ok, teach me some things about broadcash transmitter audio chains.

[W2XR]

Forgot to mention Bruce is right with regard to the inovonics 222.  that's the final limiter box I run and I'd find it very hard to operate without it.  

My own experience with the smiley face eq contour has been unsatisfactory.  A lot of this stuff with audio is subjective and based on individual voice characteristics so the YMMV caveat applies.  There are also hundreds or thousands of different products out there with millions of possible variations so any question can get a flood of "this is what works for me" type answers.  That's fine but it is important to keep in mind that unlike boatanchor receivers say, you'll never be able to try everything in this lifetime, so you have to just focus on setting up something that seems to get the job done for you, then move on to other things.  

Anyway, in my case the problem I ran into with the smiley face and my voice was that I felt like I hit my modulation peaks with low frequencies so most of the transmissions were peaking with muddy lows.  It sounded okay with the right passband and frequency response on the rx, 12 KHz for example with 0 to 6 KHz audio, but if someone listened with an 8 KHz passband, they'd just get the low stuff and weak mids and I'd be much less intelligible.   I decided the goal should be an eq setting that gave me a flat response from around 200 hz up to 4, 5 or 6 KHz (depending on band crowding) and a 10 dB rise from 80 Hz up to 200.    This sounds okay on anything from a 5 KHz passband up to 12 KHz.   My voice except when I first get up in the morning, is mostly midrange and essing  high frequency so this gets me how I sound in real life I think.  

Rob

Hi Rob,

This will probably appear sacreligious to some, but I like the Inovonics 222 better than the Orban Optimod-AM I have.

I wish I had saved my dough and never bought the Optimod. As I have stated in this forum on several previous occasions, the Optimod may work fantastically with class-E and other later generation technology transmitters, but in my experience it is not a good processor when used with a plate modulated rig.

The Inovonics 222 sounds great and really grabs ahold of the audio, and sounds every bit as good on-air for voice as the Optimod. Music quality through the Optimod is a different story entirely, due to the wider bandwidth requirements associated with music, and the multi-band processing advantages (reduced IMD among other things) are readily apparent here. I can let the rig modulate up to and beyond 125% in the positive direction with the 222, and it will never exceed 95% in the negative direction. As you know, even if you were to purchase it brand-new, the price is still quite reasonable, and they are somewhat common on the used market. It has been in continous production since the late 1980s, so factory support is not an issue.

Rob, do you use the internal NRSC boost in your 222? I don't, perferring instead to use an external EQ ahead of the processor for my HF boost and overall EQ.

I have several different tube and solid-state processors here that I have acquired over the years, but I always find myself coming back to the 222, and it far and away gets the most airtime at W2XR (what little airtime there is). Like yourself, I would find it difficult living without it.

73,

Bruce

[K1JJ]

Bruce,

I boost the 50-90hz range about 3db less than I boost the 3-5kc area.  So, the S curve starts at say, 0db at 50-90hz, drops off down to -8 to -10db to bottom between the muddy 110-800hz area, then slowly rises linearly up to 3-5kc where it flattens out to 5kc at +3db boost. Then cuts sharply down from there.  Yes, your absolute settings will be different, of course, based on your voice.

As you well know, first the rig must be CLEAN - transparent.  Otherwise all this talk is a waste. A stock Valiant, etc is not what I call clean and boosting lows and highs is simply gonna produce mud and trash. Best to run a rig like that flat. It's the flat and super clean rigs from 30hz-8kc I'm talking about here.

I look at it this way. When we get on the air, how many unsolicited compliments do we get on our audio? Guys like 3rd BA Bob/K1KBW are always getting them. He has a tremendous voice for radio and he set it up right with the proper gear, including an EQ. Then there's guys like QIX who have AVERAGE voices, but have their rigs set up perfectly for their voices and get comps all the time - as much as Bob.

Sure we can run a D-104 into a stock Apache and talk to the world and get out fine, if that's all we want. But the comps will never come unless it's out of pity. I think most all hams wanna sound like a $million bux, am I wrong? So MUCH depends upon our voice. The worse it is, the more tailoring required to hit a certain pleasing level to the ear.  We have a choice to be satisfied with a simple arrangement that sounds "OK" or a more complex and well tailored sound that rivals the broadcash stations. There's nothing wrong with either - it's just that a distinction must be made here. What one wants to do and his goals are what matter.

T

[W2XR]

Bruce,

I boost the 50-90hz range about 3db less than I boost the 3-5kc area.  So, the S curve starts at say, 0db at 50-90hz, drops off down to -8 to -10db to bottom between the muddy 110-800hz area, then slowly rises linearly up to 3-5kc where it flattens out to 5kc at +3db boost. Then cuts sharply down from there.  Yes, your absolute settings will be different, of course, based on your voice.

As you well know, first the rig must be CLEAN - transparent.  Otherwise all this talk is a waste. A stock Valiant, etc is not what I call clean and boosting lows and highs is simply gonna produce mud and trash. Best to run a rig like that flat. It's the flat and super clean rigs from 30hz-8kc I'm talking about here.

