A NEW KW TUBE Transmitter AM PDM BUILD - the beginning...

[WBear2GCR]

<snip>
In order to minimize telegraphed vibration, the plan is to take the motor and squirrel cage assembly, and create an enclosure and air flow plenum by cutting and laminating several sheets of one-inch thick styrofoam.  This assembly will then be enclosed in two-inch thick cellular foam rubber, before mounting in the rack cabinet.  

To reduce the intake and exhaust airflow noise, a labyrinth airflow path is planned.  The entire transmitter will be enclosed in a six-foot high rack cabinet, 22 inches wide, and 33 inches deep.  The floor of the cabinet is 80% open, and the cabinet frame is made up of cast and extruded aluminum sections, with removable sheet aluminum skins for the sides.  A sheet of 3/4 inch thick plywood will be used as a base for the plate transformer and other HV power supply components.  Around these components, 3/4 inch holes will be drilled to allow air to enter from beneath the plywood, through a pleated air conditioner filter.  By providing many moderate sized holes around the components, a large volume of air may enter, flowing at a lower speed than the intake or exhaust ports of the blower assembly.
 
The sides of the cabinet, and the front and rear, where possible, will also be lined with two-inch thick cellular foam rubber sheets.
 
Air will flow into the enclosed blower assembly, and then on to a plenum enclosing the 3CX2500F3 modulator, and the 4-400A driver tube.  The exhaust side of the plenum will also be perforated with 1/2 inch to 3/4 inch holes over a large area.  
Two or three layers of fiberglass sheet insulation will be placed above the tube plenum, in a manner creating a labyrinth path whereby the air flow much change direction several times, thus muffling the noise from the motion of the air, as well as the higher frequency noise caused by the tube heat radiator fin assembly.  The airflow will be sufficient that the cellular foam rubber and styrofoam are not exposed to extreme temperatures. </snip>

A few thoughts.

I would advise strongly against using any plastic foam inside a transmitter.
It's a severe toxic hazard in the event of a thermal event (aka fire).

In addition it is a poor material for sound absorption, and poor for blocking
sound transmission.

Also, it degrades with temperature and exposure to oxygen and UV.

To drop noise in a duct there are a variety of methods - they tend to be documented
in acoustics books and the more common ones in HVAC texts and papers.

The simplest idea is to create what works out to be a low pass filter. The simplest
one is in the form of a duct that opens briefly to a larger volume, and then continues
out to a smaller volume. Turns in a duct that are NOT lined with absorptive material that is
effective at the frequencies of interest, and effectively so, will not attenuate
much of anything.

What is an effective sound absorbing material?
Most are marginal.
Foam is one that is reasonably poor.
Every material that is commonly available works best in fairly thick applications.
Thinner applications work mostly at HF, and even then not so well.

One of the best turns out to be wool felt. If that can not be found then
cotton felt is not bad. There is also nowadays sold synthetic and combination
synthetic & cotton felts. Some of it has been fire retardant treated for domestic
and commercial installations.

Just like car mufflers, a tuned Helmholtz type filter system can be effective, and made
mostly from metal or similar. This is done all the time in HVAC installs to get rid of
fan noise, especially where high velocities are needed. Should be possible to apply
for us.

Rockwool and fiberglass are actually poor acoustically, and not effective
at lower frequencies until they are used in fairly thick applications.

There is "solid" material sold for use in domestic and commercial buildings similar
to the ceiling tiles but thicker, in 1 to 4" thick sheets. It's either fiberglass or
rockwool with some sort of binder. In some situations these might be useful.

So, stay away from styrofoam or Urea based "foams", latex "foam rubber" has
a short half-life so not a good one either.

Preventing sound transmission through surfaces is similar but a bit different.
The people who do sound studios and speaker enclosures have studied this in
depth.

Two main methods:
- decoupling of two surfaces
- constrained layer

The constrained layer idea is a way to turn vibrational energy into heat within
a composite surface. Essentially it's a high pass - low pass - high pass - low pass,
etc. So, a soft layer, stiff layer, soft layer, stiff layer... etc.

Sometimes space permitting one can incorporate one layer of granular material, like
fine sand as a layer. The sand which vibrates turns the vibration into heat. So that
would be between layers of something like metal - silicone rubber - metal - silicone rubber -
metal - sand - metal, silicone rubber - metal. Or wood/rubber/wood etc...

The decoupling of two surfaces is the classic method - aka "shock mounts".
Even if it holds up a wall or the entire floor or ceiling. For example.

We'd engineer the operation of a tube, or the design of a tank circuit, it's a good
idea to apply the same to sound control and reduction...

                           _-_-bear

The absorption of and resistance to transmission of sound is often shown in graphs
of attenuation vs. frequency WRT thickness...fyi.

EDIT: forgot to mention many foams are mildly electrostatically charged, and even those
that are not tend to pick up airborne dirt and grease quickly, and are not easily cleaned.
This creates an additional concern.

