The AM Forum

I have a Hammond 2220 -  rated 15K: 3000,5000,7000 ohms -  modulation transformer. It is marked 175 watts but out weighs another 150 watts transformer that I have by 3 times. It is a bigh heavy chunk of iron

I am looking to use it with two 811As at 1250 volts to modulate a single 813 at 1750 volts.


The driving impedance of the 813 is going to be higher than 7000 ohms. The turns ratio looks like 1.46 , 1.74 and 2.23 at the three taps.

Can I make this work?


Pat
N4LTA
N4LTA

811s modulating an 813 is the same as an ART-13.
http://amwindow.org/tech/htm/modtran/misc/art13.htm (http://amwindow.org/tech/htm/modtran/misc/art13.htm)
15k primary and 7.3k secondary is what the ART-13 mod transformer has. the 7k secondary on the hammond should work. what i've wondered but never heard of this being done before is if you could use one of the other taps on the hammond to feed the screen of the 813 through a dropping resistor. the ART-13 transformer has a tertiary winding to modulate the screen.

Using a single 813 at 1750 volts  at 175 mA is a 10 K driving impedance. I think the ART 13 used a bit lower plate voltage but I am not sure. Seems like to get power out of the 813 I will need a higher plate voltage and that will screw up the match.

I have a couple of 4-65 sockets and some tubes and it looks like a pair of them at 1750 volts is a better match but I wanted to run an 813.


I have several 700 series Hammond plate transformers that I am trying to use.


Pat
N4LTA

A pair of 813s would solve your problem and you could run them light and modulate the heck out of them.

My plate supply is limited  but I think your idea may be what I need to do. The plate transformer is a Hammond 724 rated 1500 at 245 mA CCS or a Hammond 726 rated 1790 volts at 175 mA.

Hammond says multiply the mA ratings by 1.5 for CCS but that may be a stretch although they are usually conservative.

That would be 1790 at 260 mA or 1500 volts at 365 mA

The 4-65As are more efficient at the lower plate voltage than most other tetrodes.

Pat
N4LTA

If you are using a different voltage on the final than on the modulator, you need to adjust the ratio accordingly.

In this case, with a lower voltage on the modulator, you need a smaller step down ratio in the mod transformer.

The effective ration is sqrt(15000/7000) * (1750/1250) = 2.04.

I think you would have trouble getting 100% modulation with that setup -- too much step down.

Note the ART-13 runs both the final and the modulator off the same 1150vdc supply.

Dave

That was the reason that I asked. I don't think the transformer will match a single 813 with two 811As.

Two 813s at 1750  volts at 350 Ma  would be 5000 ohms and that would match.

The power input would be  612 watts which would require a minimum of 306 watts to 100% modulate (probably not enough in practice)

Two 4-65As  at 1750 volts at 240 mA would be 7250 ohms which would closely match the 7K tap.

Power input would be 420 watts which would require a minimum of 210 watts audio which "might" be possible with the Hammond transformer.

Dropping the plate voltage to 1500 at 240 mA  would help dropping the minumum audio to 180 watts.

Still not sure the modulation transformer has the oomph? It is a big heavy chunk of iron though.

Pat

1750V@175mA is about 300W input, and yields a plate impedance of 10000 ohms as you stated. 811A's at 1250V want to see an impedance of about 9000 ohms, but with a plate impedance of 10000 ohms, your 811's would be seeing an impedance of approx 21000 ohms using that transformer. That is probably too much of a mismatch to use that transformer with those tubes at those settings.

HOWEVER, if you run the 813 at 1500V and 200mA, you still get 300W input and the plate impedance is lowered to 7500 ohms. If you then raise the 811 plate voltage to 1500V, now they want to see an ideal impedance of 12400 ohms. Using that transformer on the 7000 ohm tap with the 813 running 1500V@200mA, your 811A's will see an impedance of about 16000 ohms. This would actually be a pretty good match and you would have plenty of spare power from those 811A's.

My 254W rig is actually running at about those settings. Plate voltage on my 811A modulators is about 1400V. They want to see a plate impedance of about 11000 ohms, but they are actually working into an impedance of about 16000 ohms. No problem. They seem to work FB into that impedance.