I look at it this way. When we get on the air, how many unsolicited compliments do we get on our audio? Guys like 3rd BA Bob/K1KBW are always getting them. He has a tremendous voice for radio and he set it up right with the proper gear, including an EQ. Then there's guys like QIX who have AVERAGE voices, but have their rigs set up perfectly for their voices and get comps all the time - as much as Bob.

Sure we can run a D-104 into a stock Apache and talk to the world and get out fine, if that's all we want. But the comps will never come unless it's out of pity. I think most all hams wanna sound like a $million bux, am I wrong? So MUCH depends upon our voice. The worse it is, the more tailoring required to hit a certain pleasing level to the ear.  We have a choice to be satisfied with a simple arrangement that sounds "OK" or a more complex and well tailored sound that rivals the broadcash stations. There's nothing wrong with either - it's just that a distinction must be made here. What one wants to do and his goals are what matter.

T

Hi Tom,

Thanks for the usual thoughtful reply!

Completely agreed. When it comes to audio, EQ will not make a silk purse out of a sows ear.

I boost the 3 to 5 khz region about 6 dB in my rig, so I'll try about 3 dB boost in the 50 to 70 hz region as a starting point, and hear how she sounds. For what it is worth, I have never changed the EQ settings since I added 6 dB of negative feedback (NFB) to the modulator/driver section about 3 years ago.

I think that to really appreciate the advantages of EQ, NFB in the transmitter makes a big difference. I had to completely re-EQ the rig after I added the NFB. Much less HF boost was required, and the HF EQ contour changed noticably after the feedback was included. As to be expected, the low-end response was improved with the NFB, too. It's always easier to EQ a black box that has an essentially flat audio response to begin with.

Just my 2 cents.

73,

Bruce

[N2DTS]

kbw is one on the guys who are often 25 KHz wide.
I hear plenty of people who sound just as good with half (or less) the bandwidth.

When he is on, I often turn the stuff off as he takes up the entire ghetto.
I have no idea what he normally runs, but I suspect there is no limit to the high end, and its also boosted...

Brett

[K5UJ]

Bruce,

Thanks, yes the 222 is a great box but it is important that I qualify what I write by stating that I run my audio gear into a low level solid state JA exciter rig via an isolation transformer.  I feed the balanced modulator chip directly via an RCA photo jack at the back of the rig.  I don't have great pipes.  In fact my voice is or can be nasal and irritating.  There are some hams with these rich baritone voices, we all know a few, who could kill just running a mic into a stock Viking 1.    I keep trying to get by with as few boxes as possible, eliminating devices when I can figure out a way to do without them but I'm still down to probably too many.  Orban wrote once that fewer is better.  I run the mic preamp into a graphic eq that doubles as a bandwidth control.  I pretty much cut everything off below 80 Hz but with my voice there isn't much there anyway unless I have a head cold.    The equalizer box also has a great noise gate.  This drives an old Orban 422A compressor/limiter/desser which feeds the 222.    The Orban limiter is a studio limiter with a delay.  By itself it is not adequate for preventing negative overmodulation in AM.  The deesser is an Orban proprietary design that I didn't think did any good but I swept it one day and found it kicked in at 4 KHz so I started running it

I operate the 222 with the NRSC pre-emphasis, however mine is the 222-03 (or maybe it is the 222-02) that has the low pass filter that cuts off at 6.4 KHz.  That's my stopgap.   No matter what I do, nothing gets above 6.4 KHz but usually I never put out anything higher than 5 KHz.  I run the 422A at a pretty high compression ratio to get some density.  that with the 222 often performing a 6 to 12 dB gain reduction at 130% positive lets me (I hope) achieve a high average audio level.   
Interesting what you wrote about the 222 v. the Optimod.  I jokingly refer the 222/422A combo as the poor man's Optimod, although the 422A is not a multi-band compressor.

I also have a CRL PMC400A that I got about 10 days ago at a hamfest.  Have not had a chance to do anything with it (it needs a little TLC before I power it up).  It has tilt control for high level plate rigs.  How important is this?  I'm in the dark mostly beyond vaguely thinking this has something to do with phase.

I have to train myself to simply speak normally across the mic and let the audio gear do the heavy lifting.  After years I am very much in the habit of almost yelling into the mic and calming down to speak in a normal conversational tone is really hard for me to learn.   I once went on a tour of the studio of one of our 1-A 50 kw stations here in town.  You can imagine all the processing, the optimods, the RE20s and so on.  Listening on the broadcast receiver the talent sounded as if they were trying to fill a room.   At the station standing next to them while they delivered their copy on the air I discovered they were working up close to the mic and speaking in low modulated tones.   You get these rich nice voice sounds, not by speaking like some carnival barker but by working the mic up close and talking low.  Well, for me. old operating habits die hard    One amusing thing about the broadcash station--I observed an old 9100 in an equipment rack and expressed surprise to see it there instead of out at the tx site.   The CE happened to be nearby and he explained it is used to process headphone audio back to the talent because they want an "on-air" sound in their monitors.  I teased him about this a bit for using a $4 or $5K box for that purpose but when I got home the joke was on me--the monitor to my cans from my mixer board is taken off the 222 just ahead of the isolation transformer!  So I'm a hypocrite.   