[K1JJ]

Interesting info, Bear.

Another solution to fan noise:   I just completed my outside installation of a Dayton 2C647D blower for the big linear amp and did measurements .  From peak power to fan noise, (S/N on ssb) looking at the SDR spectrum analyzer, I am seeing -45 dB delta. I am using the exact same mic and audio chain as on AM.  This is using simple flexible neoprene 4" ducting, about 3' long from the wall outside the house to the tube chassis inlet.

I've found that at least -30 dB of signal to noise is needed to avoid the "machine shop" sound. -40 to -45 dB  is superb. At -45 dB there is barely any sign of air noise, especially on the air when we mix in room ambience that can include residual hum, scope fans and other sounds that can be classified as a "quiet room."

When I go outside and listen to the sound of the blower; the slight mechanical whine and the sound of the air inlet, I estimate it is at least 80% of the noise energy being generated by the blower. The other 20% is inside the shack.   The air noise and mechanical sounds must be dissipated somewhere, and in this case most of it is outside.  

As for cold air, the temp starts at say, 30F outside and after an old buzzard transmission peaks out to about 110F above the tube.  That is warm to the touch, but not anywhere like starting at 70F.  So, for noise and temperature, this method is a reasonable solution for me. It is quiet and gives a temperature edge from fall to spring.  In addition, the cold air lets me run the blower at about 1/2 normal speed. This helps to cut noise tremendously. So there is a confluence of advantages that all add up to a silent sound when using a remote blower mounted outside the house during the cooler months.

As for tube seals, I think the delta change between 30F to 110F   vs:   70F to 150F  will affect the tube seals about the same. I've axed a few guys smarter than me about this and they either did not know or didn't think that starting a tube at 30F was a problem. I ran this external anode tube for about 2 years this way back ten years ago with no problems.

This also brings new meaning to "filament cold start."   I use a Variac to bring the fils up slowly in addition to a delay timer that doesn't let the tube key up with HV until seven minutes of fil heating have passed.

The good news is with the filament heated up and running - and the outside air at 30F, the air above the tube hovers around 55F.  So the runup from 55F to 110F is not too extreme.

T

[W3SLK]

Call me nutz but could you make a small audio amplifier, (I'm talking pocket transistor radio size), with some sort of common emitter circuit whereby you invert the noise created by the duct work and cancel it? Chust thinkin!

[K1JJ]

Call me nutz but could you make a small audio amplifier, (I'm talking pocket transistor radio size), with some sort of common emitter circuit whereby you invert the noise created by the duct work and cancel it? Chust thinkin!

Yep, that would probably work.   A microphone facing the blower area and then added to a mixer with the real signal 180 degrees out would do it.  They do this trick on luxury jets and cars I understand and who knows where else.  I'm surprised there isn't an MFJ box made specifically for this purpose.

Maybe the difficulty is adjusting and maintaining proper audio phase, (and variable amplitude needs across the specturm) but in this digital "pure signal" whirl with a computer screen interface it should be do-able to a fine art.

T

[Opcom]

This is using simple flexible fiberglass 4" ducting, about 3' long from the wall outside the house to the tube chassis inlet.

Do you mean the house HVAC type or the aircraft type?

[K1JJ]

This is using simple flexible fiberglass 4" ducting, about 3' long from the wall outside the house to the tube chassis inlet.

Do you mean the house HVAC type or the aircraft type?

Hi Pat,

Did I say fiberglass?  It's actually 450 degrees F neoprene silicon at $16/ ft ducting. Orange.   It has internal solid wire spiral but has a smooth inside.  I got the high temp cuz at one time I was pulling air out of the top of the tube.

 https://www.pegasusautoracing.com/productselection.asp?Product=3621&utm_source=google&utm_medium=cpc&utm_campaign=3621&gclid=EAIaIQobChMIqPXKubqB7gIVB7zACh0oWwfoEAQYBSABEgLA8PD_BwE

Nice stuff.

T

[Opcom]

This is using simple flexible fiberglass 4" ducting, about 3' long from the wall outside the house to the tube chassis inlet.

Do you mean the house HVAC type or the aircraft type?

Hi Pat,

Did I say fiberglass?  It's actually 450 degrees F neoprene silicon at $16/ ft ducting. Orange.   It has internal solid wire spiral but has a smooth inside.  I got the high temp cuz at one time I was pulling air out of the top of the tube.

 https://www.pegasusautoracing.com/productselection.asp?Product=3621&utm_source=google&utm_medium=cpc&utm_campaign=3621&gclid=EAIaIQobChMIqPXKubqB7gIVB7zACh0oWwfoEAQYBSABEgLA8PD_BwE

Nice stuff.

T

That's what I mean by 'aircraft' stuff. It's the best.

