modified heising modulation

[modulation_apprentice]

right now i have a modified heising circuit setup with a 35hy choke, 2uf cap, and an audio transformer, on the audio transformer 8 ohms side is fed by my rack and pa amp. then on the hi impedance side of the audio transformer one leg is grounded and the other leg goes threw the cap to the reactor. normal setup i'm guessing. what i am wondering is can someone point me in the correct direction to add negative peak limiting.

thanks

[Pete, WA2CWA]

What do we call you? Name or call so we have some idea who we're talking to.

[modulation_apprentice]

ow didnt introduce myself sorry about that name is brandon thanks for asking.

[WD8BIL]

There's a couple Peak Clippers here, Brandon. With your setup I don't see why it wouldn't work across the reactor considering it is basically doing the job the mod transformer's secondary would be doing in the "standard" setup.

[steve_qix]

Hi !

The 3 diode circuit shown on the am window is not a great circuit.  There have been numerous technical discussions about this circuit.

Here is a better circuit that I used at a many broadcast stations, including a 5kW station.  It will not put undue stress on the modulator or related components    The advantage is that the modulation is controlled exactly, due to the stiff power supply, and the resistor can be very small in wattage.  This circuit also protects the modulation transformer during the entire audio cycle, and keeps a constant load on the modulator.

[modulation_apprentice]

as far as the current ratings of the diodes are those minimum current ratings or can you use bigger? the reason i ask i happen to have some huge diodes here with .45 amp ratings but didnt know if that big of a difference would affect the circuit?

[steve_qix]

as far as the current ratings of the diodes are those minimum current ratings or can you use bigger? the reason i ask i happen to have some huge diodes here with .45 amp ratings but didnt know if that big of a difference would affect the circuit?

The current rating depends on the current in the RF amplifier.  You need to figure that the current will double at 100% modulation, and will TRIPLE for 200% (positive) modulation - that is PEAK current, not average.  The average current will be somewhere around 1.5 times the DC, under modulation.

An RF amplifier running at 200mA should really have a 400mA diode minimum (to assure some safety factor), and really a 600mA or 1A diode is preferred.  However, for a 200mA RF amp, a 450mA diode would certainly work in my opinion :-)

Regards,

Steve

[modulation_apprentice]

as far as the current ratings of the diodes are those minimum current ratings or can you use bigger? the reason i ask i happen to have some huge diodes here with .45 amp ratings but didnt know if that big of a difference would affect the circuit?

The current rating depends on the current in the RF amplifier.  You need to figure that the current will double at 100% modulation, and will TRIPLE for 200% (positive) modulation - that is PEAK current, not average.  The average current will be somewhere around 1.5 times the DC, under modulation.

An RF amplifier running at 200mA should really have a 400mA diode minimum (to assure some safety factor), and really a 600mA or 1A diode is preferred.  However, for a 200mA RF amp, a 450mA diode would certainly work in my opinion :-)

Regards,

Steve

ok sounds good the reason i was wondering is just the fact that there is alot more of a differnce in ma's.  the 6146 that i am feeding direct from the modified heising circuit is only pulling .030 mills..  give or take a few for a 10 watt carrier. and the diode is almost half an amp. i mean its total overkill but if it would work i would be able to start putting it together this weekend, if not then be watching the mail.

[steve_qix]

I don't see why the diodes you have won't work !  The current draw is so low it does not seem to be an issue.  Most diodes are fast enough to switch at audio frequencies.

I noticed the current is very low.  You may be able to get away without the modified heising circuit.  Most transformers will handle a certain amount of unbalanced DC,  and 30mA is pretty small !!  Some of the transformer experts may have further opinions about this, but I would try it without the heising first and see how it goes !

Regards,

Steve

[WD8BIL]

The 3 diode circuit shown on the am window is not a great circuit.  There have been numerous technical discussions about this circuit.

Ummmm...... Steve......

