Homebrew 4-1000A plate modulated by a pair - Questions

[WA1GFZ]

Jeff,
You have one nice sounding BC 610.

[K1JJ]

Tom,

I found Stu's detailed description in this thread to be extremely helpful regarding the Heising cap.

http://amfone.net/Amforum/index.php?topic=18722.0

Jeff

Thanks, Jeff - that did the trick on the Heising cap value. . In addition, I wasn't aware that when using Heising, no matter how large the reactor,  the inductance of the mod transformer was the limiting factor to the low end.  Heck I have 70H of reactor going in there and thought that was hot sh*t...   But it makes sense, of course.

BTW, I meant to mention that the other night when we chatted briefly your rig sounded quite pristine. It was almost 60 over and the audio was back to the normal transparent norm. Was that the newer BC rig or the BC-610 as Frank mentioned?


Johnny - Yep we've had some crazy QSOs when you're aboard too. When Big Country, The Derb, Brentina, you  and a few others (like the chairman of the board and the prime minister) get going, you never know what's gonna happen. I shud of known you were crazy the first time you intrduced yourself at Timonium in a serious manner as W3FSH -" W3 Fat Smelly Ham"...

T

[AB2EZ]

Tom
Jeff

Et al. Again... I very much appreciate the kind words.

Jeff provided a pointer to my prior post on this subject... which makes this post a little redundant. [I posted this before I saw Jeff's post]. However... I'll leave it here anyway. Note the factor "K" which I have added here, and which is defined below.

On the subject of the size of the Heising capacitor. You may not notice the difference in the sound... between using a 10-20 uF capacitor and a 2uF capacitor, but the reason for using the smaller capacitor is as follows:

If you use a large capacitor, then frequency response of the transmitter will be 3dB down at the point where the reactance of the Heising choke in parallel with the magnetizing reactance of the modulation transformer (as measured on the secondary side of the modulation transformer)  equals the modulation resistance of the rf output stage.

For example:

Suppose: R= 3500 volts/0.4A = 8750 ohms; the Heising choke has an inductance of 70H; and the magnetizing inductance of the modulation transformer (as measured from the secondary side) is 35H.

The combined inductance (in parallel) of the Heising choke and the magnetizing inductance of the modulation transformer is: 23.33H [Note that making the Heising choke larger is not going to increase the value of the parallel combination very much]  

Then the low frequency 3dB cut off is 8750 ohms/[2 x pi x 23.33H) = 59.7 Hz.

[That's how the numbers come out when the modulation resistance of the RF output stage is that high]

If the Heising capacitor has a value of 20uF, then the combination of the Heising capacitor, in series with the Heising choke, and in series with the magnetizing inductance of the modulation transformer will have a series resonance at a frequency of  3.5 Hz. The "Q" of this series resonant circuit will be around 3.85. I.e., at the resonant frequency of  3.5 Hz the impedance of the 20 uF Heising capacitor will be around 2274 ohms; which is (1/3.85) x the modulation resistance of the rf stage.

What this means is that a sub-sonic oscillation of the modulator (at 3.5 Hz), or any other sub-sonic audio injected at a frequency near 3.5 Hz, can cause a very large voltage to appear across the secondary of the modulation transformer, across the Heising choke, and across the Heising capacitor.

If that never happens (sunny day scenario), then its not a problem. If it happens ... e.g. a truck goes rumbling by your house causing ~3.5 Hz mechanical vibrations of your transmitter... that get converted into low frequency audio by microphonic effects (rainy day scenario)... then your expensive components may destroy themselves. This is the Tacoma Narrows Bridge scenario... playing out in your transmitter.

The small capacitor approach has two benefits

1. It moves the series resonant frequency up to a higher frequency... where the Q of the series resonant circuit will be close to 1 (no huge voltages produced by an audio input at that series resonant frequency).

2. The low frequency rolloff characteristic is actually slightly better... i.e. the modulation frequency response will be flat down the the frequency of 59.7Hz (instead of 3dB down). Below that point, it will roll off fairly quickly.

The size of the small capacitor is chosen as follows (this is an approximation):

The impedance of capacitor at the low frequency rolloff point (59.7Hz in this example) should equal (1/K) x the modulation resistance of the rf stage (8750 ohms in this example), where K is [the inductance of the Heising choke and the magnetizing inductance of the mod transformer in series / [the inductance of the Heising reactor and the magnetizing inductance of the mod transformer in parallel]

In this example: K = [70H + 35H]/[23.33H] = 4.5

So the impedance of the Heising capacitor should be 8750 ohms/ 4.5, at the frequency 59.7 Hz.

