What is causing this audio highs attenuation/ distortion problem?

[K1JJ]

Joe,

Yes, you are right about the 528E not being a fast limiter. I can go "Cluck - tock" with my tongue and see a severe flat top / over modulation from that fast rise time.  I needs a Dominator box... :-)


Well, I decided to finish up this thread with a summary listing of Fabio II features.  Kinda like the new car models:


List of Features in Fabio II, a 4-1000A plate modulated by 4-1000As:

1)   Pair of 4-1000As,  in final class C,  backed off to ½ or 1/4 power for cleanliness.

2)   Pair of tetrode-connected, push-pull  4-1000A modulators in AB1 run conservatively.

3)   Solid state WA1GFZ Mosfet audio driver. (Mod xfmr is the only audio transformer in rig)

4)   DDS VFO, Mosfet RF driver = 80 watts RF drive, all SS.   Driver has 600 watt capability with variable air flow.

5)   Hall effect failsafe shutdown over-current protection for RF final plate, screen and grid.  Also modulator plate, screen, and grid protected with separate sensors.

6)   Neutralized RF final

7)   Vacuum variables and dial counters for plate and loading

   Nine meters for full metering

9)   Viewing windows

10)   Six step sequencer (key-up, key-down) which allows variable delay for each step and also forward and backward order can be changed.   (W2NBC design)

11)   RF deck on pedestal for easy turning 360 degrees for servicing

12)   Aluminun cabinets are easy off in seconds for servicing.

13)   1 KW RCA 144 pound broadscash mod transformer

14)   150 H Heising modulation reactor    -   20 H screen self-modulating choke

15)   All squirrel  cage blowers have new bearings and run quiet on Variacs

16)   Modulator screen voltage electronically regulated. Grid bias uses diodes in cathode with selectable switching.

17)   Unique key-up and key-down arc quenching system using vacuum relays and large 200 watt power resistors across the mod iron and screen choke.   

18)   15KV 30A LARGE vacuum DPDT relay used for RF antenna switching. Good isolation for receiver accidents.

19)   CRL PMC-300A with separate negative and positive peak limiters.

20)   RE-20 mike

21)   Variacs on both screen supplies

22)   Silver plated edge-wound tank coil and 8” heavy duty ceramic bandswitch

23)   Standard  white panels, black lettering, IBM blue cabinets

24)   Modulator on casters for easy servicing

25)   HV supply easily switched / tapped for 2200, 3200 or 3800V.

26)   140 uF  @ 10 KV   total HV supply capacitance. (Choke switched in or out as needed)   Regulation excellent

27)   Modulators run as tetrodes, 590 volts screen voltage. Using 13dB of negative feedback around the modulators back to the low level SS driver.

28) THD (harmonic distortion) better than -35dB 2nd harmonic, -55 dB 3rd harmonic...   IMD (inter-modulation distortion) better than -35 dB 3rd order, -55 dB 5th as measured on HPSDR spectrum analyser.

29) Audio frequency sine wave response flat within 1 dB from 15 Hz to 10 Khz at 100% modulation. 10 Khz to 12 Khz  
80% modulation with clean waveforms. Triangle waveform looks perfect at 200 / 1000 Hz.  

30)   DDS VFO and RF output spurious down at least -80 dB from carrier.

31)   Side channel splatter well behaved and down at least =45dB when out of selected audio bandpass. No secondary side humps as seen on the SDR spec analyzer.

32)  ALL balanced audio cables thru-out the audio chain.  Used to run unbalanced for years.

33)   The babes just love Fabio.

T

[KA2DZT]

Lot of extras in this rig

What's the sticker price with the shipping??

Tax and title are extra.

[W3GMS]

Kudos Tom! 

It looks like its ready to head down runway 29er put the maul down and get it airborne! 

As you know, its a lot of work to get everything right, but feels so good when you have reached that point.   Unfortunately when the going gets tough, lots of folks just quit and throw in the towel on the project.  You worked through the issues and learned a lot in the process, so whats not to like! 

Joe, GMS
 

[K1JJ]

Jeff / W2NBC has been nice enough to spend time with me tailoring my AM audio chain.  He has been a tremendous source of audio information.