Tom,

I second Bruce -- ur comments always well thought out.  I swept my rig last spring and again last month with everything bypassed or in proof mode and it is flat from 20 hz up to somewhere above 10KHz (I quit paying close attention when it got way up there).   Indeed, the rig shouldn't be the bandwidth limiting bottleneck.  Problems come up when hams try to force audio past a transmitter.   And I must say, I enjoy the rich wide response sounds of a station strapping in here.  My low frequency comments were geared towards firstly, myself, since for me, any lows would be artificial (and sound pumped up and fake) and secondly, for when condx are lousy and a more penetrating "dx" type sound is needed.   

73
Rob

[W2XR]

Bruce,

Thanks, yes the 222 is a great box but it is important that I qualify what I write by stating that I run my audio gear into a low level solid state JA exciter rig via an isolation transformer.  I feed the balanced modulator chip directly via an RCA photo jack at the back of the rig.  I don't have great pipes.  In fact my voice is or can be nasal and irritating.  There are some hams with these rich baritone voices, we all know a few, who could kill just running a mic into a stock Viking 1.    I keep trying to get by with as few boxes as possible, eliminating devices when I can figure out a way to do without them but I'm still down to probably too many.  Orban wrote once that fewer is better.  I run the mic preamp into a graphic eq that doubles as a bandwidth control.  I pretty much cut everything off below 80 Hz but with my voice there isn't much there anyway unless I have a head cold.    The equalizer box also has a great noise gate.  This drives an old Orban 422A compressor/limiter/desser which feeds the 222.    The Orban limiter is a studio limiter with a delay.  By itself it is not adequate for preventing negative overmodulation in AM.  The deesser is an Orban proprietary design that I didn't think did any good but I swept it one day and found it kicked in at 4 KHz so I started running it

I operate the 222 with the NRSC pre-emphasis, however mine is the 222-03 (or maybe it is the 222-02) that has the low pass filter that cuts off at 6.4 KHz.  That's my stopgap.   No matter what I do, nothing gets above 6.4 KHz but usually I never put out anything higher than 5 KHz.  I run the 422A at a pretty high compression ratio to get some density.  that with the 222 often performing a 6 to 12 dB gain reduction at 130% positive lets me (I hope) achieve a high average audio level.  
Interesting what you wrote about the 222 v. the Optimod.  I jokingly refer the 222/422A combo as the poor man's Optimod, although the 422A is not a multi-band compressor.

I also have a CRL PMC400A that I got about 10 days ago at a hamfest.  Have not had a chance to do anything with it (it needs a little TLC before I power it up).  It has tilt control for high level plate rigs.  How important is this?  I'm in the dark mostly beyond vaguely thinking this has something to do with phase.

I have to train myself to simply speak normally across the mic and let the audio gear do the heavy lifting.  After years I am very much in the habit of almost yelling into the mic and calming down to speak in a normal conversational tone is really hard for me to learn.   I once went on a tour of the studio of one of our 1-A 50 kw stations here in town.  You can imagine all the processing, the optimods, the RE20s and so on.  Listening on the broadcast receiver the talent sounded as if they were trying to fill a room.   At the station standing next to them while they delivered their copy on the air I discovered they were working up close to the mic and speaking in low modulated tones.   You get these rich nice voice sounds, not by speaking like some carnival barker but by working the mic up close and talking low.  Well, for me. old operating habits die hard    One amusing thing about the broadcash station--I observed an old 9100 in an equipment rack and expressed surprise to see it there instead of out at the tx site.   The CE happened to be nearby and he explained it is used to process headphone audio back to the talent because they want an "on-air" sound in their monitors.  I teased him about this a bit for using a $4 or $5K box for that purpose but when I got home the joke was on me--the monitor to my cans from my mixer board is taken off the 222 just ahead of the isolation transformer!  So I'm a hypocrite.  

Tom,

I second Bruce -- ur comments always well thought out.  I swept my rig last spring and again last month with everything bypassed or in proof mode and it is flat from 20 hz up to somewhere above 10KHz (I quit paying close attention when it got way up there).   Indeed, the rig shouldn't be the bandwidth limiting bottleneck.  Problems come up when hams try to force audio past a transmitter.   And I must say, I enjoy the rich wide response sounds of a station strapping in here.  My low frequency comments were geared towards firstly, myself, since for me, any lows would be artificial (and sound pumped up and fake) and secondly, for when condx are lousy and a more penetrating "dx" type sound is needed.  

73
Rob

Hi Rob,

Thanks for the nice, detailed reply!

I think the Inovonics 222 that you have may be the unit designed for the international shortwave broadcast service, based upon the fact that it brickwalls out at 6.4 Khz when the low-pass filter is inserted. My 222 is the NRSC-compliant version, with the 9.5 Khz low-pass filter.

I do not use the internal low-pass filter. Inovonics claims the filter network is compensated to prevent ringing, but I still do notice a slight bit of ringing when the filter is switched on and in the signal path. In fact, I can drive the audio about 1 dB harder to my rig with the filter switched out, and gain about 1 dB of loudness. I know it's only 1 dB, but I guess I have read too much of what Bob Orban has espoused about maximizing modulation density and loudness in AM broadcasting. I use my outboard EQ to gently roll-off the HF response.

Rob, what is a JA exciter? I guess it is some kind of SSB rig, based upon your comment of feeding the processed audio to the balanced modulator section of the rig.