Bottom line: run the 813 at 1500V@200mA and the 811A's also at 1500V and that transformer will do fine.

Ron W8ACR

.

It is amazing trying to fit it all together. I just came from the garage looking through my "good" stuff.

I found a 1150 volt plate transformer rated 245 ma DC, two brand new Hammond 726s, two brand new Hammond 724s, a new Peter Dahl 2150 at 300 mA CCS made for a CB linear.

The 724s are 1500 volt 240 mA  and the 726s are 1790 Volts at 175 mA CCS.  The Hammonds are rated at 115 volt primary and state ICAS rating is 1.5 x CCS.

Looks like what I need is a plate transformer that is about 1500V at 600 mA to run both the modulator and RF deck.

Anybody want to trade?

That or run both the 724s  - one for each. Wonder if I could parallel them? Series I'd do - Parallel I'd have to think about?


Pat
N4LTA

You might be able to parallel the 724s, since it would be a pair of identical transformers made by the same manufacturer, so the turns ratio, winding resistance, core characteristics, etc should be identical.  Try hooking them up in parallel with no other load attached. Disconnect the midtap and one side of the secondary, but keep the primaries connected in parallel.  Apply full line voltage. Measure the a.c. current to the paralleled primaries, or if you don't have an a.c. ammeter, try running a 40-watt incandescent lamp in series with the a.c. line and note the brilliance.  It should just barely glow if at all.  Even if you have an ac ammeter, the 40-watt lamp would be a good idea anyway, to serve as a current limiting  resistor in case the paralleled windings got connected together backwards.

Now, shut off the power, connect the full secondary windings together in parallel, and then turn the power back on.  If the lamp lights to full brilliance and the current reading takes a big jump, you probably have the windings  reversed.  Again, with the power disconnected, try reversing the primary leads to ONE of the transformers and see if the lamp glows dimmer with the power back on.  Once you are sure the paralleled windings are connected together with the right polarity, compare the primary current reading (or brilliance of the lamp) with the paralleled secondary windings connected together to the current/brilliance with them separated. If there is very little or no difference in total primary current or brilliance of the lamp with the secondaries paralleled and and with them separated, you should be able to run them in parallel with no problem.  If there is a substantial increase in primary current with the secondaries connected together in parallel, that means there is a slight but significant difference in the voltage output from the two transformers, and they wouldn't work efficiently in parallel.

In case there does appear to be too much of a difference between the transformers to strap them together directly in parallel, there was a recent thread on this board by someone who suggested "steering diodes" to equalise the loads. That might be worth a try.

Pat,
      dont foret that the plate to plate impedance is also directly affected by the load on the secondary winding if you are not isolating it from the HV. (Modified Heising) By placing a heavier load on the secondary, you effectively lower the primary impedance. The turns ratio is by far more important as you still need to maintain the voltage swing to properly modulate the final.

Impedance will change with the load applied, but the turns ratio is absolute. The heavier you load whatever you use for a final will lower the operating impedance that the modulator plates see. I have always worried more about turns ratio than published impedances. With a Modified Heising circuit with the DC biass removed from the secondary, this is where the actual impedances really come into play.

As far as the turns ratios go you need to get enough leverage to make the Munky Swing. I always equate it to the fulcrum / pivot location on a see-saw
as you move the pivot point away from center one side gains an advantage over the other.

Also, IIRC the 4-65 is a fairly high inpedance tube preferring higher voltage / lower current applications. I would expect them to be a bit anemic at 12-1500v.  

Sounds as if you're on the right track!  That modulation transformer has too much of a step down ratio to have the modulator DC voltage lower than the RF amplifier DC voltage, so using a common power supply voltage is a must.  Assuming you could turn the tubes on all the way, you're looking at a 1:1.36 voltage step-up.  That's somewhat more than 125% positive in realistic terms, but not a lot more.  125% gets eaten up mighty fast in a high fidelity system and the human voice.  But, it should work if you watch for positive peak clipping and keep the modulation levels below this point.

