The AM Forum
March 29, 2024, 10:56:10 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2   Go Down
  Print  
Author Topic: The 3-Diode Negative Peak Limiter Circuit - Experiences ?  (Read 10284 times)
0 Members and 1 Guest are viewing this topic.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« on: March 29, 2020, 01:51:24 AM »

I just built up a 3-diode NPL circuit.  I have a Variac adjustable supply for the keep-alive circuit and diodes rated at 4X the high voltage.  Didn't install it yet.

This is for my 4-1000A plate modulated rig. I want it as a safety net in case I go over 100% negative. However, I would like to lean into it at times to see how much it will take before generating IMD sidebands.  My low level DSP bandwidth filter is working FB at about +- 5.5 to 6.0 KHz, but I want to improve things further, by doing some limiting at the high level point.

I have used this NPL circuit in many rigs in the past and found when I hit it too hard it produced some splatter. However, I am told that this is the best way to go for a high level plate modulated rig - better than a low level limiter. I want to keep an open mind and try it again.

For those of you who have tried one, how did it perform when used as an occasional neg peak safety net or when hitting the negative peaks hard to see higher positive peaks? Did you ever run IMD 2-tone tests or harmonic distortion tests to see the effects?   Or watch the bandwidth on a spectrum analyzer as you leaned into it?  Have you ever been in a Turkish prison?  (Airplane)

In the meantime, I should have mine running in a few days at most - with some test results.

Thanks.

T

The 3-diode NPL circuit:

http://www.amwindow.org/tech/htm/3diodeka.htm

Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #1 on: March 29, 2020, 09:53:50 AM »


I run one in my Val-I-Ain't Duce... works fine.
Never tested it.

One fellow (who will remain nameless) has a web presence/site that claims a
reduction in "nasties" with the addition of a single resistor. You can probably
find that info... would be neat if you tested his premise.

                     _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
w4bfs
W4 Beans For Supper
Contributing
Member
*
Offline Offline

Posts: 1433


more inpoot often yields more outpoot


« Reply #2 on: March 29, 2020, 10:02:29 AM »

a qso with K8VYZ years ago while he used a 3 diode circuit indicated that one of the diodes needed to be fast in order to prevent distortion.  He used a damper diode from tv service.  Maybe others remember more.
Logged

Beefus

O would some power the gift give us
to see ourselves as others see us.
It would from many blunders free us.         Robert Burns
N1BCG
Member

Offline Offline

Posts: 836


« Reply #3 on: March 29, 2020, 10:49:18 AM »

The resistor addition is important to round out the crossover point between negative modulation and keep-alive supply. This softens the clipping action for a cleaner result at some expense of peak control and a somewhat higher + D.C. supply might be needed.

Generally this is only an issue if the clipper is abused, ie, for artificially generating extreme positive peaks. Watt-Watchers who use SSB PEP meters on AM love the bedazzling result.
Logged
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #4 on: March 29, 2020, 11:09:46 AM »

Ashtabulah Bill used what looks like a 3 diode in the video.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #5 on: March 29, 2020, 12:51:44 PM »

Thanks for the comments.

Interesting on the potential need for one of the diodes to be fast-acting.  I'm using 6A, 1KV diodes marked  "TC   T6A100L".   I've read that the standard 1N4007 1A, 1KV diodes are good up to about 15 KHz, but there is no consistent data on the web.    Which of the diodes would likely need to be fast acting,  D2, the series diode?


If I get unacceptable IMD numbers during the tests, I may try some Schottky or something faster than the standard power supply diodes I am now using..


Also, if I wanted to quickly bypass the 3 diode NPL for an A/B test, what is the easiest way to do this? Would a jumper across D2 do it or would it require a more complicated disconnect to become invisible for quick out-of-circuit tests?

Yes, the 3-diode NPL appears to be a great way to avoid mod transformer damage form negative spikes and splatter from rogue neg peaks. Though, I have heard and seen terrible splatter generated trying to hammer a diode limiter/clipper to generate big positive peaks. But these cases are usually low level diode clippers. There definitely needs to be more "measured" experimentation using this 3-diode circuit on high level plate modulated rigs. I understand if the modulator is DC coupled (like a class E design) then low level clipping works OK. (good control)  A tube plate modulated rig is not DC coupled, thus the need for high level limiting/ clipping.