You mentioned that your recently acquired CRL PMC400A has a tilt adjustment. My Optimod-AM 9100A also has adjustable compensation for transmitter tilt and low-frequency roll-off, and I have set these adjustments per the installation manual for the unit. They do help, no question, but I still cannot achieve the kind of fail-safe control of the audio that the Inovonics 222 provides. In the manual for this unit, Orban goes into the technical issues associated with sophisticated multi-band processing and soft clipping and their effect on what he refers to as "older, pre-1965 AM transmitters" (read plate modulated), and that these transmitters are best retired from service and replaced with rigs better suited to the demands imposed by rigorous audio processing, etc. Based upon my experience with the Optimod, I must agree. There are just too many issues related to transmitter power supply ringing and droop, ringing in the audio magnetics (mod xfmr, driver xfmr, mod reactor, and so forth), that preclude such aggressive processing schemes from ultimately being completely successful with plate modulated transmitters.

If your rig happens to be of the plate modulated variety, please let me know how you make out with the CRL processor, and the effect the tilt adjustment has on it's modulation performance. I'd be very curious!

Best 73,

Bruce

[K5UJ]

HI Bruce, thanks -- a quick reply to a few things -- I'll have to play around with the low pass filter bypass and see what happens.

JA exciter -- sorry for the ambiguity -- I was referring to my Yaesu Mk V FT1000MP, the 200 w. ssb rig Yaesu made and sold before they came out with their super expensive rigs and the FT2000 series.   I run it as an analog low level AM exciter at about 20 w. to drive 2 x 3-500zg.  So the tilt control in the CRL PMC will be wasted on me at this point.

Yes my 222 is intended for European SW BC service.  They have another one with a 4.5 or 5 KHz LPF that is for American SW BC, then there is the medium wave 10 KHz NRSC like what you have and I think there is one other, I want to say it is a 7.5 KHz LPF but don't hold me to that.  This is all somewhere on the Inovonics website.  BTW, I was told that it is not hard to turn a 10 KHz 222 into a narrower one and Inovonics will furnish customers with the parts and instructions, but I have not contacted them to verify that. 

That's very interesting that the Optimod didn't control your peaks as well as the 222.   I would have never expected that.  I figured for the money those 9000 and 9100 go for they'd be the ultimate. 

These days any AM bc station that can afford it, has gone solid state so I guess they can all max out their processing and some of them sound like it    It doesn't bother me but most tone it down over concerns with listener fatigue.  I prefer to hear some punch myself.

73

Rob 

[Steve - WB3HUZ]

I also have a CRL PMC400A that I got about 10 days ago at a hamfest.  Have not had a chance to do anything with it (it needs a little TLC before I power it up).  It has tilt control for high level plate rigs.  How important is this?  I'm in the dark mostly beyond vaguely thinking this has something to do with phase.

I have the PMC 300. I think the 400 has similar controls. The 300 has two HF boost positions. I forget the exact amount but one provides about 2x the boost as the other position. You can set the frequency at which the boost begins on the back. I set mine to 1 kHz. It's a linearly increasing slope to the boost from that point.

The overshoot and tilt controls are separate from the HF boost control. One supposedly reduces the overshoot found in plate modulated rigs. The tilt compensates for low frequency phase shift and frequency response limitations of plate modulated rigs. These controls should not be needed on your FT1000.

The PMC should also have a lowpass filter in it. The default setting is 10 kHz or maybe the NRSC curve on the newer boxes. You can change the corner frequency by changing a resistor/cap or two.

[w3jn]

Rob, in my experience the tilt control on the PMC didn't do much, if anything, on either my GPT-750 with the broadcash iron, or the T-368 (stock RF section, driving the grids of the modulators directly from a solid state amp).

The PMC coupled with the Spectral Energy Processor (SEP) will make ya LOUD if ya want.  With the T-3 I find it best to run the NRSC boost off, but if you have a rig with good high response you might try it.

[flintstone mop]

Yeah OM
The key is to try to get an A.M. limiter and some EQ. A mediocore mic would be a good start.
The A.M. limiter is more than a berhinger or consumer limiter. It is tailored for A.M. broadcast and will limit excessive high freqs that you might be inadvertantly pumping into the TX with the EQ. This will prevent splatter and excessive sssssss or shhhhhh over the air.

G'day
Phred

[W1DAN]

Guys:

I knew this thread would explode!

After spending years being a processing nut, I have two ideas to put forth that may or may not add to this thread:

1) a D104 and  negative limiter/clipper/LPF can work wonders. Cheap and effective. Do not underestimate it.

2) If you wanna play with boxes, all the processing mentioned is viable. For tube rigs I have found that the audio power bandwidth (the ability for the modulator to modulate at full power across the audio spectrum) can easily be limiting. This is why the 9100A that W2XR mentioned does not work as well as he would have thought. A number of years ago I purchased an Orban 9105A-this was a processor designed for international shortwave service, and sacrificed quality for added loudness..kinda like a D104 on steroids. It made almost square waves of the audio. Well, I connected this to a Viking 2 via a SS amp and a backwards Hi Fi output transformer feeding the grids of the 807 modulators. I had about 20db of negative feedback around the SS/modulator and had a frequency response of about 50hz-8kc. The 9105A made the Viking 2 fall to it's knees modulation wise! I could not make 100% modulation due to the spectral density and loudness. This was primarily due to modulation transformer core saturation.

