The AM Forum

I wish I had time to draw this out in a usable form for the forum ... look at this one closely ... it is a push pull plate modulated stage using 813 tubes .... all seems to be orthodox until you look at the plate lead rf bypass cap (usually one lead to plate choke rf cold end, one lead to ground) .... it seems that the bypass cap is going to the screen connection and then to ground for rf but is a coupling cap for audio with low freq break point at 125 Hz

Also note that this 800 watt dc in stage is spec'd to be modulated with 150 w of audio

Whatduya think?

here it is:

C8 should go to ground and the B+ lead to C7 is also wrong. The rotor of C7 should be grounded in this shunt fed circuit.

L3 is there so the screen can self modulate to some degree which is better IMO than a big resistor off the B+.

T2 may be a typo and 450W was meant.

There should also be a RFC or 2-3K carbon resistor right at the input coil CT and bypassed at the cold end.

M1 is reading cathode current so grid and screen currents have to be subtracted to give plate current.

All in all its a pretty half assed circuit as published.

Carl

" T2 may be a typo and 450W was meant. "

Its the same value , 150 Watts, in the TT-4 Manual I've got.

klc

Point us to a full assed circuit please!  ;D

" Point us to a full assed circuit please!  ;D "

Here it is


http://www.amwindow.org/tech/htm/813/813.htm


klc

Yup, that's pretty full assed alright. Thank you very much!

On the 813 modulators, where does the bias come from? I see a -75v supply but the input modulation transformer centertap is grounded. What gives Jeeves?

At first glance it looks like the screen would be grossly overmodulated.  But looking at the values of coupling and by-pass capacitors, the grid is apparently self-modulated by L3, but with additional audio from the plate modulator coupled to the grid through C8, and forming a capacitive voltage divider with C5 and C6. The circuit looks like it would be ineffective at lower audio frequencies, given the low value of C8, so maybe this is to compensate for the high frequency roll-off of screen modulation that would result from the screen by-pass capacitors.

As for the audio power, they are probably talking about average audio power for voice modulation.  It would take a little more than 400 watts of audio to modulate it with a sine wave.  That was an old trick,  recommended by RCA right after WW2, to "voice modulate" a transmitter with smaller, cheaper modulator tubes running over-voltage on the plates.  I recall an ad, claiming a pair of 203-A's could be made to "voice modulate" a full KW DC input.

Kind of like what the leen-yar manufacturers used to do back in the early 60s when slopbucket was just first catching on in a big way, claiming their 2 or 3 sweep tube p.o.s. would run a "FULL KW INPUT (p.e.p.)".

With normal voice modulation, the thing would hit maybe 150 watts input on the plate meter at voice peaks, but if you whistled in the mike, the tubes would go into Chernobyl mode.


C7 and L4 are wired correctly. The full modulated +HV DC is series-fed to the final plates through the tank coil. This allows rf choke L4 to handle very little rf voltage and have negligible effect on the tank circuit tuning, since it is tapped onto the coil at a voltage node. With shunt-feed and rf blocking capacitor, the rf plate choke is always a problem with a multi-band amplifier, since the plate choke almost inevitably has a resonant point somewhere within the frequency range of the transmitter. With series feed, the rotor plates of the variable capacitor are raised to the same +HV potential as the stator, so that no modulated DC appears across the capacitor; the capacitor only has to handle the modulated RF, and much closer spacing can be used than if the rotor and frame of the capacitor were grounded. In the later case, the stator plates carry full +HV potential PLUS full rf voltage, effectively doubling the total peak voltage that appears between the plates of the capacitor.

The 813s in triode connection at 2KV plate voltage become zero bias triodes. No bias supply is needed.

Interesting, thank you.

You're welcome. That's the beauty of this modulator. Simple, easy yet effective and it has plenty of reserve power (headroom). It can be built on a seperate chassis and used to mod a number of different finals.

