Homebrew AM Transmitter - 60 Hz Hum Help Please
Ty Unes
http://2xl.homelinux.com/riverty/gallery/main.php?g2_itemId=1063
The TX works great and has good audio. Transmitted signal covers my house
area very well on a wire antenna. There's just one problem:
Simply put, with or without an audio signal applied, the TX has 60 Hz hum
in the audio. It's a little too much for comfortable listening. I have
not been able to isolate as to where or why and I'm asking you more
seasoned pros if you have any suggestions. Should I try DC heated
filaments instead of AC? I thought about using coaxial cable on the audio
input coming from the audio transformer and grounding the shield. Any
other suggestions that I could try?
Thank you all for your time!
rt...
Peter Wieck
Just for giggles, try:
a) Shielding your tuning cap.
b) Shielding your output cap.
c) Installing an L-shaped shield around the coil between it and the
tubes/transformer.
In that order. Aluminum beer-can or heavy-duty foil grounded to the
chassis with a alligator-clip should do it for testing purposes,
something more elegant if this works.
Peter Wieck
Melrose Park, PA
dke...@hotmail.com
Hi
Looking at your pictures, I can't clearly see but you seem to
have ground wires running every which way. I don't see
good star connections. I also don't see good isolation of the
transformers CT lead only going to the first filters lead and then
only to the next and so on.
Also, I see a capacitor between circuit ground and the chassis.
This should be only needed for AC/DC chassis. The audio
circuit should have a solid ground tie close to its input.
Of course, this depends on the source, ground loops should be
avoided.
Dwight
Engineer
Is it modulation hum, aka "tunable hum", on station at the receiver,
or is it 60 Hz getting into the audio before modulation? Solutions
are very different.
Cheers,
Roger
Jim Mueller
Is it 60Hz hum or 120Hz? If it's 120Hz, it's coming from the power
supply (assuming a full wave rectifier) and better filtering or grounding
should help. If it's 60Hz, it's coming from something else. Do you have
a schematic we could take a peak at?
60Hz is easily distinguishable from 120Hz by ear, but if you haven't paid
attention to the sound of hum your ear might not be trained to tell.
--
Jim Mueller wron...@nospam.com
To get my real email address, replace wrongname with dadoheadman.
Then replace nospam with fastmail. Lastly, replace com with us.
Ty Unes
> Just for giggles, try:
>
> a) Shielding your tuning cap.
> b) Shielding your output cap.
> c) Installing an L-shaped shield around the coil between it and the
> tubes/transformer.
>
> In that order. Aluminum beer-can or heavy-duty foil grounded to the
> chassis with a alligator-clip should do it for testing purposes,
> something more elegant if this works.
Peter,
Thank you for your response. I've been wanting to do something with the
tuner for quite a while now and your post just gave me the kick in the @$
$ I needed. It's not pretty but it does what it should. It was a chore
tuning the rig as just touching the tuner knob would add 20? 40? Khz so
one had to work with it a bit. Now, the signal tuning is much more robust.
I decided I would try and track my progress on my gallery so I moved some
pictures around and added some new ones. Version 1 and Version 2. You may
check out my mods here:
http://2xl.homelinux.com/riverty/gallery/main.php?g2_itemId=1063
It still hums good though. Still trying...
rt...
Ty Unes
> Hi
> Looking at your pictures, I can't clearly see but you seem to
> have ground wires running every which way. I don't see good star
> connections. I also don't see good isolation of the transformers CT lead
> only going to the first filters lead and then only to the next and so
> on.
> Also, I see a capacitor between circuit ground and the chassis.
> This should be only needed for AC/DC chassis. The audio circuit should
> have a solid ground tie close to its input.
> Of course, this depends on the source, ground loops should be
> avoided.
> Dwight
Dwight,
Thank you for your post.
I had everything star grounded previously and for some reason tried what
I have. It gave me the same results and I concluded that either way was
not the problem. Of course I could go back.
I don't really understand "good isolation of the transformers CT lead
only going to the first filters lead and then only to the next and so
on." Before, this was connected to chassis ground. Are you saying I
should chassis ground everything again, yet keep the CT isolated?
Shouldn't this be connected to chassis ground as well?
The source is usually a Sirius Satellite Radio. 1/8" stereo to dual RCA
on the TX.
rt...
Ty Unes
> Is it modulation hum, aka "tunable hum", on station at the receiver, or
> is it 60 Hz getting into the audio before modulation? Solutions are
> very different.
> Cheers,
> Roger
Roger,
Good question. It is NOT tunable. I would say it's getting into the audio
before modulation. I get a "whine" that I can tune out most of the time
but the ratio from "whine" to wanted audio is sufficient that it's not
much of a problem. It's mostly present at night and I attribute it to the
AM band opening up and night and my signal mixes with others.