So I built a PDM Class E rig with it's flat power bandwidth and all was happy. Sold the 9105A later. Miss it as it really cut through the QRM.

I bought a Inovonics 222 at the last NearFest (ahem $5.00! The best buy ever at that flea) and am very impressed with it. I also have the resistor chart that Inovonics put out for changing the LPF of the unit. This is not an "easy" job, but can be done. PM me and I can email it to you.

With my current airchain and with the 222 at the end, I have achieved almost the same loudness as the 9105A, but am starting to back off the knobs when condx are good..this creates a "loud enough but clean" signal that is listenable.

I use a Gregg Labs 2540 ahead of the 222 as a multiband processor like the 9100A. This EQ's my voice, but I also have to drop about 250hz by about 3-6db to have my mid bass not overpower the rest of the spectrum. For me bass is a hinderance to intelligibiliy in most conditions except strapping.

Just my thought...

73,
Dan
W1DAN

[K5UJ]

Steve, JN thanks for the info.  Dan, I recall when you got the 222 for a song.  I came close weekend before last when I got a CRL PMC400A for $25 but mine needs work.  Very interesting your experiences with the 222 and Gregg Labs and 9105.   I had been wondering how the old Johnsons etc. would handle more modern processing.  2 or 3 years ago I corresponded briefly with Bob Orban and he recommended the shortwave BC optimod for ham AM.  My 222 cuts off at 6.4 KHz which is okay for me but perhaps Bruce would be interested in the resistor mod.   

73

Rob

[W2XR]

Guys:

I knew this thread would explode!

After spending years being a processing nut, I have two ideas to put forth that may or may not add to this thread:

1) a D104 and  negative limiter/clipper/LPF can work wonders. Cheap and effective. Do not underestimate it.

2) If you wanna play with boxes, all the processing mentioned is viable. For tube rigs I have found that the audio power bandwidth (the ability for the modulator to modulate at full power across the audio spectrum) can easily be limiting. This is why the 9100A that W2XR mentioned does not work as well as he would have thought. A number of years ago I purchased an Orban 9105A-this was a processor designed for international shortwave service, and sacrificed quality for added loudness..kinda like a D104 on steroids. It made almost square waves of the audio. Well, I connected this to a Viking 2 via a SS amp and a backwards Hi Fi output transformer feeding the grids of the 807 modulators. I had about 20db of negative feedback around the SS/modulator and had a frequency response of about 50hz-8kc. The 9105A made the Viking 2 fall to it's knees modulation wise! I could not make 100% modulation due to the spectral density and loudness. This was primarily due to modulation transformer core saturation.

So I built a PDM Class E rig with it's flat power bandwidth and all was happy. Sold the 9105A later. Miss it as it really cut through the QRM.

I bought a Inovonics 222 at the last NearFest (ahem $5.00! The best buy ever at that flea) and am very impressed with it. I also have the resistor chart that Inovonics put out for changing the LPF of the unit. This is not an "easy" job, but can be done. PM me and I can email it to you.

With my current airchain and with the 222 at the end, I have achieved almost the same loudness as the 9105A, but am starting to back off the knobs when condx are good..this creates a "loud enough but clean" signal that is listenable.

I use a Gregg Labs 2540 ahead of the 222 as a multiband processor like the 9100A. This EQ's my voice, but I also have to drop about 250hz by about 3-6db to have my mid bass not overpower the rest of the spectrum. For me bass is a hinderance to intelligibiliy in most conditions except strapping.

Just my thought...

73,
Dan
W1DAN

Hi Dan,

Good to here from you, OM! Hope all is well at your end.

I am not so sure that the audio power bandwidth of the modulator in my rig is the issue insofar as the transmitter's inability to make full use of the Optimod-AM 9100A processing capabilities.

The modulator is a pair of class B 833As and I can turn the plate supply variac up to 3100 VDC. The modulation xfmr and reactor are out of a 1 KW AM BC rig, so I don't think the ability of the modulator to deliver the power Bw is the issue, relative to the DC input level to the final that I normally run. If this was the case, the audio would be grossly distorted when I use the Optimod, and it is not; it is actually very clean. Perhaps the mod xfmr in your rig was saturating when used with the Optimod; I have no idea.

I believe the explanation is precisely what Bob Orban suggests in the operating and installation manual for all of his Optimod-AM units. The real issue is the clipping feature that is inherent in his processing scheme, and although the clipping is adjustable by the user, unfortunately it is not defeatable. The combination of the clipping and how hard the multi-band compressor is driven determine in large part the density of the modulation.

I think you came fairly close to hitting the nail on the head in terms of the modulator making square pulses of the audio when you used the Optimod 9105 with your rig. To a certain extent, that is what the Optimod does; the clipping action within the processor, although it is sonically relatively benign, does tend to square off the audio peaks by it's very nature, and any transmitter with complex impedances in the audio signal path (i.e. mod xfmr, mod reactor, audio driver xfmr, DC blocking cap, etc.) , such as in a plate modulated rig, will tend to exhibit excessive ringing and overshoot in an highly unpredictable manner when excited by a clipped audio waveform.