Your improved 3 diode limiter IS one of those limiters found on the AM Window
But it is good you pointed out the differences.

[modulation_apprentice]

I don't see why the diodes you have won't work !  The current draw is so low it does not seem to be an issue.  Most diodes are fast enough to switch at audio frequencies.

I noticed the current is very low.  You may be able to get away without the modified heising circuit.  Most transformers will handle a certain amount of unbalanced DC,  and 30mA is pretty small !!  Some of the transformer experts may have further opinions about this, but I would try it without the heising first and see how it goes !

Regards,

Steve

i guess what you could call my small current project would be practice. i'm just gettin involved in this type of thing with some help of a friend and we figured start small and grow from there. more enjoyment then anything, thanks for all the help i will keep you all informed of any hopefully progress.

[Opcom]

Starting small is best from a parts-source view. There's plenty of monsters around, no need to hurry into that till you want to. I'm glad you are doing this experiment. If you use the 3-diode limiter, can you post RF envelopes to show it off? I don't recall seeing actual waveforms from that kind of circuit.

[modulation_apprentice]

yes no problem at all, all that i have is a yaesu yo-100 but i will do my best. i can't wait to see how it affects my setup.  should be interesting for sure. i have a digital camera that will take a decent 60 second video clip, so it should not be hard at all. i will keep everyone updated.

what is the voltage rating on that 40uf cap? just good enought for keepalive power supply or does it need to be good up to b+ voltage. i am guessing same value with safety factor as keep alive but just wanted to make sure.

[modulation_apprentice]

one more thing while i am thinking, as long as the keep alive voltage is rectified dc and adjustable, the 40uf non polarized cap is all the filtering i need? or should the keep alive voltage circuit have its own filter cap also? hate to ask silly questions but i want to make sure i get it right the first time and did not overlook anyth  ;Ding.

[k4kyv]

Steve, what kind of diodes did you use in the 1 kw and 5 kw transmitter?  They would have to be something on the order of 10 kv p.i.v., and D2 would have to be rated at 1 amp or more.  I'm not absolutely sure that the usual stack of silicon 1 kv power supply rectifier diodes would work.

[steve_qix]

Steve, what kind of diodes did you use in the 1 kw and 5 kw transmitter?  They would have to be something on the order of 10 kv p.i.v., and D2 would have to be rated at 1 amp or more.  I'm not absolutely sure that the usual stack of silicon 1 kv power supply rectifier diodes would work.

Absolutely correct!  I had a big stack of high current diodes for the series diodes (as I recall, 600 volts @ 5 amperes), and lots and lots of 1kv power supply diodes for the other 2 stacks.  Worked for years - never, ever, lost a diode !  Only the series diode has to be high current.  The others handle relatively little power, but do have to deal with higher voltages.

Regards,

Steve

[steve_qix]

one more thing while i am thinking, as long as the keep alive voltage is rectified dc and adjustable, the 40uf non polarized cap is all the filtering i need? or should the keep alive voltage circuit have its own filter cap also? hate to ask silly questions but i want to make sure i get it right the first time and did not overlook anyth  ;Ding.

You probably can use just the single cap. Doesn't have to be non-polarized, since the voltage never reverses :-)

Regards,

Steve

[steve_qix]

Starting small is best from a parts-source view. There's plenty of monsters around, no need to hurry into that till you want to. I'm glad you are doing this experiment. If you use the 3-diode limiter, can you post RF envelopes to show it off? I don't recall seeing actual waveforms from that kind of circuit.

Here is a modulated waveform from a transmitter equipped with negative peak limiting, and with high modulation capability.  The percentage of positive modulation shown in the picture is approximately 200%.  The negative peak clipping is very noticable in the picture.  The unmodulated carrier is shown for reference.

Without the negative peak limiter (clipper) circuit, the positive modulation is CONSIDERABLY less, and virtually every syllable causes overmodulation in the negative direction.  The negative peak limiter (clipper) prevents this!