[K1JJ]

Help! QUESTION:

I have a PS filter choke that is causing a really loud 60 Hz acoustical hum. It is so loud that it gets into the mic audio.   The supply is a 3 KV / 6 KV DC output with parallel or series primary xfmr connections.   The 100 pound transformers feed a SS diode bridge rectifier.  (6A 1KV diodes)

The choke is a big 140 pound 15H BC type that goes from the diode bridge anode connection to ground, the same ground as the filter caps as negative lead filtering. I have about 133 uF at 10KV caps. About 400K of bleeders.

I tried a different choke and it also buzzes.  When I short out the choke or disconnect it from the circuit the hum goes away.   I checked the diode rectifiers and they are OK.

The supply works well at 3KV with no choke, but when the choke see 6KV OR I draw some serious current at 3KV, the acoustical hum begins.  Even idling at 6 KV the choke hums badly.

I haven't tried the choke in the positive lead yet.   It wud take some work to do and I figgered there is really no difference when in the negative lead.

I tried one transformer alone and there is no difference which eliminates circulating current between the transformers.

There is a one-second step start with 10 ohms in the primary.

Using a X100 scope probe there is barely any AC ripple on the output cap. Clean. Smooth DC.  But across the choke I am seeing the slope of ripple and a damped oscillation. Maybe this is normal for a choke, I dunno.

Would a standard   C-L-C filter in the hot lead put less stress on the choke, thus less 60 Hz noise?

What could cause this problem?  Any tests?    

T

[w8khk]

Tom, when my dad had his radio repair shop in the early 1940s, when someone brought in a radio and asked why it hums, he typically said "It probably does not know the words".  Perhaps that is the problem with the two chokes you have tested?

But on a more serious note, I would assume the chokes have a loose structure, possibly loose core material, or the bobbin on the core may be loose, allowing the vibration.  Not knowing the makeup of your chokes, I cannot suggest whether any of the hardware can be tightened to reduce the noise level.  I agree that the problem would probably persist, whether or not the choke is in the positive or negative supply leg.  I would also assume that the filter choke would output much less hum or noise if you use capacitor input (C-L-C) instead of choke input (L-C) filtering.

The choke is just doing it's job, but perhaps it does not need to be employed in your power supply.  One advantage of having a choke input supply is that it may allow the plate transformer(s) to run cooler by reducing the peak current, and spreading the capacitor charge current over a longer portion of the AC waveform.  Without the choke, the current flows through the transformer secondary in large current peaks for the short duration of time that the transformer output voltage is greater than the charge on the storage capacitor.  This high-current peak causes more heating of the transformer windings, due to the I-squared R losses in the winding resistance.  Assuming the transformers are sized conservatively for the load placed upon them, this may not be an issue at all.  With step-start, you should be good to go, and the only remaining advantage of keeping the choke in the circuit is ballast in the event of a tornado or earthquake.

Since you plan on employing the PDM approach, additional hum reduction may be achieved by the "feed forward" feature of the PDM generator chip.  But with the size of your filter capacitor bank, and the dual plate transformers, I cannot imagine this would be necessary.  If you run the higher voltage (you mentioned having two options) your PDM series tube will run more efficiently (assuming it can handle the peak voltage) because the "on" time will be of less duration for the same amount of output voltage.

Bottom line, you may not be able to make the chokes run silent without disconnecting them from the power supply circuit.  If they do not add any value to the rig, why bother?

[K8DI]

Couple thoughts...

Can you post a scope photo across the choke?

What happens if you disconnect 2/3 of the capacitors?

Is it possible with that much capacitance that the peak charging current is well past the choke’s current capability? Not knowing what it came from, I’m speculating, but I’m going to guess the capacitor in its original usage was on the order of 10-15% of what you’ve got....

Ed

[K1JJ]

Rick and Ed,

Thanks much for the comments.

Making progress:   After a some tests I've determined that the power transformers themselves are causing the hum. When I put 240 on each transformer  (or primaries in parallel) the acoustical hum is huge. This is with no load on the transformer, just idling with the sec free.   Both are the same.

When I put 120 V  on the primaries (pri in series) the hum is almost gone.  It acts like the primaries are saturating with circulating current when 240V  is applied.   I'm setting them up now as series-series config and hoping that will solve it. Once they are quiet I can add the bridge, choke and caps.    I plan to add a Variac to allow adjustment for the various rigs.  I prefer not to by controlling it with series/par taps, but in this case I am stuck with series.

BTW, I found 10 shorted diodes on one leg and 3 shorted on another leg of the bridge rectifier stack.  Old battle scars I guess.


T

[w8khk]

When I put 120 V  on the primaries (pri in series) the hum is almost gone.  It acts like the primaries are saturating with circulating current when 240V  is applied.   I'm setting them up now as series-series config and hoping that will solve it.

What is the input voltage rating of those transformers?  Do you have to go all the way to half voltage to cure the hum?  Maybe 240 is too much, but they may quiet down at 220?  A small buck transformer and parallel primaries would work ok?  A Variac test might be very revealing as to the saturation point.  Or, perhaps the transformers hum because they don't know the words either?