Gates BC-1T questions

[KM1H]

What are the design impedances for the mod transformer and how much power will it safely handle and without saturating?

What is the value of the mod reactor?

Is the coupling cap considered adequate?

Carl

[k4kyv]

My old Gates catalogue lists the reactor as 32 Henries, although some have said that early ones are 50 H.  The mod xfmr is about a 1.35:1 turns ratio and therefore about a 1.8:1 impedance ratio.  Since the final runs at approximately 2600v @ 600 MA, the secondary impedance is around 4300 ohms, making the primary impedance a little less than 8K.  I have the exact figures but would have to dig them out next time I am out in the shack.

Since the transmitter runs about 1500 watts DC input, the mod transformer should be good for at least 750 watts of audio.  I wouldn't run it at any more than that, since I have heard reports that they are prone to crap-out, and when I got my BC1-T, I got a spare mod reactor and transformer with it, and the transformer was shorted.  

I suspect they crapped out the original and found a replacement transformer/reactor package deal or else the engineer didn't know if it was the xfmr or reactor that had crapped out and needed a quick replacement so he purchased both to minimise down time.  In the BC station, the transformer was just sitting in place and not screwed down. When I got all the stuff home and decided to put the transmitter on the air, I mounted the mod transformer on ceramic stand-offs and made sure the case/core was not in contact with ground, and placed a spark gap across the primary.  I also noticed that the side panel was uncomfortably close to the paper-covered winding of the mod transformer, only about a quarter inch or so away.  When I worked at another BC station years ago with one of these transmitters, a plate transformer blew because the a.c. primary wire to the transformer got too close to the winding, and the HV arced through the paper.  So I made some metal spacers to move the side panel about another half-inch out away from the transformer, giving it adequate spacing so cobwebs and dust would be less likely to accumulate in the gap between transformer and side panel and possibly arc over. The spaced-out panel is hardly noticeable from the outside.

I wouldn't run any DC through the secondary, although Timtron once ran about 400w input to a rig and ran the DC through one of those transformers, and said he could tell no difference in frequency response, but that it talked back pretty severely with the DC going through it.

I also have the specs out in the shack for the tapped tertiary winding used to modulate the driver stage. I'll try to remember to copy down the specs next time I am out there and post them.

I use the stock coupling cap and it seems perfectly adequate.  Others have posted that they got excessive overshoot in the audio waveform when using more coupling capacitance than specified.

[W7TFO]

Typical of all the BC-1 series, the following pertains:

Mod reactor: 32 Hy minimum @ 600 mA, and would usually measure up to 60 Hy at no load.

Mod transformer:Primary 2450 Volts CT.  Secondary 1800V.  Frequency response 40 to 15K HZ.  Primary PP load is 9.5K Ohms.  Design center was for 2X833A PP to 2x833A Parallel. 

They also have a tapped tertiary winding for tossing a little mod into the RF driver 807's

All ratings were at continuous service.

Since this type of transmitter proofs out easily within the specs of the mod iron, I'd say the coupling cap is plenty adequate....far in excess of anything factory built for amateur service.

In my days of running audio proof of performance measurements, they would run less than 3% THD @ 100% mod throughout to around 11KHz, and slowly ramp up the distortion to just under 5% at 16KHz.   More than once running over 15kHZ @ 100% mod blew up things in the antenna tuning unit.....a good way to find caps that needed replacement anyway.

Really modulating under 30Hz would cause the OL relay to trip.  40Hz would go all day.

73DG

[k4kyv]

Mod transformer:Primary 2450 Volts CT.  Secondary 1800V.  Frequency response 40 to 15K HZ.  Primary PP load is 9.5K Ohms.  Design center was for 2X833A PP to 2x833A Parallel. 

Of course, those are rms voltages.  1800 Vrms = 2545 Vpeak, which is almost exactly what would modulate the transmitter 100% at 2600 Vdc on the plate of the final.  The 9.5K figure must assume less that 600MA plate current to the final, but that impedance rating wouldn't be critical at all.

Maybe that's where the 50H figure came from.  Someone might have measured it with no DC passing through it.  I replaced mine with a Kuhlmann potted reactor rated at 50H / 600MA, because the stock reactor talked back too loudly.  The potted reactor doesn't talk back, and the mod transformer is just barely audible.  The  reactor talk-back made it unusable as a ham transmitter because I could just barely crack the audio gain before I got feedback.  I use one of those 1 KW Gates (Electro-Engineering) reactors in my 8005/805 rig at 400w input, and the reactor talkback is very annoying even at that power level.

[KM1H]

Thanks guys.

