Broadcast Audio limiter/compressor in audio chain for newbie

Discussion in 'Amplitude Modulation' started by KD6CXW, Jul 28, 2019.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: MessiPaoloni-1
ad: K5AB-Elect-1
ad: Subscribe
ad: L-MFJ
  1. K4KYV

    K4KYV Subscriber QRZ Page

    Those have pretty much gone the way of National Grid Grips, EFJ edgewound coil clips, real Astatic D-104 crystal elements, and 2-watt carbon composition resistors. The most likely source is from parted out broadcast transmitters (RA-1000, BC-1E, BC-1F, etc) or maybe some old time ham's junkbox or attic.
     
  2. W8KHK

    W8KHK Ham Member QRZ Page

    Yes, that heavy metal hardware is pretty much unobtanium. It is not likely that there will be any suitable after-market products available, due to the low demand. And we probably do not all wish to stray from hollow-state and use the MOS-FET silicon solution.

    To keep the tubes glowing, perhaps we should come up with a transformer-less design for a nice cathode-follower solution that can drive the 833 grids hard without wimping out when the grids go positive. That would take a substantial power supply, and some low-mu tubes that have the excess emission to do the job as well as transformers. I recall the RCA BTA-1M and BTA-1MX transmitters used push-pull 807s from the input transformer to the 833 grids. The first pair provided voltage gain, and the second pair ran as cathode followers driving the class-B 833 modulators. This may not have been the optimal solution, but with careful selection of tubes and power supply voltages, perhaps it would be possible to have a class-B driver solution that equaled, or perhaps surpassed, the transformer coupled driver that used the legendary 845 bottles.
     
    Last edited: Sep 6, 2019
  3. W2VW

    W2VW Ham Member QRZ Page

    I still have the audio driver xfmr from an RCA 1D here. So far all the other iron has been fine which is amazing because it was removed from service in 1967 when it was 30 years old.

    I just fired up a homebrew amp where I used the 1D's autotransformer and it's FB. The modulation reactor is used for the Johnson transmitters here.

    Anyway I was thinking it might be possible to use an off the shelf toroid type power transformer as a substitute. Having the original makes comparison a lot simpler.

     
  4. K4KYV

    K4KYV Subscriber QRZ Page

    The audio driver is the weak link in the chain of a class-B modulator. First of all, the source of audio driving power needs to have a very low internal impedance, since the grids present a widely varying impedance load to the driver, from infinity down to around 500 ohms over each audio cycle. If the internal driver impedance isn't low, the result is unacceptable distortion and splatter.

    That means the driver tubes need to have a low plate impedance, achievable either with low mu, low plate resistance triodes like the 2A3 and 845, or with screen grid tubes like the 6L6 family, with a lot of negative feedback around them.

    The driver transformer itself needs to be of special construction, so that the coupling between primary and secondary is as tight as practicable and leakage inductance as low as possible. This is usually achieved by interlacing the primary and secondary windings; a few layers of primary, overwound by a few layers of secondary, which is then overwound by a few more layers of primary, until the windings are complete, in other words as close as possible to bifilar winding.

    The turns ratio should have as much step-down as possible while still delivering enough voltage swing to fully drive the grids of the modulators to just shy of saturation (flat-topping). Increasing the step-down ratio is better than placing a swamping resistor across the secondary to maintain a more constant load, although the swamping resistor may help if you have no way to further increase the step-down but still have an excess of driving power available.

    Running two audio output transformers back-to-back, linked together with the 8-ohm speaker windings usually doesn't work very well, because audio output transformers normally don't have the interlaced windings like a properly designed class-B driver transformer (although some do), and using two transformers back-to-back leaves a lot of slack in the coupling between driver plates and modulator grids. Back-to-back coupling with 500/600 ohm windings works a little better than using 8-ohm speaker windings, although this is still not the ideal solution even though some manufacturers have made class-B driver transformers, 500 ohms line to push-pull class-B grids, just for this purpose.

    Assuming the grid impedance of each class-B triode to drop to 500 ohms at the crest of the respective audio peak, means that the transformer should be designed for 500 ohms each side of midtap, or 2000 ohms total winding impedance, with the primary impedance to match the recommended plate-to-plate load impedance of the driver tubes. Class-B driver transformers are usually rated at total primary to 1/2 secondary, since only half the winding is actively in use an any one time. A transformer designed to match a pair of 845s to class-B grids usually has around a 5:1 primary to 1/2 secondary turns ratio.

    My experience with the Gates BC1-T 807 cathode-follower driver was less than spectacular. Beside the problem that it was flaky, constantly requiring troubleshooting and having the 2-watt carbon comp resistors burn brown spots or even holes in the circuit board (vacuum tubes on printed circuit boards was never a good idea in the first place), the thing didn't fully drive the 833-As. When I replaced mine with the audio driver section from a Raytheon RA-1000, which uses a pair of 845s transformer coupled to 833A grids, I was surprised at how much the undistorted positive modulation peak capability was improved.

