A Test for modulation transformers - for correct secondary polarity connections

[K1JJ]

In the past, I've never paid attention to the secondary "Plate" and "B+"  markings on a modulation transformer. Some mod transformers have no markings at all.
According to a friend, there is a test to determine this secondary polarity. I was not given a reason yet, but perhaps it has to do with case arc-over, something like using a CT power transformer for bridge, or arc-over from sec to pri,  but not sure at this point.

Here is the test and my own results. I was able to determine the "plate" and "B+" leads by two complementary tests.


If already marked:
The secondary terminals are marked P and B.  B goes to the coupling cap to ground, P goes to the juncture of the mod reactor and the PA tubes. This is important.


If unmarked and a test is needed:

There is a way to determine which is P and which is B. Simply connect a 120VAC  suicide cord up to the primary. Connect a meter such as a Simpleton 260 as follows. Place into the 250volt AC range for openers. Connect one side of the meter to the center tap of the primary. Measure to one side of the secondary. You should see an AC voltage due to capacity between the windings. Measure to the other secondary terminal. The second reading should be either higher or lower. The terminal that has the highest voltage with respect to the center tap is the "P". The "B" has the least potential difference to the center tap. This is due to  the fact that this side of the secondary windings is closer to the primary. Results of this test can be duplicated by connecting one side of the meter to the case ground of the transformer, the other lead to be placed  to either side of the secondary to determine which side produces the greatest voltage reading . This as well is the "P" side.
  

I did the tests and both techniques pointed to the same “plate” higher voltage indications.
The BC iron weighs 59 pounds and has a 2.1 stepdown impedance ratio.
I measured 41V : 38V in the first CT to sec test.   I measured 46V: 26V  in the second case to sec test.   So I have a good indication of the plate terminal.

I now need to look at my 4X1 rig to be sure the polarity was wired  correctly back in 2012.  Seems to me it was labeled and I wired it correctly, but not sure.


 T

[w8khk]

In the past, I've never paid attention to the secondary "Plate" and "B+"  markings on a modulation transformer. Some mod transformers have no markings at all.
According to a friend, there is a test to determine this secondary polarity. I was not given a reason yet, but perhaps it has to do with case arc-over, something like using a CT power transformer for bridge, or arc-over from sec to pri,  but not sure at this point.

Tom, as I understand the issue, there are actually two reasons for observing proper polarity while connecting the secondary.  

One, as you stated, is less risk of arc-over due to more insulation resistance closer to the modulated output terminal (i.e., further from the primary winding and core) to deal with higher peak voltages.  But the primary reason is there is less capacity (between the primary/core and the output end of the secondary) resulting in less attenuation of the modulation power at higher modulation frequencies.  (This capacity is actually in parallel with the RF bypass capacitor, and the DC blocking capacitor to the pi-network tank, if present.)

Your test at 60 CPS (Hertz, for the new guys) revealed the difference in capacity.

If the test at 60 CPS is not conclusive, a greater difference in voltage between the two tests will likely result if using an audio generator at 1000 CPS or higher, and an oscilloscope instead of the "Simpleton 260" or other VOM is used for the indicator.  A bit of loading via a shunt resistor may be needed to observe a significant difference in voltage, considering the higher impedance input of an oscilloscope (or DVM).

Along the same lines, there is a "B+" end and a "Modulated" end of the winding on a modulation reactor, measured in the same manner, for exactly the same primary and secondary (no pun intended) reasons.  Here the shunt capacity of the modulation reactor can make a very noticeable difference in the overall high frequency audio response of the transmitter.

I never thought about it before, but I wonder if the same issues might also apply to a power supply inductor, whether it be swinging or smoothing.  I would suppose that the lower capacity, more distant end from the core, might have more insulation resistance and provide better immunity to arcing on input power surges if it was connected to the rectifier (source) end of the power supply circuit.

Edit:  This issue may not be relevant for some of the "Poly-Pedance", "Multi-Match" or "Vari-Match" multiple winding modulation transformers.  Depending how the windings are arranged on the core, and how the user chooses to inter connect them, there may not be a significant difference in capacity readings when measured from the various terminals.

[K1JJ]

Interesting comments, Rick, thanks.


I've never seen labels on small ham-type transformers, but it is common on big BC iron.

I had a rough time getting clean highs on my 4X1 which used a 1KW BC mod xfmr. It took a lot of effort. Maybe the polarity is still backwards in there...  :-)

OK on modulation reactors. I will try some tests later today and see.

** BTW, unrelated subject:   I have been asked by many over the years about what to use as an audio driver for the 813 rig or any plate modulated rig for that matter.  I was looking at the cheap eBay class D audio amps driving a LS-57 output transformer. A friend said he didn't  like the high frequency IMD and potential switching problems of the Chineses class D amps. He suggested using a LM-1875 20 watt chip instead to drive the 8 ohm > 5K output transformer. The specs look great.

The other alternative is using the GFZ MOSFET 11N90 audio driver board which is my plan later on.

T

[W1RKW]

GORT uses an LM1875 driving an 8ohm to xxxxK ohm transformer in reverse to the grids of the 813s.  The LM1875 is a good cheap way and provides flat and phase shift free frequency response but my audio transformer, not so much.  Never had the time to get rid of that bottle neck.  

Frank's MOSFET driver, a bit more complex but it would be the best way to go vice reverse audio transformer.