Another question....since the power transformer handles all 4 833A's in CCS its got to be pretty hefty KVA wise. So the question is...can it be run with just a big cap for the filter and some degree of steady bleeder current to keep the voltage from soaring and cooking the secondary?  If I used it that way it would be with 6A10 diode strings in a conventional FW. I'll grab the swinger if need be.

I'll also get the filament transformer since it will handle 304TL's and maybe the audio driver iron also.

[k4kyv]

Without a choke, the output voltage would run up close to 4 kv.  Choke input gives a DC output voltage of 0.9 X Vrms from the transformer.  With  capacitive input, the voltage soars to 1.414 X Vrms, but it tends to sag back down with load unless the transformer has very low leakage reactance. Many of the older transformers originally designed for choke input and tube type rectifiers have poor regulation when used with capacitive input.  I don't think you will cook the secondary of the plate transformer, but unless you have a good variac or auto transformer, the high output voltage might cook the mod transformer, reactor or some other component if it doesn't blow the filter cap, since the transmitter was designed to run with 2600v on the 833A plates. I would use step start and about 100 mfd of filter of adequate voltage rating with cap input.

Since you have the swinging choke that came with the xmtr, why not use it?

I have used solid state direct plug-in replacements for the 8008s and 866As for almost 8 years now without mishap.  I replaced the 8 mfd filter cap with a 25 mfd one, and reduced the bleeder from 100K to 50K. No step start required. I modified the rig to work on CW as well as AM; on CW I switch in the mod reactor to serve as additional filter choke inductance to maintain critical inductance and proper plate voltage key up/key down. The meters show about 100 volts difference load/no load, but the dynamic regulation still shows some jumping about with the instantaneous plate voltage when I key the xmtr on CW.

[W2XR]

My old Gates catalogue lists the reactor as 32 Henries, although some have said that early ones are 50 H.  The mod xfmr is about a 1.35:1 turns ratio and therefore about a 1.8:1 impedance ratio.  Since the final runs at approximately 2600v @ 600 MA, the secondary impedance is around 4300 ohms, making the primary impedance a little less than 8K.  I have the exact figures but would have to dig them out next time I am out in the shack.

Since the transmitter runs about 1500 watts DC input, the mod transformer should be good for at least 750 watts of audio.  I wouldn't run it at any more than that, since I have heard reports that they are prone to crap-out, and when I got my BC1-T, I got a spare mod reactor and transformer with it, and the transformer was shorted.  

I suspect they crapped out the original and found a replacement transformer/reactor package deal or else the engineer didn't know if it was the xfmr or reactor that had crapped out and needed a quick replacement so he purchased both to minimise down time.  In the BC station, the transformer was just sitting in place and not screwed down. When I got all the stuff home and decided to put the transmitter on the air, I mounted the mod transformer on ceramic stand-offs and made sure the case/core was not in contact with ground, and placed a spark gap across the primary.  I also noticed that the side panel was uncomfortably close to the paper-covered winding of the mod transformer, only about a quarter inch or so away.  When I worked at another BC station years ago with one of these transmitters, a plate transformer blew because the a.c. primary wire to the transformer got too close to the winding, and the HV arced through the paper.  So I made some metal spacers to move the side panel about another half-inch out away from the transformer, giving it adequate spacing so cobwebs and dust would be less likely to accumulate in the gap between transformer and side panel and possibly arc over. The spaced-out panel is hardly noticeable from the outside.

I wouldn't run any DC through the secondary, although Timtron once ran about 400w input to a rig and ran the DC through one of those transformers, and said he could tell no difference in frequency response, but that it talked back pretty severely with the DC going through it.

I also have the specs out in the shack for the tapped tertiary winding used to modulate the driver stage. I'll try to remember to copy down the specs next time I am out there and post them.

I use the stock coupling cap and it seems perfectly adequate.  Others have posted that they got excessive overshoot in the audio waveform when using more coupling capacitance than specified.

Carl,

I agree with Don on the specs for the mod xfmr and the mod reactor as used in the BC-1T, the BC-1G, and the BC-1H series.

I have the mod xfmrs and mod reactors I had obtained from a pair of BC-1Gs. The factory mod reactor is rated at 32 hys at 600 MaDC. I would use the stock DC blocking/coupling capacitor, which I recall is rated at 2 uf. Going much higher than this resulted in ringing distortion in my homebrew rig (2x 4-400As by 2x 833As) when using the BC-1G iron.

I believe the only 1000 watt-class rig Gates built with a 50 hy mod reactor was the superb sounding and solidly-built BC-1F. I have the original Chicago Standard mod xfmr from that rig, all 200 lbs of it, but the original Chicago 50 hy reactor was changed at some point during the rig's lifetime with a 40 hy Electro-Engineering replacement. That too is a nice, robust mod reactor, and I'm using it as the screen reactor for the 4-400A screens in my plate modulated rig. Once Gates rolled out the BC-1T, their build quality really went down the tubes, along with the subsequent kilowatt-class transmitters they had manufactured.