IE, no matter how clean the low level driver is, a lot of garbage can be generated in the final amplifier stage due to many different causes. Negative going peaks (without a limiter) or banging the 3-diode limiter too hard in the quest for big positive peaks may be important reasons.


T
 
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #6 on: March 29, 2020, 01:01:36 PM »

Didn't GFZ look into the fast diodes for this application?
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #7 on: March 30, 2020, 01:04:55 AM »

I had low expectations, BUT...  impressed so far!

I fired up the high level, 3-diode NPL tonight in the 4-1000A plate modulated rig.  I dialed in about 350 volts as the keep-alive voltage.   Under normal voice modulation the scope waveform and SDR waterfall looked the same as always, about +- 5-6 KHz wide transmit bandwidth.

I then brought the audio gain way up, socking some heavy Yallos. I was hitting the baseline quite hard.  Normally this would be severe over-modulation with a thin center line. But with the NPL working, the negative peak center line stayed slightly thicker as expected. I can adjust this thickness by the amount of keep-alive voltage.

The good thing is I was now doing well over 150% positive peaks and hitting the audio much harder than I normally do. (The REA mod monitor showed pinned positive peaks at 150% -  and the negative peaks were at the -100% stop.)  The bandwidth on the SDR analyzer stayed about the same, +-5 to 6KHz!  If I am seeing an accurate waterfall spectrum, then the NPL is doing better than I hoped.  Tomorrow I will get some other stations to give it a look and actually tune down the band looking for inter-mod.

Normally when I crank up the audio and intentionally overdrive an amplifier it almost always shows some deterioration in IMD producing some splatter. But evidently the 4-1000A modulators using regulated screen and grid supplies, -12DB of negative feedback, the WA1GFZ MOSFET audio driver board and BC mod iron can easily handle the extra clean power required when called upon.

I plan to run some 2-tone IMD tests tmw as well as some listener tests as already mentioned.

Steve/ HX:  BTW, I was told that fast acting diodes will produce more harmonics due to their faster acting clipping. Using these power supply diodes may be a better idea. I may have some good value choices in there already if these preliminary results are real.  I just wanted to use it as a safety net, but it may allow me to run the audio at a higher level without increasing the bandwidth. That would be quite a bonus.

Timtron  said he was able to optimize his 3-diode NPL by changing resistor values with a slight circuit mod.  In the meantime, Frank/GFZ is modeling the NPL to see if he can make it a softer limit. He talked about using several stages to progressively accomplish this. I am open to experimentation to get a softer effect.

More tomorrow...

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #8 on: March 30, 2020, 02:00:56 AM »


Tom,

   Your setup seems to be doing exactly what you want it to do. This stuff has been hashed over many times before, but as you stated, the 4-1 Modulator tubes are being driven cleanly, and the stiff power supplies also help. As to a softer limit with multiple thresholds, see the links pasted below. In my original implementation, back in the mid 80's, I used a 3 threshold NPL along with a Low Pass filter afterward. Could push a single 4D32 from a 100 watt resting carrier to around 800w PEP RF out!

http://amfone.net/Amforum/index.php?topic=30945.msg241375#msg241375
http://www.ing.unitn.it/~fontana/AM%20ALC.pdf

Always enjoy your posts,
Jim
Wd5JKO
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #9 on: March 30, 2020, 02:57:37 PM »

Well, as usual, 2-tone IMD tests don't take prisoners and there are no free lunches.

As I ran the tones hard into the NPL, the IMD started to creep up. I am seeing almost -45DB 3rd order at -95% negative, which is excellent cleanliness, but once the NPL line is violated at -100% negative, the IMD starts to climb.  I can get away with some extra voice audio when using the NPL, but when using tones, I cannot lay into it heavily.  So I will use it as a safety net as originally planned and ask the guys where the limit is before they start to hear sidebands crud when using voice.  As is, I can hit 150% without crud, so my expectations are high enough already.

