Drive requirements for 4cx350a

Discussion in 'General Technical Questions and Answers' started by GW7BBY, Jul 13, 2019 at 3:14 PM.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: Subscribe
ad: Left-2
ad: Left-3
ad: MessiPaoloni-1
  1. GW7BBY

    GW7BBY Ham Member QRZ Page

    I'm working on a rare Marconi Challenger Transmitter. Output is 1.2KW from 4 off 4CX350a's in parallel. As only 12 of these transmitters were sold documentation is hard to find. I have an operator's manual which would normally give the circuit diagrams, however the crucial pages covering the final stages are missing! I have deduced that the valves are shot as I can't set up the initial conditions: 100mA Ia with Vg1 between -20 and -27V. If I attempt to tune up they overheat and the HT trips out (correctly!). I get about 150 - 200Watts of RF before the HT trips. I get no increase in Ia when tuning and the drive, as shown on the meter barely registers. I have a new set of 4CX350s on order, so now I come to the crux of my enquiry - before plugging in these expensive new valves I want to ensure that all other conditions are correct. (HT is 2.2KV, Screens are a solid 400V, G1 Bias, nominally -27v, adjustable to -18v)
    I am particularly confused over the drive requirements. Briefly, the four valves are in parallel in grounded cathode configuration ( I do not have the manufacturer's circuit, I've traced it out manually). There is a synthesizer/Mixer/amplifier stage which works correctly and supplies 6Vpp, CW or SSB out to a Drive amplifier. The drive amplifier then raises supplies the necessary drive for the valves (circuit attached). Despite the 4cx350s being faulty I would have expected to to more drive from this amplifier. The active device is a CTC BAL000010 dual npn transistor in a 744a-01 package. I can't find a datasheet for this transistor, but it is probably similar to MRF392/393 which are 100Watt devices. It is mounted on a very hefty heatsink. The circuit looks like a 100Watt amp, I note that L4 in the 24V collector feed is rated at 3Amps.
    What I don't really understand is that the 4CX350 datasheet states a Peak RF Grid voltage of 21 volts and a Grid current (Max) of 2mA which does not, to me, equate to 100Watts. Assuming a Grid impedence in the order of 1000Ohms only about 4 Watts woiuld be required. Other sources that I have found suggest that only "a whiff of RF" is needed to drive these valves to full power, others seem to suggest that substatially more may be required. I've successfully interfaced solid state and valve amplifiers in the past (much lower power though) and my understanding is that voltage swing is more important for valve amplifers than outright power owing to the high input impedence at G1.
    I hope this is not too much of a ramble, can someone explain, how much voltage is required at the Control Grid, and how much power please?
    Many Thanks - Michael.

    Attached Files:

  2. WB2WIK

    WB2WIK Premium Subscriber QRZ Page

    Just one bad tube of the bunch will screw everything up, and these are very fragile tubes with respect to grid dissipation -- they haven't any! They're voltage driven and intended for AB1 linear operation only.

    Your DC parameters sound fine but we have no idea what the circuit is. How is the driver stage matched to the high input Z of the tubes? Is this an HF bandswitching amplifier, or what is it?
  3. SM0AOM

    SM0AOM Ham Member QRZ Page

    This amplifier looks remarkably similar to the ITT-Standard Radio ST1600 marine transmitter that used a solid-state driver with about 25 W output fed int0 4 parallel 4CX250B.

    Your drive requirements should be around 22-25 V peak over a hundred ohms or so. There probably is a step-up transformer somewhere between the transistor driver and the control grids, and the schematic shows a 1:2 turns ratio autotransformer between the driver collectors and the PA grids, so it should be possible to get about 40V peak without any greater problems.

    For all practical purposes, the drive power is the peak-value squared divided with twice the grid resistor value or about 625/200 or 3 watts. Losses have to be added to this.

    4CX350A:s are very temperamental, and often suffer from grid emission due to deposition of cathode material on the grid wires. Vacuum can also suffer from this reason. Both causes show up as DC operating point instability.

    One must be very careful with current limiting of especially the grids, as they are rather fragile.

    The reason for the low drive indication could also be a damaged drive amplifier, as the drive often is monitored by an RF voltmeter directly at the grids.

    My suggestion is to connect an oscilloscope with 1:10 probe directly across the grid circuit with the 4CX350A;s removed, and watch if the drive can be increased to about 25 V peak without problems.

