Mod Reactor Inductance

[n4wc]

Is there a formula for determining the value (henrys) of a mod reactor for a modified Heising modulation system.  I've read that 10 henrys per 1000 ohm plate load.  Is that right?  What's the smallest reactor one can go to.  The Globe Scout 65() (I know it's real Heising) uses maybe 5 henrys.  What is the "window" min to max?

Bill   N4WC

[k4kyv]

Is there a formula for determining the value (henrys) of a mod reactor for a modified Heising modulation system.  I've read that 10 henrys per 1000 ohm plate load.

For broadcast quality, the minimum is roughly 8 henries per 1000 ohms of plate load.  But that's for down to 50~ flat response.  If you are satisfied with less low end, you can proportionally reduce the inductance.  For example, if you are concerned only with having it flat down to 100~, 4 ohms per 1000 ohms will do.  If you are satisfied with space-shuttle audio that goes down only to 300~, then 1.33 henries per 1000 ohms would work.

But according to UTC, per their LS-series catalogue, you should design your audio for flat response at least one octave above and below the intended frequency range, to avoid objectionable phase shift distortion.  So, for example, you wish to transmit a 100~ to 5000~ range, the system should still be capable of 50-10,000 Hz.  For the space shuttle sound but with minimal distortion from 300-3000 Hz, you should design the audio at minimum for 150-6000 Hz.

That explains why some of the older broadcast transmitters from the 30's, particularly RCA's, had such incredible low-end response, to well below 30~.

The later Gates BC-1 series used 32-henry mod reactors, while the older BC-1F used 50 Henries, and possibly on some of the early ones, a huge Thordarson Tru-Fidelity reactor rated at 100 Henries.  They all run somewhere about 2600 volts @ 600 mills.

[WA1HZK]

Keep in mind a string of power supply chokes does just fine also. I mount my Mod Iron and Reactor iron on nice dry plywwood. Keep the cores away from ground. This is a good place to put the 2-4 MFD high voltage heising cap also. Keeps the flow of electrons going through the tube instead of the iron cores to ground. This is especially true when re-using older than dirt BC iron.
Keith
WA1HZK

[AB2EZ]

I few things to add to the comments already provided:

1. In a modified Heising configuration (where you are using a mod transformer and a Heising reactor) the low frequency performance is determined by the parallel combination of the Heising reactor and the "magnetizing inductance" of the mod transformer (as viewed from the secondary side of the transformer).

For example, the mod transformer in the Ranger has a magnetizing inductance of ~ 8H

If you use a Heising reactor with an inductance of 20H, then the total inductance will be 1/[(1/8) + (1/20)] = 5.71H

If you use a Heising reactor with an inductance of 50H, the the total inductance will be 1/[(1/8) + (1/50)] = 6.9H.

So... the magnetizing inductance of the modulation transformer is going to limit the low frequency response even if you make the Heising reactor very large.

In my Ranger, I usually use a 20H Heising reactor in combination with a backward connected external toroidal ferrite filament transformer (costs $10.00) acting as a modulation transformer. The magnetizing inductance of the ferrite transformer is so high that the low frequency performance of the modulator is totally controlled by the 20H Heising reactor.

2. For optimal low end frequency response (flatness down to the frequency at which the response starts to roll off), the Heising capacitor should be chosen to be a value that has a capacitive reactance (1/2pifC) that is equal to the inductive reactance of the Heising reactor at the frequency f where the inductive reactance equals the modulation resistance.

Example 1: Ranger with its stock modulation transformer

Plate modulation resistance: 600 volts / 120 ma = 5000 ohms

Effective Heising reactor inductance (parallel combination of a 20H Heising reactor and the 8H magnetizing inductance of the stock Ranger mod transformer) = 5.71H

Frequency at which the inductive reactance of the effective Heising inductor equals the modulation resistance:  2 x pi x f x 5.71 = 5000 ohms. Therefore f = 139 Hz

Best value to the Heising capacitor (produces a flat frequency response down to 139 Hz):
1/(2 x pi x 139 x C) = 5000 ohms.  Therefore, C = 0.23 uF

Note: Although somewhat non-intuitive, using a larger value for C will make the frequency response roll off more at low frequencies (at least for frequencies above 139 Hz); and can also result in a resonance at very low frequencies that places a very large voltage across the secondary of the mod transformer and across the Heising reactor if you happen to apply significant audio at that resonant frequency.

Example 2: Ranger with 20H Heising reactor and backward connected ferrite toroidal filament transformer used as a modulation transformer

Plate modulation resistance: 600 volts / 120 ma = 5000 ohms

Effective Heising reactor inductance = 20H

Frequency at which the inductive reactance of the effective Heising inductor equals the modulation resistance:  2 x pi x f x 20 = 5000 ohms. Therefore f = 40Hz

Best value to the Heising capacitor (produces a flat frequency response down to 40 Hz):
1/(2 x pi x 40 x C) = 5000 ohms.  Therefore, C = 0.8 uF

Best regards
Stu
 

[Jim KF2SY]

backward connected external toroidal ferrite filament transformer

Hello Stu,
Do you have a link for this company?  Could you just use one of these toroidal xfmrs as the mod. reactor choke?
Wonder what is the inductance is ?  For the price and the small size, sounds interesting....