The actual impedance numbers on the transformer aren't cast in stone.  With any transformer, you can run a higher impedance if the transformer has sufficient inductance, and not excessive stray capacitance, or you can run a lower impedance if the transformer has the current capability (assuming the same power), and not TOO much inductance (or leakage inductance) to hurt the high frequency response or introduce too much phase shift.  I suspect with this transformer you can deviate sufficiently do do what you want.

Assuming the modulator tubes have sufficient power output and current capability, they will, within limits, work into whatever impedance you need.  Triodes are much better in this regard, which is why I always use triodes as modulators.  Go for tubes that have the highest possible plate current and the lowest possible plate impedance.  In this way, you have a lot of current capability available to work into whatever load you have, and also to work into modulation transformers which may have insufficient inductance to properly pass very low audio frequencies without a certain degree of coaxing.  

When comparing tube data, it's the highest current and lowest impedance that matters (assuming the voltage rating is sufficient).  In general, for audio power service, tubes will be able to work into higher impedances or with lower currents than they are rated because they will be delivering less power.

As with all transformer-coupled modulators, the most important component in the system is the AUDIO DRIVER.  With class B or AB2, a direct coupled audio driver is far superior to anything else available and they are not that hard to build.

It sounds as if you have a good collection of iron over there  ;)

Slab,

Eimac specs on the 4-65 says the opposite . They claim the internal spacing makes the tub work efficienctly at lower voltage - but I have no experience.

I have looked for a couple of years for some decent mod iron and the Hammond is the best that I have come up with. It is big at 28 pounds and likely can do over 175 watts but I don't want to use it if it puts me in a corner.

It turns out that a better look  shows that the other transformer is a Hammond 720 - 1250 volts at 300mA - so I can't parallel.

I am sorting out my options - I can get a new Hammond 737 for $325 - 3000 volts at 400 mA CCS and get about 615 mA CCS at 1500 volts - but I still wonder about the turns ratio thing - it's iffy for $325.

I suspect that the Hammond mod transformer was for running two 812s to modulate two 811As maybe.

The audio power is not there for 813s or 4-400s?

I have been having this dilema for 2 years - just can't get all the puzzle to fit together and sure as hell don't want to do it more than once.  

Pat
N4LTA

Steve,
Since the step down of the transformer is too much of a difference to run a lower voltage on the modulators than what is on the final, what if the modulators were run at a higher voltage than the final? I've been wondering about this the last few days.

Oh yes, that can definitely be done and it works as expected - the only thing is, you lose the advantages of a common power supply.

Steve,

I was typing while you posted.

Too much iron - I got a Hammond dealership long ago and spend way to much of my own money buying this stuff for me to use.

The larger transformer can run both the modulator and transmitter and might be the right way to go.

Pat

the 724 sounds like it would be poifikt for a pair of 812 finals modded by 811As, this was a very popular setup years ago and was easily capable of 400w carrier. The 737 sounds like it could handle anything you wanted to throw at it, but there is the cost factor. But, also, $325 for a new 1200w mod transformer is pretty reasonable.

1500v @ 240mils is 360w of audio, so that would be good up to 720w DC inpoot for whatever you used as a final. If the turns ratio is somewhere under 1.5:1 you could probably do a 4-400 modded by a pair of triode connected 813s happily or just about anything that falls into (or you can adjust to) those operating parameters.

Dont forget that the factory specs for a tube are only guidelines and not carved in stone. You can push them a little harder, or run them a lot easier by adjusting operating voltages, biass on both the control and screen grids, and how you load them. You DONT necessarily have to run them at the design maximum ratings, and the design curves are just starting points.

An easy test to determine the turns ratio of an unknown transformer is quite simple. Just apply 120vac to the full primary and read what you get on the secondary and divide it out. Example: if you put 120 on the primary and get 60v on the secondary you know that it is 2:1. (or something very close)

As far as the 4-65 goes, I have never looked at the Eimac charts for them as I never had any interest in them. I just thought they were kinda wimpy. But most of the applications I have ever seen them used in were just about always higher voltage lower current situations, usually running them at somewhere between 2000 and 2500v.  
  