I am still not sure why I can hit the voice audio very heavily and the bandwidth does not increase much on the SDR spectrum waterfall. The main difference is heavy tone average loading vs: light voice averages. I would think IMD crud would show up a little more on voice. But I'm not complaining cuz more audio with acceptable bandwidth is what I am looking for.

All in all, so far I consider the high level 3-diode circuit a good add-on to any plate modulated rig and will experiment further to find the limits and permanent settings. After all, it does not start working until the negative peaks cut off at  ~ -95%.  Just like my low level DSP limiter, it is invisible until hammered. I like that.


Jim: OK on your past limiter experiments… thanks for the info!


T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #10 on: March 30, 2020, 07:01:45 PM »

I got some real on-air signal reports today.  I ran the rig normally and then hammered the audio to see what it would do.

Steve, QIX noticed that I was hitting  -100% and crud being generated. As it turns out, I needed to increase the keep-alive voltage. This made a huge difference and the REA mod monitor was now limiting to -95% negative like it should.  I was able to get some good 150% positive peaks with -95% limiting, with good audio quality and 5-6KHz bandwidth.  The higher keep-alive voltage made all the difference.

It was suggested that the keep-alive should be regulated. It normally goes from <1 mA  to a heavy load feeding the negative cycle of the final.  I measured a drop from 450VDC to 415 under mild modulation... and down to 350VDC under a big yallo.  That is a lot of drop. Maybe I do need a regulator.

** Maybe that's why the tones are so much worse than the voice IMD... keep-alive regulation is MUCH worse with tones...  I will try regulated with a 11N90 MOSFET next.

Also the keep-alive voltage was very high. Someone used only 50 VDC when using fast acting diodes. Are my slow diodes dropping I/R losses thus requiring higher voltage?

All in all, so far I like the results and have been running my audio level higher - while getting good audio and bandwidth reports from the guys.  

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #11 on: March 31, 2020, 12:28:43 AM »

I know there are quite a few following this thread because you told me so in emails or on the air.  This 3-diode NPL is something that some will want to try on their rigs. So I'm going to do my best to keep you informed.

So before you buy any parts or build anything, let me order and try some fast-acting diodes and build up a regulated keep-alive supply first.  We might be able to improve on this design.

I put out a call to two of my engr friends asking what the best diodes would be ... fast recovery, axial leads and at least 1KV at 2A.  If anyone has a specific diode suggestion, please post it here.

With this 4-1000A rig, the missing link  has been the ability to crank up the audio when needed. I got a good taste of it yesterday and could hear a huge difference. The added distortion was minimal, so it has big potential perhaps with the faster diodes and regulated supply.  I liked the feel of increasing the audio gain and watching the big peaks rise with the -95% negative limiting, high level.

We will see if these two mods make a difference.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #12 on: March 31, 2020, 12:58:56 AM »

I did the 3 diode keep alive on solid state radios years ago, when I first heard about the circuit.

You'll do well to use a regulator on your bias supply.  I found that although it doesn't pull much current, it did have a tendency to draw the bias supply down fairly quickly.

I was working at 12 to 28 vdc, and found an lm317 with a boost transistor worked fine.

Maybe an LR8 regulator would fit the bill for you.  I'm now using them with a TIP50 for screen regulation on 4cx250 and similar tubes. This would give you huge amounts of variability (20 to 400 or so). Not sure what you're running for scream voltage on your 2x1, but maybe you could tap off that to supply an LR8? 

I believe the keep alive supply has to operate at an audio cyclic rate.

Never used fast acting diodes on mine, and gave up once I built my positive peak stretcher.

--Shane
KD6VXI
Logged
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #13 on: March 31, 2020, 02:44:38 AM »

In my DX-100 I made a change in the 3 diode NPL circuit. Rather than have a stiff keep-alive voltage supply, which I imagined may clip the peaks, I have a series dropping resistor from the supply to the NPL with a hold-up capacitor on the supply side. The idea was to have a soft limiting effect determined by the time constant of those components rather than hard limiter clipping. If the voltages and time constant are set correctly for voice, this should yield a gentle compression of the negative peaks.