The other factor has to do with the power supply in a plate modulated transmitter. The filter capacitance and filter choke form a resonant circuit, and this is also susceptible to ringing, particularly when excited by an audio waveform that attempts to approximate a square pulse.

I can use the Optimod with my transmitter, and it sounds great; make no mistake about that. As I had mentioned previously, the low-frequency Eq and tilt adjustments do make a noticable improvement in the ability of the rig to make good use of the Optimod, but the improvement is not completely acceptable whereas it really should be. These adjustments are to compensate for the overshoot and extreme LF roll-off inherent in the transmitter. The problem with the Optimod and my transmitter is that it is difficult to have very tight control of the negative-going peaks while attempting to maintain a reasonably high level of modulation density. As I had mentioned in a previous post, I can obtain essentially absolute control over the negative-going peaks with the Inovonics 222, so that is why this unit remains my processor of choice. And for voice-only applications, the 222 sounds every bit as good as the multi-band Optimod, with none of the issues I had described above to contend with.

The PDM, class-E, and other later technology transmitters do not suffer from these issues to anywhere near the same degree as a plate modulated rig, as you had verified once you built up your own PDM rig. That is why I am not getting rid of the Optimod; I very probably will one day build up a class-E rig, and then I can take full advantage of the benefits provided by the Optimod.

I suggest anyone wishing to review this particular subject in greater detail download the operating manual for any of the Orban Optimod-Am processors. Orban goes into this very subject of "pre-1965 transmitters" (read plate modulated rigs) and the problems they present when using the kind of aggressive processing their processors can provide.

73,

Bruce

[W2XR]

Steve, JN thanks for the info.  Dan, I recall when you got the 222 for a song.  I came close weekend before last when I got a CRL PMC400A for $25 but mine needs work.  Very interesting your experiences with the 222 and Gregg Labs and 9105.   I had been wondering how the old Johnsons etc. would handle more modern processing.  2 or 3 years ago I corresponded briefly with Bob Orban and he recommended the shortwave BC optimod for ham AM.  My 222 cuts off at 6.4 KHz which is okay for me but perhaps Bruce would be interested in the resistor mod.   

73

Rob

Hi Rob,

I'm in good shape with my 222. It is the US/NRSC model, and the LPF cuts off at the NRSC upper limit of 9.5 khz. And as indicated earlier, I don't use the internal LPF.

73,

Bruce

[N2DTS]

Here is my favorite guy, taking out about 25 KHz with his fantastic audio.
I cant figure out how to do a screen capture with the other programs which show more...


Brett

[Steve - WB3HUZ]

Why don't you just confront Bob on the air, send him an email, or go knock on his door? Your continued pissing and moaning about him here is growing old.

Here is my favorite guy, taking out about 25 KHz with his fantastic audio.
I cant figure out how to do a screen capture with the other programs which show more...


Brett

[N2DTS]

He said he does not care.
I only brought it up as he was one mentioned here who has great audio.
It DOES sound good.

Tim (hlr) told me he does not limit his highs, I am not sure most people do, but he never looks wide.
He was telling me to open my audio up to 10KC and people said I was not wide, so something else is going on...

Brett

[Steve - WB3HUZ]

Running audio out to 10 kHz is a waste. Most receivers cannot be opened up wide enough to even hear it. The few that can may only be opened up so wide on the rare occasion of no QRM in that passband. When will people learn this?

[K5UJ]

I can handle 10 Khz of audio on rx but it only sounds okay in the daytime on 75 meters with short daytime propagation when the band is not very crowded.   Even then for voice it is pointless because much above 4 or 5 KHz you just have consonant essing.  You can still have a full audio sound on voice and taper off the high end somewhere between 4 and 5 KHz.  Not much is needed above 5 in my opinion.  It looks to me like he's simply running the full NRSC mask for medium wave AM here in North America.  He may just have broadcast audio gear that he runs as is, and does not realize that for ham voice operating he can use some kind of low pass filtering and be fine with half the bandwidth.   Dan may know more about this but one thing I recall about the Orban 9105 is that it has a rolloff at 5 Khz because it is for shortwave broadcasting.   I think that is correct.  That's probably one reason why Bob Orban recommended it for ham AM.  I think I told him in a nice way that not many hams can afford Orban processing, even used.

[W1DAN]

Hi Bruce (and others):

I am doing well, just very busy with work and school these days. Thanks for asking.

Good points on what you are experiencing with your Optimod. My Viking 2 definitely was saturating on the low end, and also had fairly quick LF drop-off below about 50hz. Also, I had two transformers in my modulator section, so phase changes happened quickly at the extremes of the frequency response. For my experience, the RMS power that the Orban 9105A was asking of my modulator was too much. The saturation and frequency/phase changes on the low end caused my 100% negative modulation point to be hard to determine. With a more linear modulator this point is better defined. When I had the 9100A, this symptom was less pronounced than with the 9105A. BTW, my power supply caps in the Viking 2 were new and increased in capacity. The lights dimmed quickly when I hit the plate switch as the caps charged.