[AB2EZ]

Steve

Two questions:

1. Would the design, attached below (as a variation of your design) also work. If it would, the advantage would be that you would only need one HV diode stack.

2. Does it make a difference if the modulator is essentially a voltage source?
I.e., feedback control implemented to make the modulator's output voltage follow the modulator's input voltage, independent of the load impedance (within reason).

For example, if I look at the Class H modulator design used with my Class E transmitter...  it is designed as (I believe) a current source. I.e., the 0.5 ohm source resistors force the current to follow the input voltage according to di (change in output current) = dv (change in input voltage/0.5 (ohms). Thus the Class H modulator would probably become very "unhappy" (accentuated non-linear distortion) if it suddenly looked into an essentially infinite differential load resistance on negative peaks.

On the other hand, my modern Samson audio amplifier behaves more like a voltage source... so if, on negative peaks, the differential load resistance it is driving becomes infinite (i.e., increased negative voltage does not result in increased negative current) ... it might not (does not, as far as I can tell) accentuate the effect of negative peak clipping.

Therefore... could it be at a voltage source type modulator would not benefit from a 3-diode limiter... whereas, a current source type modulator (or a classic vacuum tube modulator with little or no negative feedback) would greatly benefit from having a keep-alive circuit that inserts the missing load resistance on negative peaks?

Best regards
Stu

[steve_qix]

Hi Stu,

I believe the circuit you posted will work.  It assumes there is no stray capacitance that would prevent the voltage appearing at the non-modulated end of the resistor from following the voltage at the modulated end of the resistor 

This will work with all voltage source type of modulators, which I believe virtually all modulators that we use should be !!

The class H modulator is definitely a voltage source.  There is quite a bit of feedback around the loop, and the combination of all of the .5 ohm resistors in parallel (in mine, less that .1 ohms) is a very low internal resistance with respect to the load (about 5 ohms).  The feedback lowers the internal resistance even more!  I use the class H modulator for testing new, first-time-out RF decks, because it will work equally well into a wide range of loads.

The class H modulator, as documented has a negative peak limiter buit in, which essentially causes the modulator current to fall to 0 (zero) once the keepalive diode conducts.  Anyway - different discussion best reserved for Email !

But, back to the original question - it looks very workable to me !  I couldn't see any issues with your circuit, Stu, and, assuming the keepalive voltage is quite a bit higher than the diode voltage drops, should work equally as well as my 3 diode limiter.   And, since both circuits keep a constant load on the modulator (VERY important), even a relatively poor modulator with lots of internal resistance (no feedback, high impedance, etc.) should work just fine with either circuit (assuming the modulator has the reserve power for the additional positive peaks, that is!!)   

Regards,

Steve

[AB2EZ]

Steve

Thanks!

Stu

[W9GT]

I have a question.  What should the voltage rating be for the audio coupling capacitor (between the mod xfmr sec and the mod reactor) in the "modified Heising configuration"?  For a max plate voltage of 3KV DC on the final RF amplifier,  I was planning to use a big oil cap 2MFD @ 5KV or 7500VDC wkg volts.  Perhaps that is not a high enough voltage rating?  Should you calculate the minimum voltage rating of that capacitor (PLUS a safety factor) as equal to the final plate voltage, or should it be rated for peak (x 1.414) or maybe even peak to peak (X 2.828)?

Thanks and 73,  Jack, W9GT

[steve_qix]

I have a question.  What should the voltage rating be for the audio coupling capacitor (between the mod xfmr sec and the mod reactor) in the "modified Heising configuration"?  For a max plate voltage of 3KV DC on the final RF amplifier,  I was planning to use a big oil cap 2MFD @ 5KV or 7500VDC wkg volts.  Perhaps that is not a high enough voltage rating?  Should you calculate the minimum voltage rating of that capacitor (PLUS a safety factor) as equal to the final plate voltage, or should it be rated for peak (x 1.414) or maybe even peak to peak (X 2.828)?