Unfortunately, I haven't tested the circuit to see how it performs (my o-scope crapped out). It may take some iterations to find the right values, if indeed the concept works at all.

Perhaps a hard limiting point could also be included in the circuit — a "drop-dead keep-alive voltage"(Grin) — from a stiff supply that would come in after the soft NP limiting took place. It's a good place for experimentation.
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
WU2D
Contributing
Member
*
Offline Offline

Posts: 1800


CW is just a narrower version of AM


« Reply #14 on: March 31, 2020, 07:08:03 AM »

Steve described this to me by describing the circuit one line at a time over the air - draw this then draw that and pretty soon I had the circuit in pencil in front of me. This would have been around 1989. The radio was a TCS transmitter that I had modified for an outboard 1625 modulator so it was 2 1625s in the radio modulated by a pair. The transformer was a 2:1 step down - one of those 6.95 military grey jobs from Fair Radio I think out of the 618. Let me tell you, that 3 diode circuit worked slick!
Logged

These are the good old days of AM
KA2DZT
Member

Offline Offline

Posts: 2192


« Reply #15 on: March 31, 2020, 10:35:34 AM »

Tom

You need fast diodes,  I use TV damper diodes, a string of them to increase the PIV.  I only use a single diode (string) to apply a keep alive voltage on the class C end of the mod xfmr.  You do not need any other diodes.  

The keep alive PS should be about 10% of the HV.  The trick is to make the keep alive PS with a series of dropping resistors so the voltage sags under heavy modulation.  This voltage sag prevents the hard knee and rounds the edges that prevents splatter.  Each of the dropping resistors should have a filter cap.  Start with 10-20mfds and reduce down to a few mfds for the last cap.  Dropping resistors can be anything 2-3kohms.  Maybe 8-10kohm total.

I've been using this set-up and it works perfect.  The internal impedance of the keep alive PS keeps a load on the mod xfmr under heavy negative modulation.  No need for a 3 diode limiter.

Fred
Logged
N1BCG
Member

Offline Offline

Posts: 836


« Reply #16 on: March 31, 2020, 01:49:05 PM »

Some things to consider on this topic:

1) RF PA triodes and tetrodes arrive at -100% modulation (carrier cutoff) at different voltages, not at zero,  so the LVB+ used will be different for these configurations.

2) Asymmetry changes over the vocal spectrum so setting "polarity" will only be correct *most* of the time. The rest of the time, the negative peaks will dominate. This is what all-pass filters address and it's why these circuits can see anything from no action to plenty depending on your voice.

3) High-level clippers should be used to prevent carrier cutoff not to generate artificial positive modulation. This type of circuit was used successfully with two Gates BC1G transmitters on the broadcast band. The F.C.C. was fine with this modification because it kept the -100% peak flasher from coming on and that's it.
Logged
KB1VWC
Contributing
Member
*
Offline Offline

Posts: 40


« Reply #17 on: March 31, 2020, 03:39:09 PM »

  Let face it, any form of clipping causes distortion.  If you limit the negative excursion of the waveform, you are introducing 2nd and 3rd harmonic distortion into the waveform. The harder the knee, (squarer the bottom..faster dV/dT)  higher order and larger in amplitude harmonics are generated. I thought most of this circuit was to keep the mod iron from becoming unloaded during the occasional negative excursion past the baseline and blowing up the mod iron; not to increase positive peaks to something that no standard diode detector can handle. Most of them have serious distortion issues above 80%. If you are using an SDR or sync detector, at least it won't sound so bad.
   A spectrum analyzer watching the area around the carrier frequency with show the "stuff" outside of the desired bandwidth. However, our typical high level push-pull modulation, standard mod iron, even with negative feedback looped around it, you are still generating some higher order harmonics just from the normal audio process. You will easily see all of these artifacts with a spectrum analyzer or sdr with a single or two tone test. You won't see or detect an occasion voice artifact outside the normal bandwidth without a quasi-peak detector/analyzer or the equivalent on an sdr. 
  So all of that being said, one bounce off the ionosphere, can anyone really hear the difference? Even if you have your Telefunken Smooth plate 12AX7 in the first audio? The high end response is limited normally by the plate bypass cap/screen bypass caps. High end filtering is being done "naturally". So diodes that have a recovery time good to 15khz should be more than adequate. TV dampers seem like a good one or FR107's (1A 1000v fast recovery) commonly used switching power supplies. So if we add a few more harmonics or artifacts outside of our recommended bandwidth, as long as they are 30db down, who cares?  There are quite a few sideband stations out there that are wider than "fabio" has ever been.