The stock Electro-Engineering mod iron as used in the BC-1T, -1G, and -1H was of excellent quality, and I agree with W7TFO/Dennis's assessment as to the low-distortion and wide audio bandwidth of that transmitter. I see similar measured performance in my homebrew transmitter. The BC-1T, -1G, and -1H audio driver circuit is of a more practical and better design overall when compared to the circuit I am using in my rig, which I pirated from the BC-1F; push-pull class A 845s, transformer coupled to the class B 833A modulators, but only 6 dB of voltage feedback is required with that design, when compared to the 23 dB of NFB used in the cathode-coupled 807 audio driver circuit in the BC-1T and later BC-1 series rigs.

I think you too will enjoy good success with those Electro-Engineering BC-1T modulator magnetics. Should make a really FB sounding transmitter.

73,

Bruce

[KM1H]

If I take the plate transformer it wont be used with the mod components, I already have 2700VDC @ 1.5A available out of a big choke input supply that handles the RF and mod needs already.

With the Gates plate transformer Im concerned about the secondary impedance more than leakage reactance. Without sufficient load it can raise the secondary temperature if the wire size is insufficient. It will be interesting to see if it is a late 50's original or a more modern replacement that has a lower DC secondary resistance. Since it apparently has a 1A or so CCS secondary I can bleed it a bit heavy and still get sufficient B+ and regulation to run a 8877 on 222MHz CW/SSB.

My primary interest in the mod iron is to get everything together in one rack. Right now Im still using that 7500W mod iron I picked up several years ago and its got its own dedicated floor space which could be better utilized.

Carl

[k4kyv]

With the Gates plate transformer Im concerned about the secondary impedance more than leakage reactance. Without sufficient load it can raise the secondary temperature if the wire size is insufficient. It will be interesting to see if it is a late 50's original or a more modern replacement that has a lower DC secondary resistance. Since it apparently has a 1A or so CCS secondary I can bleed it a bit heavy and still get sufficient B+ and regulation to run a 8877 on 222MHz CW/SSB.

I have never heard of a plate transformer running hot because of insufficient load, or small wire size in the secondary winding causing it to run warm at minimum load. If you are aware of any technical references having been published on that subject, please let us know where this information can be found.

I would think small wire size, which would result in higher ohmic resistance in the winding, would cause just the opposite effect. With a light load, the transformer would tend to run cool, and it would heat up with a heavier load. Normally, temperature rise in a transformer is due to resistive losses in the wire making up the windings, with iron core losses having a measurable but smaller effect.

The most likely problem caused by too light a load impedance on the secondary would be if it is running right at the ragged edge of the insulation breakdown voltage, so that the unloaded output voltage might be just enough to cause the winding to arc through.  But any transformer even halfway designed will have plenty of safety margin in its insulation, except maybe a very old transformer with badly deteriorated paper insulation.

Too low a winding impedance can be a problem when it results in insufficient turns per volt (or in other words too many volts per turn for the core size), and the core saturates with the a.c. voltage applied.  This is likely to happen if you try to run a transformer at double voltage, for example, using a 5V filament transformer with 110V primary to obtain 10 volts with primary connected to a 220V line.  The primary winding would likely draw dangerously heavy current even with no load on the secondary.  Sometimes you can see this effect using a variac.  Put an a.c. ammeter in series with the transformer primary with nothing connected to the secondary, and feed it with the variac.  The current reading should be very low, barely perceptible, just enough to excite the core with 60~.  Run up the voltage beyond the nominal primary voltage and eventually a point will be reached where the a.c. primary current suddenly begins to increase rapidly with further increase in primary voltage. This is the same saturation effect used in the Sola Constant Voltage transformer; that's why they tend to be noisy and run hot.

I can think of plenty of cases when a plate transformer runs almost  completely unloaded, with only the bleeder current or bleeder + static plate current: a class-C CW final biased to cutoff, a class-B SSB linear, or the class-B modulator in an AM transmitter with separate supplies for modulator and final. This will certainly be the case if the HV stays on all the time with the final biased to cut-off during stand-by. I have never heard of the plate transformer in one of those supplies warming up when there is little or no load, and then cooling down with full load.

[W1ATR]

These are the two chunks stripped out of a BC-1F a few years ago. Donno if the nameplates will give you anything useful or not.

http://amfone.net/Amforum/index.php?action=gallery;su=user;cat=47;u=1019

[k4kyv]

You never know what you are getting with a BC1-F, although stock iron is bound to be heavy.  I have seen pictures and descriptions of ones with what looks like Thordarson Tru-Fidelity iron, as well as Chicago iron, both types open frame/cast end-bell construction.  Wonder if that potted iron is stock or replacement?