I am tempted to order the 737 and build a good stout 1500 volt supply and start with the Hammond transformer running two 813s at 1500 volts. The modulator will not have the power to fully modulated the pair of 813s but maybe I can find a larger mod transformer later.

Two 813s at full bore at 1500 volts is about 350 ma - 4300 ohms and an input power of  525 watts - needing a minimum of 262 watts of audio.

Two 811As at 1500 volts ICAS can do 340 watts but the mod iron is the limit. The tubes can do the job.

The Hammond iron weighs about 28 pound and I have a Thordarson 150 watt transformer that weighs 15 pounds - so I may be able to push the Hammond a bit.

Still a compromise. 


Pat
N4LTA

Experimenting is the way we learn things. (and have a lot of fun)  ;)

Yep - But please don't let my wife know how much fun the Hammond 737 is.


Pat
N4LTA

According to the catalog page VTP posted, that mod transformer is a "Professional Type" -- probably rated conservatively.

It appears to be intended for use with 811's

I have a Kenyon T478 modulation transformer which is similar size, weight, and impedance ratings.

It is also recommended for use with 811s.

And it's rated at 225 Watts.

My transforment was made before the 811A tubes came out, and if you look up the specs for 811's (without the A), the max output in class B is 220 watts.

Also, Kenyon has other mod transformers with the same case size and weight that are specified for up to 300 Watts.

My guess is you probably can push your transformer a little.

If you are using a Heizing reactor, you can probably push it even more.

Dave

Yes, That's what I am thinking also. I always have tended to believe that weight is critical to power rating. The Hammond 737 is good for 1200 VA CCS  and weighs 28 pounds. That is the same weight of the Hammond Mod transformer rated at 175 watts.

I'd like to try a modified Heising but again theres that 30 H or so choke to find. I could series some Hammond 10H chokes but that again get pricey. I have a high voltage oil cap.


I ordered the 737 and it should be here early next week - so I will get started with the modulator deck. I have a filament transformer and some ceramic bayonet sockets (Chinese not Johnson - I used to have some but can't find them. I have a 15 watt Hammond audio op transformer but will have to study Steve's recommendation on a FET type driver as that may be a better way to go. I have a couple of large chassis and rack panels. I'll probably use five 1000 volt 3 amp diodes in series for each rectifier. Don't have a large enough oild cap or caps so I'll use a string of electrolytics. I really have not had problems in the past with electrolytics as long as they are conservatively rated as to voltage.


I'll need a negative supply to slightly bias the 811As at 1500 volts - I have some PC Board type 6/12 volt transformers that can handle that.



Pat
N4LTA

<snip> "I ordered the 737 and it should be here early next week" <snip>

Hmmmmm............ For the right price I wont tell your wife  ;D  ;D  ;D

If I dissapear  - you all will know why.

Ok. but there are advantages to having two supplies, too. for one, you can be sure that you wouldn't run out of headroom if you have two supplies, one for audio and another for RF. and if you make the modulator supply's voltage variable, you can control how much positive peaks you have by pushing on more voltage.

Don't want to hijack this thread, but now you guys got me wondering if my thinking is clear on this. A couple of posts have mentioned that it is difficult to achieve 100% modulation when the plate voltage of the modulators is lower than the final RF voltage. Take the following example:

Modulator - 811A's at 1300V=optimum plate to plate impedance around 10000 Ohms

RF final - 810 running 2000V@200mA - 400W input, 10000 ohm plate impedance

811A's at 1300V can produce 300W of audio

Using a 1:1 modulation transformer, will the 811A's be able to modulate the RF 100%? I would have answered yes, definitely, since at 400W input, the 810 should only need about 200W of audio to get to 100% modulation and the 811A's can easily do more than 200W. So how do the differing plate voltages come into play here?

My 254W rig runs about 2200-2300V@200mA on the RF final and 1400V on the modulators. IIRC, the taps I'm using provide a turns ratio of about 1.2/1 PRI to SEC stepdown. I seem to have plenty of audio and am apparently modulating very well. For what it's worth, I am running a 50H mod reactor/modified Heising system. I have plate voltage but no plate current on the secondary of the mod transformer.

Comments?