This is an on-air recording of Fabio II - Jeff made this afternoon on 75M, using a flat SDR at  + -  12 Khz  bandwidth.   I must say I am very pleased with the sound -  considering my average voice. My transmitted audio intentionally rolls off sharply after 7 KHz in this recording.

Notice the lows, sss's, ch's and ffff's.   Getting cleaner and the hard work is paying off.

T

[KA2DZT]

Tom,

I was listening a few nights back when you were on with Timtron, Tim was comparing different mikes.  Your audio sounded excellent, almost as good as mine.

Fred

[K1JJ]

Tom,

Your audio sounded excellent, almost as good as mine.
Fred

Thanks, Fred.   Yes, it would be a daunting task to challenge your audiophile throne...  


Well, one of the last thangs I need to add here is a reduction in background blower noise. It ain't bad now, but I'd like try to reduce it another 4-5 dB.  I have an older DBX 904  noise gate on its way.  Normally noise gates are unnerving on AM, but I'm thinking of using a slow decay and keep some of the noise always there. (down -4 to 5 dB) This may give the illusion of a quieter background. We will have to experiment and see.

T

[N2DTS]

The ideal thing is to have no blowers.
A silent rig is nice, and not just for the guys on the far end.
I built a few rigs that had no fans or blowers at all, and often use the 3x4d32 rig that has no blowers/fans.
The big rig has blowers but they are small and slow, and do not make much noise.

I use the noise gate in the vx2000, and it seems good if you do not over use it.
Its not a gate, its more of a reverse compressor, it reduces the gain when not speaking, kind of slowly, and does not cut anything off like a gate, just reduces the level some.


I hear guys on the air using amps that have blowers, loud and anoying even for me, it would drive me crazy to have that noise level in the shack.

So how big can you go without fans?
My pair of 813's needs no air, and 4 would do 1200 watts carrier easy.
Four 100TH's as modulators would work, with power to spare, and be quiet.

Can you run 4-1000's without air?
You could run 4-125's and 4-400's without air.

[W2VW]

After 20 years of living with air handlers in the shack I have again remoted mine.

Now the acoustic transformer hum can be heard in its full glory.

[K1JJ]

Yes, 833As X 833As would make a nice, silent rig. Blow a little muffin fan air on the seals and they wud be FB.

I suppose a 4X1 in class C could survive without chimney air, maybe some seal air -  but a pair in AB1 modulator service that idled at a few hundred mils might not do so well.  Eimac does spec a chimney and forced air.

Well, at least my quad 6146 PDM rig is silent and will be get used a lot during the summer.

Another thang to think about is the HEAT generated in the summer. This is another reason for remote control out in the garage.   Of course, a class E rig solves all these problems.  But Rico Suave is long gone to another master.
I remember Rico (24 pill class E rig) running in  Jul / Aug with not much temperature rise in the shack. Amazing.

An expander with slow attack and decay might work here on AM. I'll know in a few days.

Some acoustic transformer hum is part of the scene here too.   We all need to find an old telephone booth. Stick it in the shack and operate inside.

T

[WD5JKO]

I suppose a 4X1 in class C could survive without chimney air, maybe some seal air -  but a pair in AB1 modulator service that idled at a few hundred mils might not do so well.  Eimac does spec a chimney and forced air.

   I recall a high voltage supply using the pulse version of a 4-1000 (4PR1000?) that was immersed in a tank of insulating transformer oil. It was silent, and no fans blowing. Just immerse the rig into an oil tank, and re-position with the front panel up facing the ceiling. This will also solve any arcing problem. 

Jim
Wd5JKO

[KA2DZT]

Tom,

If you put cotton in your ears you probably won't hear the fan noise.  I listened to your signal and did not hear any fan noise.

Fred

[W3GMS]

Yes, 833As X 833As would make a nice, silent rig. Blow a little muffin fan air on the seals and they wud be FB.
An expander with slow attack and decay might work here on AM. I'll know in a few days.
Some acoustic transformer hum is part of the scene here too.   We all need to find an old telephone booth. Stick it in the shack and operate inside.
T

Tom,
Did you ever think about building a wall with a sloped double pane window in the shack.  You would then have the transmitter room and a studio.  You could still see all the visuals of the transmitter through the large window in the wall separating the two rooms.   Set up a second Mic in the transmitter side of the room and mix in just the right amount of ambiance if you like a little bit of transmitter noise!   