Thanks and 73,  Jack, W9GT

Generally, I see capacitors rated somewhat higher than the DC voltage.  In theory, there should not be a lot of voltage across the capacitor, save the DC - and then, only in some modified heising configurations (where the mod transformer secondary is tied to ground through the capacitor,  or if the secondary is grounded and the cap is floating). 

In the configuration where the mod xfmr secondary is tied to the DC side of the choke, and the cap is connected from the other end of the mod xfmr secondary to the top end of the choke, the voltage rating of the cap can be lower (no DC across the cap and the AC should be very small). 

I tend to prefer the latter configuration because the capacitor is not charging up to the DC plate voltage through the mod xfmr secondary.

Hope this helps !

Regards,

Steve

[AB2EZ]

Jack
Steve

There is another issue. I learned about this when I read one of the ancient references (I'll add the title to this post later... it is available for download at no cost).

Even though the inductor is large and the capacitor is large... there is a low frequency at which these form a resonant circuit.

For example, if the inductor is chosen as 20H (and assuming the modulation transformer has a magnetizing inductance that is higher than 20H) and the capacitor is chosen as 2uF (more on the choice of capacitor value below)... then the resonant frequency will be 25.2 Hz. At that frequency, there will be an additional voltage across the capacitor which can be very large... limited by the Q of the circuit formed by the capacitor and the inductor and the plate resistance of the tube(s).

Separately (but related)... the best performance (flatness of frequency response at the low end) is obtained when the capacitor value is chosen to resonate with the inductance (the combination of the Heising choke's inductance and the modulation transformer's magnetizing inductance, in parallel) at the frequency where the impedance of that inductance is equal to the plate resistance.

Example (used with my plate modulated legal limit transmitter):

Plate resistance = 1600 volts / 300 ma ~ 5333 ohms
Heising reactor value: 50H
Magnetizing inductance of the modulation transformer: greater than 50H

Total of: Heising reactor inductance and modulation transformer magnetizing inductance in parallel: ~50H (neglecting the effect of the magnetizing inductance of the modulation transformer)

Frequency at which the inductor's reactance = the plate resistance: 17Hz

Optimum value of coupling capacitor (5333 ohms at 17Hz):1.8uF

When modulated with a sine wave at 17 Hz, the peak voltage across the capacitor will equal the sum of the B+ and the peak modulation voltage ~ 2 x the B+, for 100% modulation

Therefore, it is a good idea to choose the capacitor to have a rating of at least 3 x the B+ to allow some margin.

In my transmitter, I used ten (10) 18uF electrolytic capacitors in series, each rated at 450 volts. I used a 100k ohm balancing resistor across each capacitor. This is  2.8x the B+... but so far, so good. In any event, I don't plan to apply a 100% modulating sine wave at 17Hz... so that helps.

Stu 

[steve_qix]

This cannot be emphasized !! Watch out for those resonances !!!!  There is some info somewhere (maybe it is Stu's reference) about this on the 'net (I've seen it, but don't remember where !).

All of the old broadcast transmitters that I've seen use modified heising, and they use surprisingly small capacitors - taking advantage of an optimal resonance with respect to the load.  Outside of this optimal zone, BAD THINGS CAN HAPPEN blowing components left and right due to high voltages (speaking from personal experience here)!!!  Don't kill your mod transformer, reactor or heising cap by creating a resonance with no where for the voltage to go 

When I've been in doubt, I've used a rather large heising capacitor (with respect to the impedance of the load) configured such that no DC appears across the cap.  The capacitor voltage hasn't been an issue [still, the voltage rating of the cap was a couple thousand volts with 4000VDC appied to the RF amp], but again, with such a large cap (and a pretty big heising reactor also), the low frequency resonance was not a problem.

Regards,

Steve