Thanks, Ron W8ACR

I think that I should add a modified Heising type choke for this rig to help the modulation transformer.

Large chokes are very hard to find so I am probably going to have to series the Hammond 10H 500 mA chokes.

How many do I really need?


Is there a formula for calculating the choke size?


Pat
N4LTA

the rule of thumb that i saw is 10H for every 1000 ohm of RF final impedance.

That means I need 4  - 500 mA 10H chokes - That is about what I was thinking.

Boy, This is going to be heavy, Anyone know of another source for modulation chokes?

Pat

Dennis (the Iron Menace) w7tfo can help...

Peter

Audio iron is in short supply right now.

Power & filament iron is plentiful.

73DG

Any power chokes?

Pat

Pat,

A modulation choke needs a pretty high voltage spec.

The peak voltage at 100% modulation is twice the B+, or 3KV in your case.

To be safe, probably figure 5KV chokes.

I'm not sure which Hammond chokes you are considering, but the ones I found listed on their site were rated at 1000VDC.

This is too low -- I'm not even sure they would work if insulated from ground as the audio voltage across the choke would be pretty high when used as a mod reactor.

Dave

they might work if you insulate them from ground. i was looking at those same chokes, i wish they would put on there what they hi-pot tested them at, if they do such a thing with their chokes.

I have used them at 2000 volts in the past sitting on 1" insulators.

There just isn't much currently manufactured iron around. My thought would be to mount all four on a plate and insulate the plate above ground. They are expensive - wholesale $55 each - I hate to blow up four of them.

Pat
N4LTA

Can an individual buy transformer parts like the E and I sections, laminate and all that stuff and just assemble their own transformers?  Maybe some kit offerings where one can sorta customize the ratings would be an interesting biz.

Pat:
Start by using all of the copper a mod transformer offers, don't worry about impedance numbers. End to end windings on primary and secondary, don't bother with the taps. Then measure, scope it and tweak it from there. You likely won't find a better setting than using the whole of both windings.

You can use Zeners in the cathode return of the 811s to properly bias them at more than 1200 plate volts.  Very easy to do. OR you can use forward biased regular diodes at .6 volts per each. At 1200 volts on them, zero bias on 811s will prolly be fine.

My ART-13 makes a good 150 watts out on 75, with 1200 volts on both the 811s and 813. My friend, KE0MT gets near 250 watts out of his, with near 2 KV plate voltage and a blower on the toobs.

Bill

Thanks Bill. I had forgotton about using zeners. That will be easier than a bias supply.

I may have found a modulation choke which will help matters.

I am passing some time this weekend making a "scale model" to play with.

If it isn't too ugly, I might post a picture.

Pat
N4LTA

Decided to build up a "scale model" of the twin 813 rig to experiment with and a Slab said to learn on. Using two 807s at 400 volts to keep the driving impedance low. I am not worried about max performance on this one. Has a Hammond 278 that was out of the box - 400  - 0 - 400 V at 200 Ma - so A choke input supply for 400 volts was in order. I also wanted to experiment to see how much the standard Hammond  chokes swing at low load.

I built a small PC board with two 220 uF/450 volt snap ins, four 1N5408s (two in series in each leg) and a little DC supply in case I need it for relays. The caps were balanced with two  five watt 25 K resistors.

The resistors will pull 8 mA at 400 volts - not enough for a bleeder to bring the choke input supply to the critical value. The choke is a 10H - so if it is 10 H at light load - it should take 40 mA to make things work.

Fired it up and got 497 volts (with 8 mA load) which was a little better than I expected. With this light load, I expected it to be close to the peak  of 565 volts. I then stuck two 10 watt 10 K resistors in series on the output and it settled down to 401 volts nicely. So that makes the total load 28 mA and the choke input filter is within the critical current. That tells me the choke is at least  14 H at light load - which "might" indicate that I can get 40% more inductance out of the Hammond chokes at light load. It does swing somewhat.

So now to the RF section......

Pat
N4LTA

So where can we ordinary people get these chokes for $55?

They are about twice that much at the on-line retail sites.

Where is the best place to get good deals on Hammond xfmrs and chokes?

I am a Hammond dealer

PM me for a AMer discount