Years ago when I was around 15 and my Mom had enough Ham Radio in the house Dad and I built an outside radio building.  I put a wall in that building separating the shack into two rooms.   I could easily see the two 6' racks housing the final and modulators through the large window.  No background noise and it made it very comfortable operating. 

When Dad passed away and Mom went into a retirement establishment I moved that building up to our home.  Looking forward to operating from that space about 125' from the house. 

Joe - W3GMS           

[K1JJ]

Jim, yes, it wud be impractical, but I wonder how glass (4-1000A) would handle being immersed in oil?   Maybe if it slowly heated up and down it would be OK.  It's just when we pour a cool liquid on hot glass does it crack.  Interesting but difficult to do...      What a job to change a tube or troubleshoot the rig.  You probably remember some of the old time hams using metal tubes upside down in water.


Fred, the reason for noise reduction efforts is I'd like to increase the density of my audio using the multi-band processor, but this brings up background noise.  If you listen to the audio.wav I posted a few back, you will hear the white noise in the background.  It would probably come up at least 4-6 dB if I do.    At least it's white noise and not bearing noise. Last week I brought all the blowers in for new bearings.  Made quite a difference.

T

[K1JJ]

Hi Joe,

A few guys have suggested your glass separation idea too.  It would work FB, but I just don't have the room in this shack to do it unless I broke down a wall and added the adjoining room.  My homebrew SSB linear is also a noise problem cuz I like to run a little processing for DXing.

I once looked into sound barrier curtains that cost upwards of $1K as well as acoustic tiles, carpets, etc.  But they would hold in the heat... I'd need outside ventilation, etc. I guess nothing beats the glass window cuz it keeps the rig close by, viewable and dead silent.

For now I may add some carpets on the hard tile floors and be satisfied with running lower density audio. After a certain threshold, blower noise is REALLY distracting on the air and should be quenched.  I find that noise shud be down -30dB or more on ssb from peak signal is important.   On AM, at least -40dB down is acceptable.  For a quick ssb reading, just tune in your own signal and look at the s-meter at full output and then key up with just blower noise without voice modulation for the difference in dB.  


Sounds like a nice set up that you and your dad built.

T

[KB2WIG]

T,

Ya could just put it in the garage, or build a small room. And hook up yer mon ittor. Is Fobio II shy?


klc

[KL7OF]

Tom that rig sounds very nice...I was playing the clip on the computer and Jan says "that guy sounds very good..smooth"...I told her it was FABIO......Nice Job  !
  I remoted the fan on my FRT 39...Put it outside on the shack wall with a weather cover...ducted the air inside...very quiet now...I can keep the exhaust air in the shack in the winter and run it outside for summer time ....Steve

[N2DTS]

Needs the cones of silence!

[W3GMS]

Hi Joe,

A few guys have suggested your glass separation idea too.  It would work FB, but I just don't have the room in this shack to do it unless I broke down a wall and added the adjoining room.  My homebrew SSB linear is also a noise problem cuz I like to run a little processing for DXing.

I once looked into sound barrier curtains that cost upwards of $1K as well as acoustic tiles, carpets, etc.  But they would hold in the heat... I'd need outside ventilation, etc. I guess nothing beats the glass window cuz it keeps the rig close by, viewable and dead silent.

For now I may add some carpets on the hard tile floors and be satisfied with running lower density audio. After a certain threshold, blower noise is REALLY distracting on the air and should be quenched.  I find that noise shud be down -30dB or more on ssb from peak signal is important.   On AM, at least -40dB down is acceptable.  For a quick ssb reading, just tune in your own signal and look at the s-meter at full output and then key up without audio for the difference in dB.  


Sounds like a nice set up that you and your dad built.

T

For some reason I find blower noise much more annoying on SSB than on AM.  For that reason, I always use the downward expander on my Symetrix 528E when I get on SSB.  Nothing is worse than hearing a SSB signal that say is 20 over and when he stops talking the S meter barely moves!  