Steve
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #18 on: March 31, 2020, 04:08:48 PM »

Hi Steve and Clark -

I think youse guys are correct.

I just added in a stiff regulated keep-alive supply and tried some two-tone tests again.  The KA voltage was stable at 375 VDC but the results were exactly the same as with an unregulated supply. .  When the NPL point is approached the IMD rises off the floor and gets worse and worse as the audio is increased past this point.  It doesn't matter if the NPL is limited to -95% or even -90% or the KA voltage is stable. When using tones and the invisible -100% point is reached, the IMD goes to hell on the SDR spectrum scope.

Yes, it will save the mod iron on random negative peaks which is a good thing. And it appears to let me run the voice audio up slightly with no extra bandwidth that I can see.  So I will keep it inline.  I don't plan on switching to faster diodes cuz the results will probably be the same.

The only thing that stays in my head is the ability for a low level soft limiter to work very well to limit peaks and still stay clean. I use one now in DSP and it works FB. I imagine that duplicating this in the high level position would be very difficult considering the amount of HV parts needed and the requirement for a physically large low pass filter.

So, I will use the 3-diode NPL for mod iron protection and see if it helps me run a slightly louder audio during rough conditions. (at the expense of slightly higher IMD)  I can presently hit a "semi-natural" 150% positive at -95% negative while maintaining +-5.5 to 6 KHz bandwidth, with loud audio and mod iron protection.  That's FB with me.

Presently, I have my DSP Butterworth low level LP audio filter starting a roll-off at +-4.5 KHz. The last bit of sibilance usually ends at about +-6 KHz, as per design.  As long as the NPL doesn't bother these tight parameters under normal use, I will be happy.

Bottom line is IMD tone tests don't take prisoners and there's no free lunches.   Wink

T

*** Funny how we went thru this almost exactly 10 years ago - I'm even using the same NPL board today.   Totally forgotten:
http://amfone.net/Amforum/index.php?topic=23159.0


**  Steve/QIX summed it up nicely 10 years ago about the 3-diode circuit:  "... Once the system starts clipping, you will obviously generate some audio harmonics which can make the signal wider, but this will occur with ALL waveform modification systems." 


Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #19 on: April 08, 2020, 11:10:06 PM »

I was talking with Frank/GFZ about my 3-diode circuit. He agreed that the IMD garbage seems to start wherever the 3-diode circuit is set.  IE, if set at -90%, the IMD starts at -90%, not at -100%.   I am using it just as protection for the mod iron and did not like the IMD generated when leaned into. But we were warned about that already.

I came across another simple circuit that uses just a string of diodes connected to the modulated HV and a resistor to ground. Negative cycle loading is an old idea. I can see how it would put a safe resistive load on the mod transformer when at -100%.   I'd like to give it a try.

Question: How would I know it is working by looking at a scope or other indication?  What would happen when it is hit hard at -100%...  produce IMD?
How would the resistor power and values be calculated?   I see they used about X5 times the HV peak modulation voltage to calculate the diode string PIV rating.

From my calculations using their  Valiant parameters of 630V, 280 mA and 180 watts input, I see they used about the plate impedance for the loading resistor value.  They used about 1/9 the input power to calculate the 20 watts as the load resistor power rating.  Sound close enuff if I scale it up for the 4X1 rig?


If anything I could get rid of some more shack clutter with the 3-diode supply and diodes and replace it with something simpler.

Additional info - Frank has a working neg cyc loading circuit in his own Valiant and said:  "I referenced the load to ground so I could put a Zener in line with the resistor. I put a Dialco
led indicator across the LED [Zener?] as a base line monitor. It flashes if I go below the baseline."     This would make a flashing indicator at -100%
(I will have to ask him about this, but sounds like an interesting modification.)