    You can DC test the 4CX350A:s one-by-one in the amplifier if you watch out for overheating due to reduced air flow and keep the testing periods very short.

    If you can locate an AVO Mk V Valve Tester, it is possible to test the valves individually under static DC conditions. Wear-out phenomena show up immediately.

    Last edited: Jul 13, 2019 at 5:20 PM
  4. GW7BBY

    GW7BBY Ham Member QRZ Page

    Thanks both for your prompt replies - very helpful. Interesting that the drive amplifier is similar to the ITT ST1600 as , of course, the Marconi Challenger is a marine transmitter.
    WB2WIK: The attached circuit in my first post is for the Driver Amplifier, I don't have the manufacturer's circuit for the Finals, but I've attached a copy of my manually traced circuit (Finals003.jpg). I'm assuming tyhe impedence matching is done by a broadband matching transformer (T1) on the Drive amplifier Circuit as the output, with the -27V bias impressed on it is passed via pin4 at the bottom of the diagram directly to the the 4CX350 grids. The feedpoint being the centre of the "X" on the attached picture of the underside of the valve box (P1100809.jpg).
    Karl-Arne: I'll be working on it again on Monday, if I can physically get to the grid circuit I will monitor it with a scope. Although, a monitoring hole is provided, it's still a bit difficult to access so I might have to remove the valve box with the drive amp inside and run it up on the bench. Your comments about the drive requirements accord with my gut feeling - I've seen some contradictory suggestions which undermined my confidence a bit!
    Damaged drive amplifier is one of my concerns. Although the active device, BAL0000L10 seems hard to get and possibly expensive, it wouldn't be hard to replce the whole drive amplifier with something different if necessary. We heven't reached that point yet.
    It had crossed my mind that it would be possible to test the valves one by one in the amplifier - I would have to take great care to remove the unused anode clamps owing to the danger of them contacting something they shouldn't! I might know someone with an AVO MkV, we only have an old Taylor Valve tester. One of the valves does show considerable discolouration due to heating, the others seem OK in that respect so I might be lucky.
    Other negative indications for the Drive Amplifier are that resistors in the Pi matching network/attenuator at the input were burned out - I've now replaced them.
    I'll report my findings on Monday after working on the transmitter.
    Thanks both for your comments. - Michael.

    Attached Files:

  5. KM1H

    KM1H Ham Member QRZ Page

    Ive used the Svetlana GS-36B, aka 4CX400A, which is not a real EIA registered number, in several 4CX250B, 4CX350A, 8930, and same based tetrode applications with excellent results including a AN-6155 on 70 cm, and are a very rugged and reliable tube. With a socket change they adapt well into various 8122 amps.

    With 4 tubes in parallel is there any provision for individually adjusting grid voltages or must matched tubes be used?
    Note that matching often does not hold over heavy use and could cause problems including failures.

  6. KM1H

    KM1H Ham Member QRZ Page

    Do those tubes use the Eimac sockets with built in screen bypass caps? I do not see any parasitic suppressors in the anode leads but the low marine frequencies used may not have required them. Ive little experience with commercial marine amps, mostly military and ham/commercial in the 1.8-30 MHz range.

    Is that a burn mark at the bottom of the anode choke?

  7. SM0AOM

    SM0AOM Ham Member QRZ Page

    Most transmitters this era used the Eimac SK-620 air-system socket, and VHF parasitics were kept in check by some well-placed ferrites and decoupling resistors.

    Below is the "innards" of the contemporary ITT-Standard Radio ST1610 servo-tuned marine transmitter, using 4 4CX250B finals.


    This particular transmitter did not use neutralisation and RF negative feedback in order to simplify the circuit and permit servo-tuning.

    Instead there were liberal amounts of resistive grid-swamping.

    The manually tuned predecessors, the ST1200 and ST1400, used tube driver stages and bridge neutralisation. Also some RF negative feedback was applied.

    All final tubes had individual bias adjustments, but many R/O:s and field service engineers felt that there were reliability problems, especially toward the end of their life-time when tube quality had gone down-hill.

    The whole ST1200/1400/1600 series transmitters were closely related to the servo-tuned CT1000 transmitter that had the same tube complement and was rated for ISB operation.
    Conservative ratings and a lot of negative feedback kept the first adjacent channel suppression
    to better than -45 dB.

    SRT CT1000 1 kW KV sandare.jpg

    The CT1000 saw limited civilian use on some ocean liners such as the Gripsholm and Queen Elisabeth II as the ST1430 remote controlled HF transmitter.