    The 807 cathode follower circuit should be OK for a 250-watt transmitter, but at a full 1000 watts in broadcast service, something more robust would be better. Maybe a pair of 828s with around 1200 volts on the plates.

    If you can find a proper driver transformer but not the 845s, a pair of 211s will work reasonably well until you can find some 845s. All you have to do is change the bias voltage. Chinese-made 845s are available for the audiophool market at a fairly reasonable price, about $150 for a matched pair last time I checked. Grant, W4BVT tried a pair of Chinese 845s in his rig and said they worked just as well as his n.o.s. RCAs, but no guarantee they would last as long as the RCAs.

    A pair of 6L6s with adequate negative feedback work about as well as a pair of 2A3s or 6B4Gs without feedback.
     
    W2NBC, WZ5Q, K5UJ and 2 others like this.
  5. N2DTS

    N2DTS Ham Member QRZ Page

    Or just do AB1.
    I tried to talk Don into dumping all that class B junk and going AB1 last night on 40 meters, but he had to run to dinner.
    I think a number of fine broadcast transmitters used various tubes in AB1, was it a power limitation that had class B so popular?
    For all the talk about critical iron and low mu tubes, burnt resistors and wimpy drive, you would think AB1 would be top choice....
     
  6. W8KHK

    W8KHK Ham Member QRZ Page

    It is probably not possible to get any reasonable power from a triode modulator in AB1. AB2 and B are more efficient and yield much greater power output with triodes. On the other hand, tetrodes and pentodes do very well in AB1, and, from what I have seen, they can be cleaner when they do not run grid current.

    A couple examples of broadcast transmitters in the 1 KW class running AB1 modulators are the Collins 20V-3 and the RCA BTA-1R.
    RCA uses a pair of 2E26 tubes as voltage amplifiers, while the Collins uses a pair of 6SJ7s. I am sure the 6SJ7 would do well in the RCA circuit, but they probably chose the 2E26 to reduce the number of spare tube types, as the 2E26 is also used in the RF section. Both transmitters run a pair of 4-400A finals, modulated by a pair. It is impressive that only one stage of amplification is needed between the input transformer and the modulator grid!

    The tetrode transmitting tube came along after the triode, so many of the earlier transmitter versions used triodes even after screen-grid tubes were available. Of course, if running the tetrode, you need a regulated screen supply and additional protection circuitry. This is probably less expensive than the driver for a class B triode, especially when done properly. There is probably very little incentive or reason to switch from triodes to tetrodes or pentodes when it works fine and sounds great, unless the smoke leaks out of an unobtanium component. That is of course why we have a junk box of shelves with all that iron organized for that unthinkable event.
     
    K5UJ likes this.
  7. K5UJ

    K5UJ Ham Member QRZ Page

    Class B is awesome if you have the iron; AB1 gets the job done but oh man, you got to have the bottles that can deal with the heat (304TLs) or you need some serious cooling. AB1 is great in the dead of winter :)
     
    W8KHK likes this.
  8. W8KHK

    W8KHK Ham Member QRZ Page

    Yup! But the additional dissipation only heats the shack while you are talking. Need an old buzzard transmission to run the rig as a shack heater. I wonder how much heat is generated by the several stages of amplification needed to drive the class-B tubes, including tube heaters and the power supplies? And they typically run all the time, not just during transmissions. This would be good in the winter too.

    I sure hope I have my Class-A series modulator working before the freezing weather sets in. Only one modulator tube to heat the shack, and no transformers. Talk about efficiency - it is 100 percent efficient at heating the shack. Even the 400 watt filament of the triode contributes to the warm cozy atmosphere.
     
  9. N2DTS

    N2DTS Ham Member QRZ Page

    Well, I am thinking screen grid tubes in AB1, like the 4x150/4cx250b tubes will do 600 watts of audio, 2000 volts 90ma resting per tube, 250 peak ma per tube.
    4-250's will do 500 watts of audio in AB1, 4-400's more, and resting power does not seem excessive.
    Output can be quite clean with just a stiff screen supply, or some VR tubes can be used.

    I have built a few class B modulators driven by a solid state amp into a good transformer (UTC LS) and I always thought the
    AB1 stuff sounded cleaner...
     
  10. K4KYV

    K4KYV Subscriber QRZ Page

    For more than about 100 watts I have always preferred triodes in everything. Since from the outset, I sought out the low mu triodes and driver transformers when they were plentiful and cheap at flea markets and hamfests I now have an adequate supply on hand. The stiff bias supply is less hassle to build than a regulated screen supply. The same with the rf final; the neutralisation of the triode is less hassle than supplying voltage to the screen and having to be careful not to exceed safe screen current when tuning up a tetrode or pentode.
     
    WZ5Q and (deleted member) like this.

Share This Page

ad: QSKLLC-1