T-368 audio driver question

[W9BHI]

I would like to drive the grids of the 4-125's using a small
solid state hifi amplifier.
I guess I would need a matching transformer of some sort.
Something like an output transformer connected up backwards?
What impedance primary for the grids?
Power rating?
How many watts would the amp have to produce to drive the the grids to get 100% modulation?
Inquiring minds want to know.
Thanks,
Don W9BHI

[w4bfs]

hi Don .... took a look at the modulator ckt .... looks like the 4-125s are op AB1 given the screen regs and high value bias pot .... shouldn't have to gen power just V swing .... did you look at the T368 mods listed in the Transmitter section ? .... looks like the Tron has wrestled with the problem .... can you imagine Tim in the WWF ?   take it to the mat (belch)

[AB2EZ]

Use the following Antek toroidal transformer ($11.00 + shipping) http://www.antekinc.com/pdf/AN-0206.pdf as a voltage step up transformer.

I have used these types of transformers in audio step up applications, in conjunction with 8 ohm output audio amplifiers, with great success. They behave as almost ideal transformers, with a very high magnetizing inductance, and a frequency response that extends from below 30Hz to above 7500 Hz.

Feed the 6V windings in parallel with the 8 ohm output of your audio amplifier.

Place the 115V windings in series (make sure to set the polarity correct: connect + to - at the midpoint). Connect one end to each 4-125A grid. Feed the -96V (or whatever it is) bias to the grids of the 4-125As via the midpoint of the two 115V windings.

This is a 6 : 230 turns ratio transformer, and it will provide a voltage step up of 230/6 = 38.3

Place a 12000 ohm 5W (or higher wattage) resistor (wire wound is fine) from one of the 4-125A grids to the other 4-125A grid. This will present (via the transformer) an 8.2 ohm load to the audio amplifier. Strictly speaking, this load resistor might not be needed... but some audio amplifiers will not work properly if they don't have a load whose value is in the ballpark of the impedance of the speaker they were designed to drive.

Since you need a peak grid-to-grid voltage of 192V to drive the 4-125A grids, you need a peak voltage across the primary of the transformer of 192V/38.3 = 5 volts.

A sine wave having 5 volts amplitude across an 8.2 ohm load corresponds to 1.5 watts of average power or 3 watts of peak power.

Therefore, your audio amplifier will have to deliver 3 watts of peak power to its 8 ohm output terminals.

Good luck!

Stu

  

[AB2EZ]

Don

In a recent update you wrote:

"I have been caught up in several projects. The modulator driver using the Antek toroid worked ok, but it needed much more than the 3 or 4 watts calculated to drive the 4-125's.

Thanks again,
Don W9BHI"

Since the step-up transformer approach is an alternative that others may be interested in... can you provide a few more details? It would be helpful to determine why this didn't work for you as expected.

E.g.:

Are you sure that the parallel primary windings were connected in the correct phase, and that the series secondary windings were connected in the correct phase?

How much audio power from the output of the audio amplifier to the input of the Antek transformer was required to drive the 4-125s?

Did you measure the peak audio voltage at the output of the audio amplifier, and the peak to peak audio voltage from grid-to-grid of the 4-125s, when they were being driven to fully modulate the transmitter?

Was the -96V bias for the grids connected to the midpoint of the two output windings of the Antek transformer (and not to either of the grids by any other paths)?

Thanks
Stu

[W9BHI]

Hello Stu,
The Antek toroid was connected as follows:
Top secondary green to bottom secondary green.
Top secondary blue to bottom secondary blue.
The audio amp output was connected to the green and blue pairs.
The top primary red was connected to one 4-125 grid.
The top primary black is connected to the bottom primary red and those two were connected to the -96 volt bias.
The bottom primary black was connected to the grid of the other 4-125.
The 12k 10 watt resistor was placed across the 4-125 grids.
I drove this with one channel of the Symetrix 20 watt per channel amplifier.
I noticed that in order to drive the modulator to 100% modulation, the clip indicator was sometimes flashing on the Symetrix amp.
I would guess that the amp was running out of steam at that point.
I did mention in my post that this setup did work but took more power to run.
BTW I did not put a scope on the output of the transformer to see what the actual PP voltage was,
I only have a 10:1 probe.

Don W9BHI

[AB2EZ]

Don

okay... thanks for the details.

I sounds to me like either: there is an additional load from grid-to-grid caused by an additional path (not including the transformer) from each grid to ground; or the audio amplifier is not working properly (the clip light comes on, but the output is only a few watts); or the transmitter is designed to require the grid of each 4-125 modulator tube to be driven positive with respect to the cathode.

The 4-125 specification sheet includes data for typical AB1 and AB2 operation.

In AB1, the grid bias is -96V, and the peak grid-to-grid voltage is 192V. The audio driving power is (of course) 0 watts, and the maximum signal audio output power is 320W.

