenge is in the cct design. Beyond that there are many more important things to accomplish with my time. At present that would be using a chain saw & wood splitter to demolish the dead Ash trees here on this place. 4.4 Acres takes a lot of work. Been here 50 yrs.:)

The image is 3W & Baxandall Tone Contriols. There is a magnetic phono preamp or mike input. The OPT is a Jensen 2430 equivalent to the H125D. Probably made for them by Hammond. Built more than 50 yrs ago, in everyday use, including today. Still sounds OK.:)
 

Attachments

  • 3 WATT AMP RHS 9W 20C.jpg
    3 WATT AMP RHS 9W 20C.jpg
    102 KB · Views: 53
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.