I have a ton of parts, just not the big modulation transformer. Where does one source this kind of iron?
Also, what rating should be considered when choosing iron? The power supply should be rated for 100% duty cycle or higher?

And the RCA note has another error:

L4 is a three turn link that should be in the center of th L5, not L4.

Phil

Considering an application engineer must have drawn it, knowing the standards I am held to as one, and the standards the RCA engineers would have been held to, it's hard to accept that there's an error in that volume. No one's perfect I guess, not even the old masters! But I do not have an issue with C8. I had something else in mind for that kind of coupling scheme involving not only a capacitor but a resistor and a transformer, but have not tried it.

Nope, C7 rotor should be going to a good RF ground. As shown connected to the B+ it is not going to work since its floating for RF.
If you want to eliminate the RFC then ground C7 rotor thru a cap and use narrower plate spacing.

RCA TX and RX tube manuals have many incorrect circuits.

I don't see how it's floating, Carl.

C7 is basically 2 caps in series in parallel with L5.
A simple parallel resonant circuit. L4 isolates the junction of C7a and C7b from L5 but allows the DC potential to be as Don described.

thanks to all who responded (so far )  .... there may still be more to this circuit than first meets the eye ....

Patrick : ... please check both transformer phasing marks, not in phase as drawn ... circuit is interesting

Don: ... a nice analysis ... if C5,6,8 act as a voltage divider (4:1) then with 2000 V swing out of secondary of mod xfmr would yield a 500 V swing to screen which would have cut off the 813's by a 100 V or so on neg swing ....splatter city .... it is possible that this circuit operates on a lesser secondary swing of say 1500 V which would require a different analysis as to functionality

KC2ZFA : ...sorry don't know your name .... thanks for posting the schematic from the manual

I promise that this was not a rhetorical question about how this works unless Carl is right and it is just imagnary ....73  

Looks like they are counting on C8, C5 and C6 to provide the rf ground to the rotor of C7. Probably not absolutely the best possible rf ground for the capacitor, but you also want to avoid, if at all possible, an additional capacitor from the rotor to ground in order not to have too much capacitance across the modulation transformer. I believe that has been an issue with high frequency audio response in the T-368, which uses series HV feed through the tank coil. Note that all 3 of those caps in the 813 final are 5 and 6 kv micas.  If they are good quality low rf impedance transmitting caps and the wiring is physically laid out properly, using 1/2" or wider copper strap instead of plain wire for all the interconnections between ground, the three caps and the rotor of C7, it should be OK if the leads can be kept reasonably short. Also, EVERY ground connection associated with that stage should be grounded to one and only one common point on the chassis, even if a shorter lead could have been used directly to chassis right where the component is mounted. Grounding C2, for example, to one point on the chassis and C4, C5 and C6 to a different point (or worse still, a multitude of different points) on the chassis is just asking for parasitics, instability and squirreliness in the final stage, since there are inevitably rf impedances and thus small differences in rf potential, at different points along the chassis, and a common rf impedance between plate and grid to ground may exist.  This can result in undesirable rf coupling and therefore feedback between the grid and plate circuits. And every grounding point in the stage should use its own separate lead to that chassis ground; it is important to avoid daisy-chained RF grounds.  

Everything should initially be laid out as symmetrically as possible, but sometimes in a push-pull final, perfect symmetry can actually result in parasitics, so that the leads to one tube may have to be slightly lengthened relative to the other to get the final to settle down.  Another precaution is to make the plate OR grid leads as short as possible, but not both. You should intentionally leave one or the other somewhat longer, to avoid inadvertently turning the final into a VHF TGTP oscillator.