Therefore, in this example, the Heising capacitor should have a value of around: 1.37 uF.

As a second example:

In my Viking Ranger, the stock modulation transformer has a magnetizing inductance (referenced to the secondary) of around 7H. Using the 9-pin socket on the back, I can run the Ranger in Heising configuration. I'm using a Heising choke with a value of 20H. The Ranger has a modulation resistance of around 600 volts / 0.120A = 5000 ohms. The parallel combination of the Heising choke (20H) and the Ranger's mod transformer's magnetizing inductance (7H) is 5.2H. The low frequency response, therefore, is flat down to around 5000 ohms /(2 x pi x 5.2H) = 153 Hz. The value of K is 27/5.2 = 5.2. The best value for the Heising capacitor is around 1.1 uF. I am using four (4) 4.7uF capacitors in series, so the Heising capacitance I am actually using with my Ranger is 1.2 uF.  At the lower rolloff frequency (153Hz), the impedance of the Heising capacitor I am actually using with my Ranger is around 833 ohms. It resonates with the series combination of the Heising inductor and the transformer's magnetizing inductance at a frequency of around 28Hz. The Q of the series combination at resonance is around 1 (i.e. the impedance of the Heising capacitor I am using, at 28Hz, is 4840 ohms)

As a third example:

My home brew plate modulated transmitter uses a pair of Russian GS-35b triodes.

The plate voltage (at carrier) is 1700 volts. The plate current (at carrier) is 300 ma. The modulation resistance is 1700 volts/300ma = 5667 ohms.

The Heising choke has a value of 50H. The magnetizing inductance of the ferrite core toroidal modulation transformer is much larger than 50H.

In this case, the approximate formula for K (given above) doesn't hold. At low frequencies, the impedance looking into the secondary of the mod transformer will be the modulation source impedance (referenced to the output of the modulation transformer); and not the impedance of the magnetizing inductance of the modulation transformer. This impedance is resistive.  Series resonance will occur at the frequency where the Heising capacitor resonates with the Heising choke. The Heising capacitor should be chosen so that frequency will be the same frequency at which the Heising choke has an impedance equal to the modulation impedance of the rf stage (5667 ohms in this example). I.e. for this case, we should set K=1

Going through the numbers...

The low frequency cutoff is: 5667 ohms/ (2 x pi x 50H) =  18 Hz
The optimum Heising capacitor is 1.6 uF

Summary

If the modulation transformer has an extremely high magnetizing inductance... i.e., much larger than the inductance of the Heising reactor... then use K=1

Otherwise use the formula for K given above

Stu

[Steve - WB3HUZ]

Another benefit of a smaller blocking cap is better time domain/impulse response. Larger caps can create large overshoot which can cause damage, at worst or limit your peak levels of modulation at best. Neither are good.

[K1JJ]

Still busy wiring stuff up tonight and came across another question:

Let's say I have two HV supply voltages available, 3500V and 6KV from the same supply.  I would like to keep the voltage on the 4X1 rig all the time whether it is being used or not.  It's operating limit is 3500V. Can I switch the HV to 6KV while the 4X1 rig has its filaments switched off? I know it won't hurt the tubes, but how about the mod iron and associated Heising parts that normally would crap out if operated at 6KV?  I plan to float the mod iron on Plexiglass so there is no HV path to ground. The other path is through the tube plates, but the fils are off.

Is this OK or am I missing something?

BTW, heard Steve/ HUZ and Frank/GFZ on tonight looking for west coast. Steve, for the vac variable flange you were talking about, just build a small frame outa alum angle and sheet and mount BOTH of them side by side or one over the other. (C1 and C2)  Look at the pics you took. I've done it for years this way - no big deal.  Strong and wide paths for conduction.

It was a good conversation - I wanted to join in but the shack is torn apart. Next week shud have the new rig running on AM.

T

[W2NBC]

Frank and Tom,

Thanks for the signal reports. You both heard different transmitters; Frank the BC-610, and Tom the Yellicrafters HT-37 done up with mods posted on the AM Window.. It definitely was great chatting with both of you!

Stu,

Your posts have really been educational, as are Don's, Tom's, Steve's, Dave's, Derb's, Slab's, Frank's, and too MANY others to mention on the Forum. The "search" function is your friend! 

