Please help me unlock the secrets of the ancient Inca...

[W8KHZ]

...modulation transformer... 

For quite some time I've been in possession of a pre-war Inca N-28 multi-match modulation transformer made by the Phelps Dodge Copper Products Corporation.  It is rated at 600 watts.  I'm contemplating using it in a homebrew rig but I don't have much information on it.  I found reference to it in an ad I found online. It describes it as follows:

"INCA N-28 Universal modulation transformer from any class B tubes (up to 600 watts audio output) to RF stage, carries full RF plate current through secondary. Primary 4000 to 16000 ohms. Secondary 2000 to 20000 ohms." 

It may be a long shot, but does anybody here have the connection info for this transformer?

If I can't come up with the connection info, I would like to get some advice from the gurus here on what the best approach is to figure out what the connection info should be.  I have traced the windings with an ohmmeter and determined that there are 4 center tapped windings. At this point I could use some help...

How do I determine what windings are primary and which are secondary? Or are they interchangeable depending on what impedance ratio you are shooting for.  I assume I can feed the windings with a signal generator to determine the ratios but how do I translate the ratios to some sort of impedance values like you would find on a connection chart. Maybe once I determine the ratios, I can extrapolate from the upper and lower values given in the ad? 

If I can get this all figured out, I would like to give it a test run by swapping it with the CVM-5 modulation in my current homebrew rig. I want to make sure its going to work before I build a transmitter around it!

Any help is appreciated!

Thanks!

Brian - W8KHZ

[WBear2GCR]

Presumably they are interchangeable, but they may have made the secondary with a higher breakdown voltage.

The way I would go at it is to use the formula for turns ratio to impedance. The turns ratio is equal to the voltage ratio(s). So you can put a 10volt source on the whole winding of one side and see if the other "whole winding" side has higher or lower voltage. If it is higher, then that is the secondary side, per the spec.

Similarly you can identify the other windings using the same formula and the voltages found, etc.

                   _-_-bear

[KL7OF]

I have one of these transformers in use as a replacement in a BC1-G Gates..833s X 833s...The tag is missing on mine..I always wondered what brand it was...I will have to look but I'm sure I strapped some  of the posts together to get the desired ratios..My transformer came out of a homebrew rig built in the 40's....250ths X 100ths...When you come up with the ratios, Let us know....I installed mine in the Gates 15 yrs ago and have forgotten everything I knew about it...
GUD LUK....Steve

[W3RSW]

Well since you've told us that,   go ahead and take that look or a take a pix and explain what wire goes to what in the rig, what posts you've strapped for your BC 1g and if the voltages, impedances, tube specs. are still stock. That will give him a starting point.

[KA2DZT]

The four windings are tapped but unlikely center taps.  The taps on multi-match type mod xfmrs are offset.

[KL7OF]

I walked out to the shack and took these pictures...The plate voltage on my Gates is stock  (2500 V) the tubes are a pair of 833s mod by a pair of 833s...In the picture you see the HV choke and cap in the foreground ......the Plate transformer is in the back on the right side and Mod reactor is on the back left side  That Mod reactor is out of a 10KW gates rig so it is much bigger the rest of the iron..On the Inca mod transformer,you can see the posts that I strapped together with aluminum straps.



On second look I see that my transformer is not the same....Mine says N26A,   The other  is an N28...Mine has 11 terminals and the other one has 12...so in the words of Gilda Radner.."NEVER MIND"

[W8KHZ]

KA2DZT, you are correct about the taps on the 4 windings being offset. I probably shouldn't have described them as being "center tapped" but I was trying to convey that they were in fact tapped.  When I get a chance I will try to post the  resistance measurements of the windings/taps. Maybe it will help in figuring this thing out!

Brian - W8KHZ

[w8khk]

Resistance measurements by themselves will not be very helpful, and may be misleading.  Voltage measurements will be far more intuitive, as they relate directly to the turns ratios, from which you are able to calculate impedance ratios.  Resistance measurements may be misleading because the same size or gauge wire may not be used for all the windings, and even if the number of turns are the same in two windings, the part that is nearer the outside of the core will have more wire length for the same number of turns, hence a larger resistance reading.