On your modulator LF limitation...what is your -3db point in your transmitter (BTW..nice modulator!)? How many audio transformers? Can you eliminate any driver transformer? With the small amount of square waves generated by the 9100A, these square waves would tilt in a plate modulator and thus the peak level of the square waves would not be at the level you want (usually higher in negative level). This is why Orban wants every good boy to have a transmitter LF frequency response down to the hertz region (I have studied his manuals and patents extensively-pretty interesting circuitry tricks). The proof of a good modulator is in attempting to pass an audio square wave through your modulator and see how faithful the waveform is. As we all know this is a difficult test to pass. In my Viking 2, I bet it would be pretty ugly. In my Class-E PDM rig, I know I have some HF ringing due to the high level PDM filter. In the case where there are bandwidth pix posted above, that transmitter is running the linear class H modulator and the frequency response has no natural filter anywhere in the chain. BTW, I brick wall my audio at about 6kc with a 24db/octave LPF similar to the Orban and Inovonics. I am thinking of bringing it down to 5kc. The 9105A has a brick wall at 4.5kc and is down about 40db at 5kc. It is my opinion that having 10kc audio through a transmitter is not recoverable on an receiver under 99% of the conditions. In an rx that has a 12kc IF, that means it can hear only 6kc of recovered audio.

One thought would be to use high level diodes to limit the negative excursions after the mod transformer. Your 100% modulation point would be well defined at the expense of a little added splatter depending on any splatter filter used. Look at Steve QIX's 3 diode limiter circuit.

As you may know, on transmitter power supplies...the resonance is usually in the hertz or sub-hertz region. This shows up as a change in carrier level when modulating and is called "carrier shift" due to "power supply bounce". Most reasonably built transmitters have little supply resonance. I know you knew that though.

In the 9100A, try backing off your multiband clipping and run that into the 222 (maybe with the pre-emphasis turned off). The 222's overshoot compensator changes the phase of the signal before it is clipped to reduce the amount of square waves created...cool trick!

On the 222..I too am very impressed with it. It smooths out my high end better than many other limiters.

For fun and weirdness, I obtained an Orban TV 8182A (and some 8182ASAP's-different story for another time). I have this after my Gregg 2540 and before the 222. The 8182A has something called a Hilbert Clipper, which is just a simplified speech processor that creates IM distortion instead of clipping distortion. This adds a few DB of apparent loudness to a voice signal. This is what the 9105A also did, but better. I kinda like my chain better as it is warmer.
 
Hope we did not hijack this thread too much. Again a D104 can be extremely respectable-simple too!

Derb-Man...just get on the air with your Master of Metal, don't worry so much!

Good to hear your thoughts Bruce (and others)!

73,
Dan
W1DAN

[KD6VXI]

Here is my favorite guy, taking out about 25 KHz with his fantastic audio.
I cant figure out how to do a screen capture with the other programs which show more...


Brett

That's LOTS of multiband compression.  I can effect the same thing with my audio chain if I add in the BBE Sonic Maximizer.  I add it in before the brick wall limiter.  It basically 'extends' your highs digitally.. (adding in things that might should be there, but usually aren't).

I also have a subsonic synth I can use, but it usually ends up sounding muddy.

BUT, to get nice wide audio like that, that ALSO sounds good, you need multi-band compression.  A third octave EQ and other doodads would help, but I'd say it's a kick ass compressor, and maybe some other type of "brilliance" adder.

It's similiar to the "torque under the curve" in a racing engine.  Making 2000 ft lbs of torque means NOTHING if you only do it in a very narrow peak.  Making 2000 ft lbs across a large RPM band means you really be screamin down the track...  Same with audio.  You can extend your bandwidth WAY out there, but if you don't use something to fill the valleys and voids, it's going to not sound as good, be VERY sibilant, and have people pissin and moanin.


--Shane

[N2DTS]

Oh!
Thats the missing piece.
I often wondered why broadcast stations look like they have lots of audio on the carrier, huge lobes on the spectrum display, and most ham AM signals have very little, even with plenty of audio and compression.

Is it correct to assume you could chop it off say at 5 Khz and still have plenty of audio, and sound good, and not be real wide?

Brett

[Steve - WB3HUZ]

Is it correct to assume you could chop it off say at 5 Khz and still have plenty of audio, and sound good, and not be real wide?

Yes.

[flintstone mop]

Heyyyyy DERB
I hope UR taking a lot of notes!!!!
Start simple and don't worry about the racks of audio equipment needed for Ham radio. As you feel comfy with a B'cast TX, then you graduate to the processors.
The Raytheon only needs 0dB of audio. Consumer audio levels, but it should/must be balanced audio. Maybe suggestions here for a type of transformer to get Tim Balanced audio out.

Fred

[W2XR]

Hi Bruce (and others):

I am doing well, just very busy with work and school these days. Thanks for asking.

Good points on what you are experiencing with your Optimod. My Viking 2 definitely was saturating on the low end, and also had fairly quick LF drop-off below about 50hz. Also, I had two transformers in my modulator section, so phase changes happened quickly at the extremes of the frequency response. For my experience, the RMS power that the Orban 9105A was asking of my modulator was too much. The saturation and frequency/phase changes on the low end caused my 100% negative modulation point to be hard to determine. With a more linear modulator this point is better defined. When I had the 9100A, this symptom was less pronounced than with the 9105A. BTW, my power supply caps in the Viking 2 were new and increased in capacity. The lights dimmed quickly when I hit the plate switch as the caps charged.