[K1JJ]

Hi Bob -

I didn't know you were using the LM-1875.

Here's a  LM-1875 kit on eBay for a few bux from China a friend sent me.  How do they do it?

It takes a month to ship, so I must decide whether to build the more complex GFZ board or take the EZ way out.

https://www.ebay.com/itm/20W-HIFI-Mono-Channel-LM1875T-Stereo-Audio-Amplifier-Board-Module-DIY-Kit/401332958044?epid=21009783134&hash=item5d714ef35c:g:0zYAAOSw42dZJHBx


T

[KC2ZFA]

alternative three ?

http://k6jca.blogspot.com/search/label/813%20AM%20Transmitter%20-%20accessories?m=0

[W1RKW]

Hi Tom,
I had thought about using an outboard audio amp to drive the transformer when building but wanted a more compact and self contained amp.  One day I was thumbing through an IC datasheet book and stumbled upon the LM1875 and bingo and went on from there.  I used the circuit provided in the datasheet. Wired it up on perf board and slapped a sizeable heatsink to the LM1875 and stuff it in a small metal box to shield it and mounted the box to the modulator chassis.  Right next to it is the audio transformer and from there onto the 813 modulators.

One of my fears was the use of an unbalanced input with the LM1875. The audio line is balanced going from the operating position to the transmitter but at the transmitter end it is converted to unbalanced using an op-amp based converter which is also contained in a metal box. The unbalanced line is about 8 or 10 inches long and there are no RF issues with the audio, surprisingly.

Let me know if you want a picture of the amp and I'll pull the modulator from the rack.  Please say yes. it might give me motivation to swap the audio transformer.  

Like you, I don't know how they sell those kits so cheaply.  I don't know what the LM1875 cost back in 2001 or so but it was certainly more than that kit  

Hi Bob -

I didn't know you were using the LM-1875.

Here's a  LM-1875 kit on eBay for a few bux from China a friend sent me.  How do they do it?

It takes a month to ship, so I must decide whether to build the more complex GFZ board or take the EZ way out.

https://www.ebay.com/itm/20W-HIFI-Mono-Channel-LM1875T-Stereo-Audio-Amplifier-Board-Module-DIY-Kit/401332958044?epid=21009783134&hash=item5d714ef35c:g:0zYAAOSw42dZJHBx


T

[K8DI]

For a different approach...what about a parts express Dayton Audio DSPB100?  100 watts, built in dsp (Analog Devices, program it to do bandpass filters, EQ, compression/limiting). It’s $35, the usb card to be able to program it is $25, but it doesn’t stay connected so you can use it with other amps/dsp boards later.  I haven’t messed with that particular amp but I’ve been playing with the usb card and a couple dsp boards with the same dsp chip.  Amazing dsp power for the cost. One of these at 100w will be loafing driving a good tube amp output transformer in reverse to hit your 813 grids...use a hifi amp transformer (vs a guitar amp one) and it should be pretty linear. And, they’re not shipping direct from China, you can get it in a couple days!

Ed

[K1JJ]

Lots of good ideas...

Bob, in your case, since you already have the LM-1875 and transformer working as a temp solution, why not put the time into building a GFZ MOSFET board so you can drop the transformer? That will make the biggest difference, especially when you add the NFB.   I think I'm going to order parts for the GFZ board from John/JSW's bill of materials list soon.

Yes, all balanced audio chains eventually turn into an unbalanced feed somewhere. Even the GFZ board now has balanced input, but the 813 grid feeds are unbalanced (though higher level).


Another friend sent me his version of the mod transformer  P  B+  polarity test:

"My Transformer Test.

Ground the primary and one secondary lead (start or finish if multi tap).  Take a series resistor say close to the plate impedance and connect it to the open secondary lead.  Drive the resistor with an AC voltage. 60 Hz will be fine but it can be anything in the audio range. The AC voltage reference is the common ground of all the other transformer leads.  Measure the voltage at the junction of the secondary and resistor with other meter lead also at ground. Record that voltage.  Now swap the two secondary connections and do the same test. The lead with the highest voltage has the highest impedance to ground. "

So we have no excuse not to get the polarity right.  


* I'm building up a new HV power supply for the 813s X 813s rig. The transformer is a 2000-0-2000 V/  1500-0-1500  1.750 KVA  675 mA 100 pound brute.   It will give me either (both cap input) 2100VDC or 2800VDC.   Perfect for the 813s.  I am using Shane's 120 VAC step start circuit for the auto PTT keying. I like keyed HV supplies... hate to leave them on.  I've been wondering if I would ever put this transformer to use. Carrying that 100 pounder up the cellar stairs was sporting.

I will start taking some pics soon. It's a frog now, but will be transformed into a handsome prince soon. Cleaned all the parts and will paint the panel.   I still gots to come up with a new name. "Hollywood" is my current choice but that could chance in a Marlborough Minute.

T

[W1RKW]

Tom,
you're absolutely right. Franks board is the right way to go.  I figured swapping the transformer would be quick and dirty with little to no drilling and blasting since time isn't on my side right now.  I have the GFZ board built and have the power transformer John is using.  Just have go through all the mod notes and install it but I  also figured on building up a new modulator chassis for it to make it a clean build.  Also have another mod transformer that I'd like to put to use or try.  Don't know if the audio transformer is the issue or the CVM-5 or both.

[W1RKW]

T,
You should name it Cool J since it's a summertime cool rig.