    CT1000 was a "staple" for a quite long time in the Army and Navy as a fixed transmitter. The last were decommissioned around 1995.

    This transmitter also became the center of an export deal with Turkey in the late 60s, where the intermediary went insolvent and finally paid off the contract sum with carpets. ITT-Standard Radio became a quite large actor on the "genuine carpet" market for some time.

  8. GW7BBY

    GW7BBY Ham Member QRZ Page

    Hi Carl, Just a quick look on the net; GS36A may be another option there are some cheap used ones around that could be useful for test purposes. I'll have to take a closer look at the spec.
    No, it's not a burn mark, its a splodge of black marker from somewhere. There is no provision to indivdually tailor the G1 bias: the set up procedure is to set the bias for a total standing anode current of 400mA (100mA each) then to individually adjust the screen voltages to bring the individual anode currents within a % of each other (have to look up the number). The book says to check these monthly.
    As far as I know the bases include screen bypass caps and there are no parasitic supressors for the anodes.
    Many Thanks - Michael.
  9. GW7BBY

    GW7BBY Ham Member QRZ Page

    That's a beast of a transmitter. By way of background inormation I've attached a picture of the radio installation with the Challenger on the right. I look after the Radio collection at a museum in West Wales UK called Internal Fire - Museum of Power. The main core of the museum is concerned with Internal Combustion engines mainly diesel, but over the last two years Steam engines have also been added - all are in working order and are run on a daily basis. Over the last 5 years or so we have established a Marine Radio collection consisting of a 1980's ship's radio room represented by the attached equipment which is the complete radio room from a ship called the "Pride of Calais", we have a representative 1950's radio room in full working order which I use regularly: Oceanspan transmitter, Atalanta and Mercury Receivers and a 1970's Radio room from aRoyal Fleet Auxilary tanker:"Orangeleaf" plus many more interesting bits see for more info on radios, or for the main museum site. Hope this isn't too off topic!

    Attached Files:

  10. GW7BBY

    GW7BBY Ham Member QRZ Page

    I've spent the day with the Challenger and I've also been doing my homework: studying the 4CX350 datasheets. I think I now have a better understanding of the requirments for this valve. Given that the grid drive requirements are, 1) Zero Watts and 2) it must be operated within the negative grid region and 3) a nominal Grid bias of -27VDC, it is understandable that the positive going input voltage swing should not exceed 27V and that it is adviseable to allow a margin of safetly by keeping the upward swing to about 20 - 22 V giving a peak to peak voltage swing of 40 - 44 V.,Thanks Karl-Arne for pointing me in this direction.

    As Karl-Arne suggested I took a series of readings of the drive amplifier output voltage from the feedpoint to the grids of the four valves.

    Test conditions: all tests done on CW mode, HT off, Valves in place, probe x10, all bands from 405KHz to 25MHz, Power levels: Low, Medium, High and Max.

    Results: Volts peak to peak; 8Vpp on low at the lower frequencies to 35Vpp at Max power on 25MHz.Below 4MHz at Maximum power 15Vpp was produced, now I know that the transmitter is restricted to 400Wats maximum below 4MHz and when I plot the 15Vpp against the characteristic curve for the 4CX350 this should resut in Ia of 250mA which seems reasonable.

    The maximum voltage developed in the 4 - 16MHz region is still only 15 - 18Vpp, I would have expected more, but at 22 and 25MHz the output goes up dramatically to 28 and 36Vpp respectively. Again plotting 35Vpp grid voltage on the characteristic curve gives an Ia of about 500mA which seems reasonable.

    The results were the same with the valves removed, applying HT with the valves in place did not affect the results.

    ALC action may come into play when the valves are functioning correctly, although I'm not convinced of this. The other consideration is that 1.5KW (pep) is specified for SSB so a somewhat greater input voltage swing may be expected for SSB. The limited documentation I have does not specify the output for CW as a separate figure.

    Although, I am suprised at some of the low readings at maximum power, the fact that the drive amplifier is capable af achieving a range from 8Vpp to 35Vpp suggests that there is not much wrong with it.

    If you good people on the forum can see anything wrong with my thinking here I would be only too pleased to be corrected.
    One thing I find confusing is why does the drive amplifier have to be so powerful (50 - 100Watts) to increase 6Vpp at its input to 35Vpp at its output when its only objective is to feed a high impedence load requiring zero Watts and next to no grid current?


Share This Page