See circuit below. (Just the diodes and resistor, skip the screen and scope stuff)

T


* Negative Cycle Loading.jpg (94.55 KB, 826x637 - viewed 325 times.)
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #20 on: April 09, 2020, 11:20:17 PM »

I installed the negative cycle loading module into the 4X1 rig today.  The LED flashes whenever I see -100% on the scope.  It looks like the circuit is working FB to protect the mod transformer. I think the advantage of this circuit is we are not tempted to lean on it to increase positive peaks like the 3-diode circuit. (human nature)  When this one hits -100%, that's  it... the carrier cuts off, period. The xfmr is protected with a resistive load.  No IMD generated if we stay above -100%... just set the low level limiter instead.  

Refer to the circuit posted below as drawn up by Frank/GFZ.   Rather than a zener, I put three regular diodes at the bottom of the string and put a 1.5V LED across it. When the LED is in the proper polarity, it flashes only when modulation hits -100% or more. It does get brighter when the negative peaks go farther below cutoff, so we need to watch it to be sure it doesn't flicker at all.  Just like my low level DSP low pass audio filter, I think a clean rig will have good control with a low level limiter and a high level transformer load (NCL)  when needed..

After a few transmissions to intentionally hit -100% for testing, the power resistor was barely luke warm to the touch. Same with the diodes. Except for generous diode PIV ratings, I don't think this circuit absorbs much power when conducting... or at least the -100% peaks are so brief and sharp, this is the case.

T

Negative cycle loading circuit below for Frank's Johnson Valiant:


* Negative peak loading.png (12.52 KB, 705x497 - viewed 322 times.)
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #21 on: April 10, 2020, 01:01:41 AM »

I like the LED flasher
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #22 on: April 10, 2020, 07:07:52 AM »

Where would I connect it? I don't have any blink'n lights yet.  Grin

See schematic.

* 813 Transmitter.pdf (185.53 KB - downloaded 110 times.)
Logged
N1BCG
Member

Offline Offline

Posts: 836


« Reply #23 on: April 10, 2020, 09:45:44 AM »

ALL AM transmitters should include a circuit like this. So basic yet so essential. The only caveat is using it with tetrodes where cutoff occurs well before the plate reaches zero volts.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #24 on: April 10, 2020, 12:55:09 PM »

ALL AM transmitters should include a circuit like this. So basic yet so essential. The only caveat is using it with tetrodes where cutoff occurs well before the plate reaches zero volts.

OK on the tetrode screen risk. I have individual current fail safe shutdowns for all screens, grids and plates in the rig. That should cover these screen events??

Yep, it is a FB circuit.  But it still requires responsibility to watch those neg peaks on the scope, REA mod monitor and flashing LED.  Or, use an effective low level limiter.  That poor IMD generated by beating on a high level limiter or by -100% negative is real and to be avoided.

By coincidence, I had a nasty audio parasitic go thru the system while testing last night.  By accident the loose audio input jack hit ground while transmitting and generated a square wave thru the system big time. Everything hummed and made loud audio parasitic noise. But the rig recovered, probably saved by the NCL circuit.  There have been many, many pieces of ham mod iron that have bit the dust this way.  (The mod iron ball gaps probably arced and saved the day too.)


John, referring to the schematic, the diode string on the left connects to the modulated B+.  Anywhere between the mod xfmr output and plate choke bypass cap will do.  The other side connects to ground.  The diodes become a "short" (actually 0.5 V per diode drop) and the power resistor is esssentially across the mod xfmr during -100% negative excursions. Without the circuit, the xfmr sees only the mod choke (if used) at -100% neg and inductive voltages can soar.

In my case, I mounted the blinking LED on the front panel of the 4X1 rig.  I expect to use the scope and REA mod monitor for constant viewing... and the blinking LED more as an indication that the NCL (negative cycle loading)  is working. The scope and REA are a better operating view of negative peak activity than the LED.


T

Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.091 seconds with 19 queries.