Which brings me to another subject: a compilation of topics that could be gleaned from the VAST resource that this group has contributed to a HUGE archive that keeps growing (with all the server space, etc.)  For instance, the "Heising cap value" and its associated topics (all Heising related) could be collected and written out as an "open-source" book and chapter form, etc. The point is that MANY of these posts are worthy of serious documentation as they really are a legacy to the ART of AM radio..


Jeff

[AB2EZ]

Tom

On the 6kv question:

1. The Heising capacitor is going to charge up to 6000 volts. It needs to be rated at about 3x the normal B+ anyway (3500volts x 3) ... so that probably won't be a problem.

2. You need to ask the question of whether there is any "rainy day scenario" where the high voltage applied to the rig can ever transition, abruptly, from 0 to 6000 volts. [For example: a loose HV connector that disconnects the supply from the rig for a short time when someone jiggles the HV cable].


2a If that happens, the Heising capacitor will be starting out with 0 volts across it, and the entire 6000 volts will appear as a 0=> 6000 volt step in voltage, across the secondary of the mod transformer. Although this is one of the reasons for using a small value of Heising capacitance (i.e., to shorten the duration of this transient), I would still worry about the implications of this.

2b.  I would worry about the effect, on the mod transformer, of having 6000 volts abruptly connected to the center tap of the mod transformer's primary.
Although one might think of the modulator tubes as open circuits (when the filaments are off), there are still many small, parasitic capacitors that need to charge up... and, for a brief period of time, each half of the primary winding may have 6000 volts across it.

3. Above 3000 volts, the electric fields in typical components become so high that there are additional types of breakdown effects that become problematic in those components (i.e. effects that are negligible at lower electric field levels). Things that we normally assume to be capable of handling these high voltages (like connectors) may deteriorate rapidly.

For the above reasons, I would be reluctant to double (almost) the voltage that is applied to the rig... even with the filaments off.

Stu

[WA1GFZ]

Tom,
A 6000V static charge can kill you. I don't like high voltage on things that are turned off for safety reasons. I would put a bleeder in the final just to be safe if there isn't a volt meter on the panel.
I think I have the flange HUZ needs, now if I could just find the next size bigger.

[K1JJ]

Stu,

Thanks for the info in favor of NOT keeping excessive HV on a rig that is turned off. Makes sense.  I will make provisions to manually select the HV for each rig with some kind of failsafe to prevent 6KV on the 4X1 AM rig by accident.

Normally I have the HV supply running all the time once turned on, but I do have a HV TR relay that sends this HV to the various rigs on key-up.  Yes, I don't like HV sitting on a rig, even when it is on but unkeyed.   There's other rainy day issues like, what if a person, animal or whatever situation found its way into the rig or interconnecting HV cable? Or if I had a heart attack and the rig was sitting there hot, even though it is sealed off?  At least with key-up HV the danger time window is greatly reduced and the HV supply is well sealed off.

Frank:  That cabinet will give HUZ a big start on his new project amp. It will make all the difference to doing a first class job in metalwork.

I've found I can build virtually anything with aluminum angle, sheet and 6-32 screws. All my air sub-chassis, vac variable flanges and other framework use the technique.  Very strapping.

Well, did up the 4X1 neutralization wiring and finished the grid circuit and plate tank/ vac variables, etc. Today it's the grid and screen supply wiring and might start on the modulator component wiring. Getting there little by little.  All parts are basically mounted and working mechanically.

Jeff: So that was an HT-37, huh? That's one of my favorite exciters from the past. It's amazing how good the AM can be generated with a rig like that, once you get into the guts. I see you're able to transistion from a big BC rig to the smaller ssb-oriented rigs. Not many guys want to do that. My fix is using an FT-102 modified with an NE-602 bal mod chip. Works DC-light, of course. But doesn't have the soul of a plate modulated dragon...

T

[WA1GFZ]

Cool Tom. Yup angle aluminum is great stuff. I'm close on my layout and even found the spare rear panel I cut years ago. Good thing I'm going to need it.
I plan to do some work after I get something to eat and recover from all the trips between the antenna and rig. I raised and lowered the 40M antenna about a dozen times. Also Got the LPDA coax cable and rotor cable in the shack.

[KC2IFR]

Just a quick note about aluminum chassis..
Welding aluminum is a problem IF u need to. A few years ago there was an infomercial about these welding rods called Alumiloy (spelling). All u had to do was use this stuff just like solder but with a torch. Well the stuff got my attention so I ordered some. For chassis work it works GREAT.
If u have to subdivide a chassis say to isolate the input from the output. this is the stuff for u.
If anyone needs some........let me know.

Bill