One thing they did do correctly, avoiding a common mistake in homebrew (and some commercial) amplifiers, was to ground the rotor of grid capacitor C2 DIRECTLY to the common chassis grounding point. A by-pass capacitor/ rf choke combination in the grid circuit AND one in the plate circuit is a TGTP oscillator waiting to happen. This does put more voltage across the plates of the grid capacitor, but here you are dealing with the rf grid voltage + grid bias, so capacitor voltage rating is not so much a problem as it is in the plate tank circuit. Also, they are correct in using wirewound resistor R without an RF choke, for the reason just mentioned above. The WW resistor inherently has some rf choking action, and its internal resistance tends to quench any tendency to cause oscillation.

BTW, I have seen push-pull final circuits in which both the tank coil and capacitor were left floating, but that always concerned me that it would be a problem if everything, including the load, were not perfectly balanced.  You can leave the cap rotor floating and ground the midtap of the coil, but I seem to recall reports of instability with that configuration, and the exact midtap point would likely have to be experimentally determined to get perfect balance with the push-pull tubes.  

And equally importantly, you NEVER want to make a direct (or even a poor) rf connection between capacitor rotor and mid tap of the coil.

Sort of on topic...
My interest has been whetted by the inclusion of the 813 schematic by KB2WIG.
I indicated earlier about having a lot of parts. Where can I find the Iron and choke(s)
necessary to make that circuit a reality? That modulation transformer must be hard to find.
Point me in the right direction and I'll start digging.

shouldn't that be for the under-chassis grounds only ?

Say C8 is grounded instead (so that the rotor of C7 is grounded through it as in the standard P-P)...are you saying that the ground end of C8 should pass through the chassis and get connected to the common ground point of C2, C5, and C6 underneath ? I thought that the under-chassis stuff had to be RF tight from the over-chassis stuff. Can you straighten me out Don ?

Thanks for all the info above !

Peter

A good question. I suppose if the chassis base is used as an rf-tight shield, for example with all the grid circuitry underneath and thoroughly isolated from all the plate stuff mounted on top, one could use one under-chassis common point for the grid circuit components and one above-chassis point for all  the plate stuff. I never built a transmitter that way, so it's something I never really thought about. If I did use separate above-chassis and below-chassis common points, I would locate them on opposite sides of the sheet metal at the same spot using the same mounting screw, and run a lead through a near-by adjacent hole from one to the other.

Everything needs to be at the same rf ground potential, and due to the skin effect, the top of the chassis is effectively isolated from the underneath side, rf-wise. A similar example of this is coax cable.  Coax can be thought of as having three-conductors: the inner wire, the interior side of the braided shield, and the exterior side of the braided shield. That's why a W2DU type balun works. The hollow ferrite cores break up the exterior braid into isolated sections at radio frequencies, while the inner conductor and inner side of the braid are unaffected.

In both my homebrew transmitters only one common point per stage is used, and for the one or two grounded components mounted on the opposite side of the metal shelf, the ground lead goes through a hole in the sheet metal to the common point on the other side.

I would pass the ground end of C8 through the chassis and connect it to the common ground point of C2, C5, and C6 underneath, and do the same for the rotor/frame of C2 if the grid tuning capacitor were also mounted on top of the chassis (but for a power-sensitive tube like the 813 I think the grid capacitor should be mounted on the underneath side). Or else mount C8 underneath the chassis and run the ungrounded lead through the chassis to the frame of C7.  The only "RF-hot" components on the top side of the chassis would be the plate leads from the 813s, the stator plates in the tuning capacitor(s), and the ends of the plate tank coil, assuming of course that the plate tank components are mounted on the top side of the chassis.

Hamfests are one place to look. Another would be to place a want ad here and other buy/sell venues. You'll want to get a 250 watt or larger transformer. You can get some model/part numbers at the link below.

http://www.amwindow.org/tech/htm/modtran/modtran.htm

doh! thanks for this reminder  :-[

Peter

Look again carefully and redraw if necessary.

Then tell me what is at the junction of C7, C8 and T2.