You post got me thinking... I'm not sure what to do about it though.
Ty Unes
> Is it 60Hz hum or 120Hz? If it's 120Hz, it's coming from the power
> supply (assuming a full wave rectifier) and better filtering or
> grounding should help. If it's 60Hz, it's coming from something else.
> Do you have a schematic we could take a peak at?
>
> 60Hz is easily distinguishable from 120Hz by ear, but if you haven't
> paid attention to the sound of hum your ear might not be trained to
> tell.
Jim,
I'm afraid my ear may be off here. I would say it's 120Hz. I've been
fooled before but 60Hz is pretty low in the audio spectrum. For me, it's
hard to tell when I'm listening to a small 4" speaker out of a 1948
Stewart Warner. I've tried on a few "new" radios as well. I say it's
120Hz.
Pictures and schematic are found here:
http://2xl.homelinux.com/riverty/gallery/main.php?g2_itemId=1063
Version 2 is current. I took these pictures last night!
The schematic is here:
http://2xl.homelinux.com/riverty/gallery/main.php?g2_itemId=871
These are pictures I took after I first built the TX. It's been through
some modifications since trying to fix and fine tune different things. I
could take new measurements if you think it's needed. But I should say
that besides the audio hum, the TX works pretty well!
Thank you very much for your help in this. I'm just at the top of my
capabilities here and hoping to gain some knowledge.
rt...
John Byrns
Ty Unes <riv...@gmail.com> wrote:
Assuming the problem isn't "modulation hum", which has been thoroughly
discussed many times previously in this group, I would take close look
at what appears to be an unshielded audio input transformer located in
close proximity to the power transformer. The audio input transformer,
especially an unshielded model lacking a hum bucking design, should be
located at a diagonally opposite corner of the chassis from the power
transformer, a chassis relayout may be called for.
--
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
nesesu
> In article <hdh8cr$4c...@news.eternal-september.org>,
>
> - Show quoted text -
Further to what John says, you could first try unbolting the AF input
transformer from the chassis and rotating it 90 degrees left and right
as well as turning it on its side and on end, with the transmitter
running, and see if the hum level changes. If that has significant
effect, then as John suggested, move it to the opposite corner from
the power transformer and put it UNDER [inside] the chassis.
Neil S.
dke...@hotmail.com
Hi
Think in terms of the ripple path. On the positive side of the filter
caps, you put resistors to isolate stages of your filter.
Most don't think about the ground side. The ripple current
to the filter caps goes through the ground path. It is better to
run the center tap wire directly to the first filter cap and then
a wire to the next and so on.
If you connect things to chassis ground, connect at the last
stage of filtering is best ( farthest away from the ripple current. ).
It does sound like your problem may be microphonics or
inductive coupling some place.
It may even be insufficient filter caps. Try increasing the
size and see if that helps.
As was mentioned, us a 'scope and determine if it is 120Hz
or 60Hz. If it is 60Hz, you can put two fixed resistors of about 200
ohms from
each lead of the filaments to ground and then a pot of about 500ohms
and a high value resistor to the audio input. This will allow you to
null
out any 60Hz. ( if the filament has a center tap, use that to ground
instead
of the two 200 ohm resistors ).
If it is 120Hz, it is bad ground paths and filtering.
Dwight
Engineer
wrote:
(snip)
> Dwight
Dwight, your point about ripple current is also important in radio
restoration. It's tempting to just drop the negative of a tubular B+
replacement cap to the nearest ground point. Don't! Make sure you
take it to the PT center tap or, for AC/DC radios, to the correct
mains input line - mostly on the VC mains switch (radio side!)
Cheers,
Roger
Ty Unes
pictures and posted here:
http://2xl.homelinux.com/riverty/gallery/main.php?g2_itemId=1098
I was able to determine that it IS 120Hz in the audio.
I removed the audio tranny from it's home on the top and placed it as far
away as possible from the power tranny. It's now bolted right behind the
RCA input jacks. Ground here is on one of the mount posts for the RCA
jack plate. I electrically restored my previous ground points and removed
the ground wires. Now the chassis is ground. All power components are on
the power side of the chassis. It still tunes good and has good audio but
hum is still present as it was before. No better, no worse.
I tried a few different CT points and found that I can make the hum
worse, but not better. I'm starting to think it might be insufficient
filtering. Two 47 mF caps seem a little small to me. Most radios use 100
mF or 60 mF or both?!?!?