Thanks,
Jim

[AB2EZ]

Jim

I purchased mine from Antek

http://www.antekinc.com/

These transformers behave as almost ideal transformers... but they cannot support any significant amount of unbalanced DC. The ferrite core has a very high permeability... which leads to the almost ideal performance. But any significant amount of unbalanced DC will saturate the core.

Therefore, they cannot be used as a Heising reactor or in any other application that involves unbalanced DC.

For modulating my Ranger, in Heising configuration, I use a 230 volt input [two (2) 115 volt windings in series] : 12 volt output [two (2) 12 volt windings in parallel] filament transformer as a 19:1 step up transformer. With the 5000 ohm load of the Ranger, the input impedance is 5000 / [ 19 x 19 ] = 13.9 ohms. I drive the transformer's input with a modern 60 watt audio amplifier. The amplifier has plenty of extra output power capability... so driving the transformer's 13.9 ohm input impedance is no problem.

For plate modulating my home brew legal limit amplifier, in Heising configuration,  I use larger versions of these transformers (rated at 1kVA) to obtain a 25:1 step up ratio. I used two off-the-shelf Antek transformers in series:   a) 63 volts : 115 volts  b) 115 volts : 1600 volts. Even with two transformers in series, I obtain almost ideal modulation performance over the frequency range 30 Hz to 10 kHz. I use a 50H 300ma Heising choke in this application, and I use a 600 watt audio amplifier to drive the transformer's input.

Note: very high voltages... extreme caution and appropriate expertise is required to do these things safely. 

Best regards
Stu

[Steve - WB3HUZ]

Note: Although somewhat non-intuitive, using a larger value for C will make the frequency response roll off more at low frequencies (at least for frequencies above 139 Hz); and can also result in a resonance at very low frequencies that places a very large voltage across the secondary of the mod transformer and across the Heising reactor if you happen to apply significant audio at that resonant frequency.

Not only that, it will cause overshoot in the time domain which will cause excessive modulation peaks. This will tend to defeat any processing to control peaks used in the lower level audio stages. There's a reason why most BC rigs used a 1 or 2 uF blocking cap. 

[n4wc]

I wonder what it would take to wind a powdered iron toroid 30 henry choke......
I assume powdered iron would handle the DC. 
I'm thinking of modulating a 50 watt input rig with either heising or modified heising and amplifying through a linear for legal limit.  This is a resource limited project.
I really appreciate the knowledge y'all have.
Thanks
Bill   N4WC

[k4kyv]

Another measure you can take with the Ranger mod xfmr to increase the inductance of the windings is to cross-laminate it.  If you don't run DC through it, there is no need for the gap in the core.  Dismantle it and restack the E and I laminations in the core like a power transformer, and that will drastically improve the low end response.

I did that with the mod xfmr in my Eico 730, and added a 40 hy reactor, using 1 or 2 mfd of coupling (based on the capacitor that was available, rather than by calculations), and it is flat to below 30~.

One problem with using power supply chokes for mod reactors is that the iron in the core may become lossy at higher audio frequencies, > 1 or 2 kHz, due to losses in the laminations, which are often thicker than those used for audio iron.  Also, some power supply iron is inherently lossy at higher frequencies, and the coil may be wound in such a way that the capacitance of the winding acts like a capacitor shunted across the coil, since high frequency audio response was not a concern in the  design of the coil. 

But if you have a good power supply choke of proper inductance, not to worry.  Place a second smaller reactor of much lower inductance (1 Hy should be plenty), but rated at the class C PA current, in series with the larger one, between the main reactor and the class C final.  At higher frequencies when the main reactor begins to drop off, the second smaller reactor will still provide plenty of inductance.  If you cannot find a small choke at the  current rating that works, try removing the core  from a spare filter choke and let it function as an air-core inductor, or take a crapped out modulation xfmr (a 100-watt unit should be sufficient for modulating over a kw), dismantle it, and wind a low inductance coil over the core.  But chances are about any 1-2 Hy inductor @ the rated plate current will work well.

[flintstone mop]

I thought Don was going to mention the fact about series connecting reactors.
This will also drop the B+ vdc going to the final P.A. If you can tolerate some drop in RF output power.

Fred

[KL7OF]

I have an open frame mod tx from a BC1-G gates 1 kw rig that has a short in the primary winding.....Any comments about using the secondary winding as a mod reactor???This transformer also has a winding on the secondary for modulating the driver tubes (2x807).....Steve

[Opcom]

get rid of the short!