While Dons last explanation winds a RF ground path thru C5,6, and 8,  'splain to me how that even comes close to good engineering by placing some level of the plate RF on the screen pins???? I also dont buy into the wandering explanation about grounding and an absolute need for a common point ground at low ham frequencies. Heck, that circuit doesnt even have any parasitic suppressors which limits its frequency anyway to low bands where the 813 parasitic paths are effectively swamped.
The only thing really needed for stable construction/performance is very low inductance grounds as well as RF connections. Any amp is going to have multiple hardware connecting to the chassis top AND bottom assuming the external tooth washers get a good grip; the more the merrier and it will make the complete surface a single RF ground. Ive followed that procedure with a single or a pair of grid driven 4X1's many times for 160-10M and nary a trace of instability.

I also detest an unswamped by a carbon resistor RFC in the grid circuit and always do that when substantial grid current is involved and a 2-3K carbon resistor alone is impractical. With a RFC and a HF effective value bypass the TGTP possibility is slim to none but it still opens the possibility of a LF oscillation which can often go undetected without an SA scan.

And why in hell would anyone use an RFC to the  L5 CT and then feed RF in to the bottom end that has to wander around before finding some sort of compromise ground?

Then please show me any published circuit that does that...on ham bands?? Maybe there are a few but even earlier RCA manuals didnt do it that way.

It looks more like something developed after a bad time at the bar the previous night. ::)

Carl

All I see at the junction of C7, C8 and T2 is the  rf choke L4.  But L4 doesn't place any plate rf on the screen pins.  The other end of the rf choke is connected to the mid-tap of the balanced tank circuit, which is at precisely a cold spot on the coil, rf-wise, so the bottom end of the choke doesn't feed any significant rf to ground.  The purpose of L4 is to carry the DC plate current to the coil, which in turn carries it to the tube plates. Since the tank circuit may not be perfectly balanced, there may be a slight amount of rf voltage at the mid-tap, but the series inductive reactance imposed by L4 and the by-passing effect of C5 and C6 effectively remove any trace of plate rf from the screen pins. The second function of L4 is to isolate the mid-tap of the coil where it is connected to the +HV, from the rotor/frame of the split-stator capacitor. You never want to directly connect the midtap of the coil to the rotor of the capacitor, hence the rf choke.

There is undoubtedly more plate RF on the screen pins via the internal screen-plate capacitance in the tube than from the midtap of the coil, through L4 and C8, on the way to ground through C5 and C6.


Did you ever try to get the parasitics out of a 160m-20m push-pull, cross-neutralised, triode final?


I do find that questionable, and I wonder what year that write-up was done.  I have never been able to build a final for the low ham frequencies without parasitic suppressors that was not squirrelly, whether using beam tetrodes or triodes, link-coupled or pi-network. Art Collins introduced the concept of parasitic suppressors some time about 1934 with one of his production amateur transmitters.  What was so amazing about that rig was how smoothly it tuned up.  The secret was the suppressors, which consisted of the familiar composition resistors with a few turns of wire wound round them.  Up to that point, the suppressors were absent from ham construction (look over the circuits described in the old Handbooks).  Hams just assumed that transmitters were inherently squirrelly to tune up, and put up with the annoyance, doing the best they could with what they had, until the Collins rig proved otherwise.


The TGTP oscillation caused by an rf choke in both the grid and plate curcuit almost always is LF, or near the operating frequency, not VHF. With a push-pull amplifier like the one here, there should be no significant rf grid current through the grid resistor, just DC, since the grid resistor is connected to the mid-point of the balanced grid tank which should be an rf cold spot. With a single-ended grid tank, the cold end of the coil should be by-passed directly to the common ground point with a capacitor, and another by-pass capacitor should be added to the other side of the grid resistor as well, where the lead to the the bias supply and metering circuits is connected.  There is no reason to use an rf choke here in any case.

Maybe in a perfect world but its still poor engineering and you still havent shown me a real life example or circuit in a HB or mag.


Nope. It was either 80-10 with 250TH's and 810's or just 6M with 100TH's. Even at 17 I listened to an OT and it worked well following his suggestions. The 100TH's were years later.