If there is a better way to connect the CT from the PT then I don't know
what it is. Seems I've tried every combination that would work. Maybe I'm
still missing something?
One comment. It seems the hum is "stable" in that nothing seems to effect
it so far. Moving components, trying different ground points and schemes,
using different locations and outlets. To me, this lends some credibility
to insufficient filtering but I don't know this. I don't have any spare
caps to try right now so that's out.
What's with the cap on the ground side of the audio tranny? This is
supposed to be there by the author for some reason.
rt...
Carter
> I was able to determine that it IS 120Hz in the audio.
Out of curiosity, how did you determine this?
> I removed the audio tranny from it's home on the top and placed it as far
> away as possible from the power tranny. It's now bolted right behind the
> RCA input jacks. Ground here is on one of the mount posts for the RCA
> jack plate. I electrically restored my previous ground points and removed
> the ground wires. Now the chassis is ground. All power components are on
> the power side of the chassis. It still tunes good and has good audio but
> hum is still present as it was before. No better, no worse.
If you are indeed sure that your hum is 120 Hz, moving transformers
around and working with ground points won't do you a bit of good. The
only cause of 120 Hz hum is poor filtering of a full-wave rectifier.
(That is why I asked above if you are -sure- it is 120 Hz).
> I tried a few different CT points and found that I can make the hum
> worse, but not better. I'm starting to think it might be insufficient
> filtering. Two 47 mF caps seem a little small to me. Most radios use 100
> mF or 60 mF or both?!?!?
Well, they may or may not be too small, but rather could be defective.
From what you have said so far about the hum being 120 Hz, you are on
the right track by pursuing power supply filtering issues.
Good luck!
Ty Unes
> Ty Unes wrote:
>
>> I was able to determine that it IS 120Hz in the audio.
>
> Out of curiosity, how did you determine this?
First off, it seemed reasonable to me in the sense that this was a 1954
model Crosley radio that I was using to test the TX with. AM mind you.
With a typical 5K bandwidth spread, 60Hz coming through that loudly just
didn't seem likely. Also, I have a good ear. I'm a musician of 30+ years
and I felt that if I had an audio reference, I could tell. So, I ran a
small program on my Linux workstation that would produce any tone
frequency I set. I set it for 60Hz and 120Hz and simply compared. There's
no question in my mind, 120Hz was easy to hear.
> If you are indeed sure that your hum is 120 Hz, moving transformers
> around and working with ground points won't do you a bit of good. The
> only cause of 120 Hz hum is poor filtering of a full-wave rectifier.
> (That is why I asked above if you are -sure- it is 120 Hz).
It's nice to "know" this. For a TX (which is new territory for me) I
wouldn't know. I agree with you though. I think I'm gonna try a pair of
100uF 450V caps here and see. What would you suggest?
rt...
Carter
Transmitter or receiver, doesn't make any difference. The key here is if
it is a full-wave rectifier, you get a 120 Hz pulsating DC waveform out
of it. If a half-wave rectifier, a 60 Hz waveform. You then run either
through your filter network to get your "pure" DC.
> I agree with you though. I think I'm gonna try a pair of
> 100uF 450V caps here and see. What would you suggest?
I have not seen the schematic of the power supply, so you are sort of on
your own. Just pay attention to the correct polarity...and use a
Variac(tm) if you have one to bring the voltage up slowly on the new
(used?) capacitors. It probably would be good to remove your existing
capacitors first, as they are probably leaky. Eye protection is always a
good idea when firing up high voltage electrolytics the first time.
Needless to say, make sure the unit is unplugged and the filter
capacitors are discharged before you go poking around.
Pete Bertini
in the transformer can decouple the transmitter's RF
current return path through AC line at the
rectification rate (half or full wave.)
Did you try looking at the modulated
envelope with a good
scope to see if there is ripple
riding on the RF waveform?
That would provide a definitive
answer, and a standard to
judge improvement by...
Peter
dke...@hotmail.com
> So I read through these posts and kinda reworked the TX a bit, took new
> pictures and posted here:
>
> http://2xl.homelinux.com/riverty/gallery/main.php?g2_itemId=1098
>
Hi
Your gounding still doesn't look optimal.
The transofrmer CT lead should be to the first
cap. there should onlt be one lead from there to
the next cap.
I see a third resistor cap as well. Not having a schematic
I don't know if you take off power before that resistor or not.
If you do take power off, that would be a good place to
connect to the chassis ground. The third filter cap should
be wired to this point and not to the chassis. The wire
should be short as well.
Additional capacitor is a good idea.
I am currious as to why you have the capacitor to the
power lead ground? With a transformer system, you
should be able to have the chassis attached to the
real ground.
Dwight