On your modulator LF limitation...what is your -3db point in your transmitter (BTW..nice modulator!)? How many audio transformers? Can you eliminate any driver transformer? With the small amount of square waves generated by the 9100A, these square waves would tilt in a plate modulator and thus the peak level of the square waves would not be at the level you want (usually higher in negative level). This is why Orban wants every good boy to have a transmitter LF frequency response down to the hertz region (I have studied his manuals and patents extensively-pretty interesting circuitry tricks). The proof of a good modulator is in attempting to pass an audio square wave through your modulator and see how faithful the waveform is. As we all know this is a difficult test to pass. In my Viking 2, I bet it would be pretty ugly. In my Class-E PDM rig, I know I have some HF ringing due to the high level PDM filter. In the case where there are bandwidth pix posted above, that transmitter is running the linear class H modulator and the frequency response has no natural filter anywhere in the chain. BTW, I brick wall my audio at about 6kc with a 24db/octave LPF similar to the Orban and Inovonics. I am thinking of bringing it down to 5kc. The 9105A has a brick wall at 4.5kc and is down about 40db at 5kc. It is my opinion that having 10kc audio through a transmitter is not recoverable on an receiver under 99% of the conditions. In an rx that has a 12kc IF, that means it can hear only 6kc of recovered audio.

One thought would be to use high level diodes to limit the negative excursions after the mod transformer. Your 100% modulation point would be well defined at the expense of a little added splatter depending on any splatter filter used. Look at Steve QIX's 3 diode limiter circuit.

As you may know, on transmitter power supplies...the resonance is usually in the hertz or sub-hertz region. This shows up as a change in carrier level when modulating and is called "carrier shift" due to "power supply bounce". Most reasonably built transmitters have little supply resonance. I know you knew that though.

In the 9100A, try backing off your multiband clipping and run that into the 222 (maybe with the pre-emphasis turned off). The 222's overshoot compensator changes the phase of the signal before it is clipped to reduce the amount of square waves created...cool trick!

On the 222..I too am very impressed with it. It smooths out my high end better than many other limiters.

For fun and weirdness, I obtained an Orban TV 8182A (and some 8182ASAP's-different story for another time). I have this after my Gregg 2540 and before the 222. The 8182A has something called a Hilbert Clipper, which is just a simplified speech processor that creates IM distortion instead of clipping distortion. This adds a few DB of apparent loudness to a voice signal. This is what the 9105A also did, but better. I kinda like my chain better as it is warmer.
 
Hope we did not hijack this thread too much. Again a D104 can be extremely respectable-simple too!

Derb-Man...just get on the air with your Master of Metal, don't worry so much!

Good to hear your thoughts Bruce (and others)!

73,
Dan
W1DAN

Hi Dan,

Thank you for your detailed reply, and I'm glad to hear that you are doing well. Out of curiousity, what are you currently studying in school?

Let me respond to your questions on a point-by-point basis:

1) The measured -3dB end-points for my rig, in the proof-of-performance mode, are about 35 hz to 12 khz.

2) Number of audio xfmrs in the signal path: (2) in the microphone preamp; audio input xfmr in the audio driver; audio driver xfmr; modulation xfmr (and mod. reactor). That makes five (5) xfmrs in the signal path. Lots of opportunity for phase shift and ringing/overshoot beyond the processor, as there are obviously (3) xfmrs after the processor.

3) Ahhhh.....the always problematic class B audio driver xfmr! I can eliminate it by going with a cathode follower audio driver ( I have the cathode follower audio driver board from a Gates BC-1G), but I'm waiting for the FET audio driver board that Steve/QIX is developing; Steve, I hope that you are listening! That would eliminate the existing audio input xfmr that also splits the phase to the push-pull driver, and the driver xfmr.

I have played at length with the clipper setting in my 9100A. I can really increase the modulation density by increasing the clipping level, as would be expected, but I am not enamored by the effect that so much clipping imparts to the audio. At any setting of the clipping, the unpredictable modulation characteristics of my transmitter are still apparent with the Optimod 9100A. At 10 to 12 dB of gain compression with the Inovonics 222 and with it's lack of clipping circuitry, it is just as effective (from a modulation density standpoint) as the multi-band Optimod-AM and sounds just as good.

I have occasionally thought of adding the 3-diode limiter circuit to my rig, primarily as a "protection of last resort" for the modulation transformer, but to me, this is analogous to putting the cart before the horse (or something like that!). Since the Inovonics 222 provides such great control of the modulation characteristic, and it's use virtually eliminates the possibility of hitting the baseline while maintaining adequate modulation density, I see no need to add the diode limiter at the present time.

Funny you should mention the Optimod-TV 8182A. I have been seriously thinking of buying one for the last month or so, as they appear to be being dumped lately, possibly with the demise of broadcast TV, and in some cases at bargain-basement prices. My reason for obtaining one; to try modifying it for AM use, and to remove the clipper circuitry, thereby making it more usable with my rig, and to have the advantages of 4-band gain compression. Have you considered doing this with your 8182A?

I think you, Rob/K5UJ, Bill/KC2IFR, and I are the biggest cheerleaders for the Inovonics 222!

73,

Bruce