TGTP is near the operating frequency, LF (usually below the BCB) is a loop where most of the energy is in the input path and the output coil acts just as a piece of wire. Often there is grid current and no or very little plate current. Ive helped debug quite a few HB amps over the years that acted that way.



Im rather well aware of that and how tank circuits work and was referring to DC current only which I had thought was obvious. Sorry if I confused you as I tend to assume too much at times and type fast.


No argument from me but a few thick headed types keep insisting on using them "since if its in the book its gotta be a perfect design". I often hear TGTP artifacts on the bands popping in and out that coincide with identifiable stations on the band monitor SA.

In my HF-300 rig, the split stator grid capacitor is elevated a few inches above the metal deck, so that the grid tuning knob is even with the plate tuning. It was extremely squirrelly even with the parasitic chokes. I had about a 3/4" wide copper strap attached to the mid-point on the  capacitor frame with the other end connected to the grounding point.  With much trial and error, I finally tamed it down by using two copper straps, each one connected to the common ground point, and the other ends each connected to one of the end plates of the grid tuning capacitor. That capacitor is only about 8" long. Just relocating those ground leads to the capacitor a few inches made the difference. Doesn't take much at VHF.

I have discussed this circuit with several hams ota ... Rex, K4JBJ has built several 813 am tx and he commented that the circuit was a way to improve the performance of a 813 .... when he said this, I recollected ot qsos about how some 813's were 'hard to modulate'
Reed W2CQH decided to look at the turns ratio in a spare ART13 mod xfmr, which uses a tertiary screen winding ... he found a 3:1 turns ratio from the output winding to the screen winding ... I will use this for the 813 data entry

**** some of this data is extrapolated from user notes and will be marked ***

tube type      dc screen voltage      ac peak modulating voltage       notes

6146                175                         105                                    *** many designs over mod the screen - see other threads

4-65                250                          250                                    Eimac data sheet

4-125               350                          210                                    Eimac data sheet

813                  350                          350                                    ART13 mod xfmr

4-250/400         400                          350                                    Eimac data sheets

4-1000              500                          250                                    Eimac data sheet


conclusions are still being formulated but here are some ideas for discussion:

1. those tetrodes that need equal amounts of dc and ac peak voltages would  work most flexibly from resistor dropping the       modulated B+

2. those tetrodes that need lesser amounts of screen modulating would benefit from resistor ratioing of mod/unmod screen V

3. choke modulating the screen (self mod) MAY be difficult in those tubes requiring equal amounts of dc and ac V and require a 'boost' circuit which is what kindled this thread in the first place

4.some families of tubes (4x150/4cx250 etc) require a low impedance source to provide the required screen ac swing to modulate cleanly

In 1989, Deano wrote about this vis-a-vis the 6146.

http://www.amwindow.org/tech/htm/scrnmod.htm

The Champ 300/350 used a seperate screen supply and a choke and sounds good on all bands without that RCA nightmare.

3. Can you elaborate on this 'boost' please? I'm not sure I understand.

4. The only amp here w/ 4X150's makes negative screen current under certain conditions. Is this why #4 is stated?

Hi Patrick ... hows the bunker of doom going ?

3. the 'boost' circuit is the shift of the plate bypass cap (C8 in this TT5 circuit) from ground to the PA screens.  It is speculated that this would increase the screen ac swing as well as restore some higher freq response

4. the note about 4X150 family is from an Eimac data sheet application note.  The negative screen current is from plate secondary emission.  keeping this under control usually requires resistive loading to ground in the screen supply (from Econoco notes)

73   John

Is this "boost" documented anywhere?

Hi Carl .... no, its just a possibility ... an ota discussion with Reed W2CQH this morning about this thread admitted that there is an opportunity to extend the am state-of-the-art here by measuring and quantifying the effects of screen voltage and current with distortion reduction in high level am ...any takers ?