In AB2, the grid bias is -42V, and the peak grid-to-grid voltage is 178V (i.e. less than in AB1 operation). The maximum signal audio driving power is 2.4W, and the maximum signal audio output power is 400W

In either case, the total drive power required, with a 12 ohm grid-to-grid resistor, would be much less than what the amplifier should be capable of providing.

In AB2, there might be some saturation effects from the transformer ... due to unbalanced DC in the secondary (i.e. if one tube draws more average grid current than the other)... but that would require a scope to observe. If there were significant transformer saturation, that might cause the required drive power to increase (v. no saturation), and the output modulation envelope would be distorted.

Best regards
Stu

[AB2EZ]

Don

I can think of another reason why you may be seeing a problem with this configuration:

The negative grid bias supply for the 4-125 modulator tubes is probably a pretty standard design... i.e. a diode-based rectifier feeding a capacitor.

If the 4-125s are intended to operate in class AB2... then whenever grid current flows into either tube from the negative bias supply... the capacitor at the output of the negative bias supply will charge up; and the voltage at the output of the grid bias supply will become more negative. This will increase the grid drive voltage required to obtain full output from the modulator.

To fix this problem, it is necessary to place an appropriate resistive load on the negative bias supply (from the point where the negative grid bias supply connects to the Antek transformer's secondary to ground) in order to bleed off the excess negative voltage. This resistor will ensure that the average current that the negative bias supply is delivering is always in the direction that will discharge the supply's output capacitor (i.e. making it less negative).

Something like a 3300 ohm 5W bleeder resistor should work. A wire wound resistor is okay. This resistor will not draw any power from the audio amplifier. All of the power dissipated in this resistor (a few watts) will come from the negative grid bias supply.

Stu

[WD5JKO]

Perhaps the audio amplifier has a slight DC offset to it, and the load being close to zero ohms DC will use up some or a lot of the amplifiers current capability budget, i.e 200mv dc offset into 0.1 ohms is 2 amps. An 8 ohm loud speaker as a load has nearly 8 ohms DC resistance, so a slight DC offset with a speaker load has a negligible effect. The primary windings of that Anetek transformer are pretty low resistance at DC.

Also, if the programming audio has an asymmetry to it, then there will be a DC component that the amp must then take on the burden of handling when the load is close to zero ohms DC.

One way around this is to have a series capacitor with the primary of the transformer. It will need to be big, and the value makes a high pass filter. I have used big electrolytics in series, like 10,000mfd @ 50V connected like this:

--------+| |--------| | +-------------

I've used diodes across each capacitor (cathode to + of cap), and without diodes...Made little difference.

With the cap there is a place for any DC component on the output to go without using up the amplifiers ampere budget.

Jim
WD5JKO

[AB2EZ]

Jim

This is a very good point. In addition, these transformers saturate very easily if there is unbalanced DC in the windings. The DC resistance on the low voltage side of one of these transformers will be a fraction of an ohm (e.g. about 0.3 ohms of series resistance in the 12 volt winding of a 12V 50 watt transformer... assuming 10% resistive losses when the current is 4A)

I have always connected the audio step up transformers I use directly across the amplifier output... but, then again, my audio amplifiers have "servo balanced" circuitry to null out the DC offset at the output of each amplifier, before a relay kicks in to connect the associated output load. I measured the DC voltage at the output of a 300 watt audio amplifier that I use. It was about 0.009V. The DC current into a winding with 0.3 ohms of resistance would be about 30mA.

Based on your comment... I have added a 1 ohm 10W resistor in series with the output of the amplifier, to reduce the DC current into the transformer's low voltage winding to around 7mA. I measured the voltage across the 1 ohm series resistor, and it was 0.007V, as expected. It may be my imagination, but I think the envelope of my RF signal is even more closely tracking the audio input to the modulator with this added resistor in place. Since my transmitter (without this added resistor in series) presents an audio frequency load to the amplifier of around 11.3 ohms, this extra resistor adds only around 8% to the power that the audio amplifier has to deliver.

Separately, although the input my audio amplifier is intentionally asymmetrical (bigger positive voice peaks), there is no average (DC) component. One reason for this is that my audio chain's low impedance, balanced output is AC-coupled to my amplifier's unbalanced, high impedance input, using a small 600 ohm : 150 ohm transformer. [There is no significance to this 4:1 impedance transformation. I'm taking the output of this small 600 ohm : 600 ohm transformer from the center tap to one end... in order to obtain two simultaneous outputs that are 180 degrees out of phase with each other]. The transformer has no significant effect on the audio frequency response for frequencies above 30Hz, and the associated 2:1 step down ratio is easily accommodated by the audio amplifier.

It would be an interesting experiment for Don to add a 4 ohm 10W resistor in series with the transformer's primary... just to see if it makes a big difference in the behavior of the amplifier. The 4 ohm resistor would only increase the required amplifier output power from around 3 watts to around 4.5 watts, if everything else is working as expected.

Stu