Agree that to much blower noise on AM is not good either.  I think Bill, DUQ has the balance pretty close to perfection.  Just enough to add a bit of flare, but not so much that you would say, "listen to the blower noise"!      

Yep, sounds like it time to bump out some walls or add an addition on!

Joe, GMS

[N2DTS]

http://www.youtube.com/watch?v=vsNR9FnxOdY

[K1JJ]

For that reason, I always use the downward expander on my Symetrix 528E when I get on SSB.  Nothing is worse than hearing a SSB signal that say is 20 over and when he stops talking the S meter barely moves!  
Joe, GMS

Thanks for the report from the XYL, Steve.  Most of them think we are weirdos, so that's a nice change...


Joe, I also have a 528E here that I use as a "mic conditioner."  I use it as a preamp and a slight touch of compression before it hits the EQ and limiting.  I just tried the 528's expander and as you know, the technique shows promise for AM - HOWEVER, it is abrupt and unnerving. Probably be FB on ssb, but on AM we need a slower attack and decay as well as a variable atten level to work with.



*  WARNING:  This particular unit (card) requires an edge connector and has no case around it. You must also supply the +- 24VDC and +- 15V.    There are complete rack units available for more $.

Here's an older $35 DBX 904  I bought on eBay the other day. Shud be here soon. There are a coupla left if anyone wants one. This has all the knobs to make a seamless transition.  (I think)   It will let us keep some residual noise in there for masking the effect:


http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=161255261744&ssPageName=ADME:L:OC:US:3160

Here's the manual. Looks like 1980s technology, but that's OK.

http://adn.harmanpro.com/product_documents/documents/498_1323992363/904%20Owners%20Manual_original.pdf


The ad says:

"The 904 is an expander gate using the dbx® OverEasy® action for a smooth onset of gating. Attenuation limit, attack and release rates and threshold are all adjustable. This popular gate also features Programmed Latch Mode which mutes a channel until an above threshold signal is present. It then latches open until reset. The 904 is a very fast voltage controlled below threshold downward expander. It senses the level of an input or keying signal and determines whether the level is below threshold. If it is, the signal gets attenuated; if not it passes at unity (0 dB) gain. The amount of signal attenuation is a function of its own level and the 904's Attenuation, Limit, Ratio, and Threshold settings."

Keep it close to the noise source. (mic)   The 528E has  "preamp out / compressor in"   jacks on the back to allow this box to be inserted in an early position.


T

[WD5JKO]

Jim, yes, it wud be impractical, but I wonder how glass (4-1000A) would handle being immersed in oil?   Maybe if it slowly heated up and down it would be OK.  It's just when we pour a cool liquid on hot glass does it crack.  Interesting but difficult to do...      What a job to change a tube or troubleshoot the rig.  You probably remember some of the old time hams using metal tubes upside down in water.

  The big tube is pretty buoyant, so the plate cap was down. This kept the tube in the socket when immersed in oil. Getting the tube in there is a blind operation, and often your arm goes all the way in such that the armpit is at the top edge of the tank and the shirt sleeve is wicking up the oil. Many of these oil tanks had high voltage supplies composed of voltage multiplier stacks. The caps were often about 5000pf @ 40KV, and the size of a hockey puck. If one of the diodes blew, then the cap next in the string remains charged...just waiting for an arm to get nearby. Too often there would be a jolting reaction followed by about 5 gallons of oil spraying around the room.

  We used to charge one of the caps up, and leave it on somebody's desk. They would hold that charge for years...someone sits down at your desk, and starts examining the paper weights. Its amazing how long it takes. Happened to me, and I threw it through a wall in an instant.

  Having a 4-1 in oil should not be a problem as long as the tube is heated up after being immersed.