I had similar challenges when using multi-match transformers, such as the Stancor 3899, the Thordarson T11M78, and the UTC S21.  I was lucky that documentation was available for these transformers.  UTC not only gives connection charts, but also schematic diagrams of the transformers, so that you can see which windings are placed in series or parallel.  Perhaps studying some of these documents (available on the AMWindow, BunkerOfDoom, BAMA Edebris, etc, would give you a better understanding of how the multi-match transformers are configured.  

Typically, the primary consists of two windings placed in series, with B+ applied to the junction between the two windings. Often, the secondary needs to be somewhat lower in impedance, and is sometimes configured with two windings in parallel.  Parallel use of the secondary windings also allows a larger current draw for the final amplifier, if necessary.

After studying schematics and impedance transformation tables for UTC transformers, then, instead of resistance measurements (which will only identify separate windings and taps) apply 5 or 10 volts to a winding from a 60HZ filament transformer.  This is a low enough source impedance that you will not need to make adjustments of the audio generator output level when comparing winding voltages.  Measure and map out the voltages across each winding, and each winding segment from each end to the tap.  Then you will see if you have four identical tapped windings (unlikely) or two pairs of like windings, each pair different.

From these measurments, you can then connect a pair of windings in series, and measure whether the voltages add or subtract.  This will enable you to identify winding direction, or phase relationships for each winding.  This is absolutely necessary to be able to properly connect windings in either series or parallel.  You can series unlike voltage windings (for the secondary RF load side) but you should have identical voltages each side of center tap for the primary, and you MUST have identical voltages if you parallel windings, or you will damage the transformer.  

Calculate the impedance ratio needed to place this transformer in your operational rig.  Then from the impedance ratio, calculate the voltage ratio needed from the entire primary to the secondary.  See any handbook for this method and use your square and square root calculator to transform the ratio.

Once you know the voltage ratio needed, you should be able to come up with a pinout wiring combination to come close to the desired voltage ratio.  It does NOT need to be exact.  Close gets you there!

One rule of thumb, is that if you want to use the transformer at its rated power level, it is necessary to use as much or all of each winding as possible.  For example, using one winding for secondary when it could be paralleled with an identical winding places all the current through the one winding, and the winding resistance  causes unnecessary heating compared to using both windings in parallel.  (The impedance ratio and voltage ratio would still be the same, but the power handling capability of the winding is doubled when placed in parallel.)

I think you will get a good idea of how all this works by studying multimatch transformers that have good documentation and schematics.  Then after you take your voltage readings and draw a diagram of your transformer with relative voltages and phasing dots, many of the folks on this forum will have enough information to guide you closer to a solution.  Hope this helps, and have fun!

[W7TFO]

Ratios & impedances have value, but as they say, the map is not the territory .

Mainly that is the reason you don't find hard and fast data on transformers like this, just a range of things to stay within.

I always start with full windings on both sides, run it with a trapezoid scope at half or 3/4 RF power.

Make & record your observations, then move it to the next tap down on either primary or secondary.

Maybe an hour or so in the workings, then you will have it done and move on.

73DG

[W8KHZ]

Thanks guys!

Some great suggestions here... I'll do some more in-depth testing in the upcoming days and will report back!

Brian - W8KHZ

[KA2DZT]

Usually one side of the xfmr will have higher impedance than the other.  So, depending which way you set it up, you can have a step-up or down from the modulators to the PA.  Best to always use the entire winding on the modulators, from that you can use some or all of the secondary for the PA.  Most of the time, using a single plate supply, you will have step-down going to the PA stage.  So, with the two secondary windings in series, you tap up from the B+ end of the two windings.  Never tap down from the class C end of the two windings.  With this set up you only have two selections, one using the whole of both windings or one tap up from the B+ end.  To get a greater step down ratio you would have to put both secondary windings in parallel when you're unlikely to use any taps, it would be the whole of both windings in parallel.

The turns ratio is what is important with mod xfmrs.  The impedance seen by the modulators in determined by the turns ratio and the load impedance of the PA stage.  The lower the PA load impedance the lower the impedance seen by the modulators for a given turns ratio.

Fred