Jim
Wd5JKO

[W3GMS]

Joe, I also have a 528E here that I use as a "mic conditioner."  I use it as a preamp and a slight touch of compression before it hits the EQ and limiting.  I just tried the 528's expander and as you know, the technique shows promise for AM - HOWEVER, it is abrupt and unnerving. Probably be FB on ssb, but on AM we need a slower attack and decay as well as a variable atten level to work with.
Here's an older $35 DBX 904  I bought on eBay the other day. Shud be here soon. There are a coupla left if anyone wants one. This has all the knobs to make a seamless transition.  (I think)   It will let us keep some residual noise in there for masking the effect:
[/quote]

Tom,
The DBX unit should, based on what I am reading, be a lot more controllable than the very simple downward expander in the 528E.  Yes, the downward expander on the 528E works very well on SSB but I  have never liked the sound of it on AM.  It tends to "flutter" at or near threshold and just creates an annoyance at least to my ears.  All the other functions on the 528E work extremely well and I really like the box.

Looks like the DBX unit you purchased is designed to plug into some sort of power module.  Although any well filtered +/- 24V DC supply should do the job.  I would likely use some +/- 24V 3 terminal linear regulators for that task. 

It will be interesting to see, with the control it provides, how transparent it will be.   

Joe, GMS 

[Opcom]

http://www.youtube.com/watch?v=vsNR9FnxOdY

yeah I know 60+ days, but are those BC-610 plate tank coils in that video, on the top of the domes?


BTW Tom, did you ever have time to make the schematic for this set esp. the modulator with the feedback and tweaks? The MP3 file sounds so good..

I notice you said it is a class C RF stage. Is this a mod applied to the GG amp, Rico II, or a totally new set?
The reason I ask that question is because I was thinking about how a GG stage could be plate modulated. The stability of a GG stage, plus the ability to plate modulate it, would allow it to be used in two modes: As a GG linear amp, or as more of the AM transmitter - biased heavily for Class C, still with cathode drive and plate modulated. Does that make any sense?

[K1JJ]

Hi Pat,

Fabio II is the pair of 4-1000As in class C  plate modulated by a pair of tetrode-connected 4-1000As. The grounded grid amp is an 8877 and another completely different rig. That's Fabio I, thus the confusion... :-)  Rico Suave was the 24 pill class E rig which has been disassembled -  and another ham presently has and will someday resurrect.

Here is the GFZ audio driver I use for the 4-1000A plated modulated rig.  (Look at the third post with the file for the schematic)

http://amfone.net/Amforum/index.php?topic=23632.0

I never did document the final fine-tuned audio feedback circuit except for what is in the schematic. I think every rig will be a little different based on the mod transformer and Heising reactor used, as well as layout. So start with the basic feedback circuit as shown and tweak from there using an audio generator, scope and spec analyzer.

The changes I made also involved stability... by adding some small value bypass caps that tamed the audio feedback down but did not affect the highs too much. I know I had it right when I was able to dial in over 12 dB of audio negative feedback from the 4-1000A plate cap back to the low level 1 volt stage of the SS driver without oscillation. It made a big difference in the rig's ability to pass lows below 100 Hz  cleanly. As you can read in the various threads, I eventually got the transmitter flat and clean, 140% modulated,  from 15 Hz to about 10 KHz.  (less % but still clean from 10-12 KHz) This is not easy when using a mod transformer. The fact that this mod transformer is the only transformer in the complete audio system certainly helps.

At first I tapped the NFB off the secondary winding output of the mod transformer thinking I could correct for some of its phase shift. But in the end, the overall sweep results were better just tapping off one of the 4-1000A modulator tube plate caps and letting the mod iron have its way..

BTW, look for Stu's work here using plate modulated grounded grid amplifiers. I have never tried it but from his tests it seems to be quite do-able.

T

[K1JJ]

After seeing some of the AM rig pictures on this website with BIG tube viewing windows, I realized I've been missing out. For the 4-1000As  modulated by 4-1000As I had been using a tiny 3" window. Today I got out the saw and opened it up.

I can't believe the effect and improvement in viewing ambience.  I plan to get some real glass cut to size with brass bezels to cover the windows.   It's a thrill to watch the modulators change color as I speak. (AB1)  The finals are in a line so I see mostly the front one, but they stay almost gray running class C.


I'd recommend anyone planning a big tube project to make that window as large as they can. It's more work later to open it up because of the aluminum filings, so do it early on.

Fabio is now on air regularly and I finally gots it running close to where it was 3 years ago, but now the highs are even cleaner.  


T

The MODULATORS:   (click on file and should open straight up and down)