I got the mod transformer  taken from one years ago (and still have it).  It is a UTC LS-691 that weighs in at 300 lbs, although I think that one may have been a replacement for  the original.

[KM1H]

I have never heard of a plate transformer running hot because of insufficient load, or small wire size in the secondary winding causing it to run warm at minimum load. If you are aware of any technical references having been published on that subject, please let us know where this information can be found.

In no particular order:

1. http://www.w8ji.com/choke_input_power_supply.htm

2. QST articles in the 30's explaining that the start up voltage in a 2000-2500VDC supply using a cap input can exceed the PIV of the 866.

3. My own experience over 25 years ago when I fried a NOS UTC destined for the 3CX1000A7 2M amp by running it into a 50uF cap and forgetting to attach the bleeder.

[W1ATR]

You never know what you are getting with a BC1-F, although stock iron is bound to be heavy.  I have seen pictures and descriptions of ones with what looks like Thordarson Tru-Fidelity iron, as well as Chicago iron, both types open frame/cast end-bell construction.  Wonder if that potted iron is stock or replacement?

I got the mod transformer  taken from one years ago (and still have it).  It is a UTC LS-691 that weighs in at 300 lbs, although I think that one may have been a replacement for  the original.

If I had to take a swing at a guess, I would say the iron I have was a replacement. The part numbers line up with my F manual, but my iron looked a lot newer than everything else in there. The mod trans feels like it's pushing 200lbs+ and the smaller mod reactor is prob in the 125lb range. I still have all the other iron as well and it's all Chicago stuff including a pile of these little 10hy? potted chokes and two nice looking steel end cap swing chokes. When I picked this load up from my buddy Dave Schneider up in Kingston, he did make it a point to say not to run any DC right on the mod trans and to only use it with the reactor.

On a side note, I also have this reactor choke out of a 20v2 that is of that cast end bell construction. It's so heavy, I have it on it's own little dolly in the corner of my shop.  

[k4kyv]

I have a couple of 845 to class-B grids driver transformers that came from BC1-Fs.  One is a potted Chicago unit and the other is a much heavier Thordarson open frame Tru-Fidelity unit with the cast iron end bells.

Dave Schneider up in Kingston, he did make it a point to say not to run any DC right on the mod trans and to only use it with the reactor.

Glad to hear Dave is still kicking.  He's the one I got my UTC LS-49 class-B driver transformer from, new in the box, probably 20 years ago.  I use it in my HF-300 rig to match a quad of 2A3s to a pair of 810s.

[k4kyv]

I have never heard of a plate transformer running hot because of insufficient load, or small wire size in the secondary winding causing it to run warm at minimum load. If you are aware of any technical references having been published on that subject, please let us know where this information can be found.

In no particular order:

1. http://www.w8ji.com/choke_input_power_supply.htm

2. QST articles in the 30's explaining that the start up voltage in a 2000-2500VDC supply using a cap input can exceed the PIV of the 866.

3. My own experience over 25 years ago when I fried a NOS UTC destined for the 3CX1000A7 2M amp by running it into a 50uF cap and forgetting to attach the bleeder.

Those are well-known problems.  Mercury vapour tubes don't like to work into capacitive input filters.  It's not so much the PIV as the peak current that eventually destroys them, although they don't tolerate excessive PIV any better than do solid state diodes.

By running the power supply with the UTC unloaded with no bleeder, you effectively ran it as capacitor input, even with the  choke in the circuit.  A choke input filter requires a reactor with at least a minimum critical inductance.  A rule of thumb for critical inductance is 1 henry for every 1000 ohms of load.  For example, if you have a 5-25 henry swinging choke, you need a bleeder resistor no higher than 25000 ohms, and make damn sure the inductance indeed reaches 25 henries with the current pulled by the bleeder.  In actual practice, the static plate current helps the bleeder along, so a 30K resistor or higher might appear to work.  But lose filament voltage on the tubes, and watch out.  Usually it is the power supply filter caps that go, because the output voltage soars from 0.9 X rms output voltage from the transformer to 1.414 times the voltage when the load is removed.

I don't think the problem was excessive heating in the plate transformer due to the size of the wire in the primary or secondary; a voltage spike caused the insulation to break down and short out the winding. With a proper rated filter choke, bleeder resistor and output capacitor in the circuit, you have nothing to worry about as far as blowing the plate transformer or rectifier regardless of the load, assuming the insulation in the transformer, choke and capacitor haven't deteriorated due to age, moisture invasion or faulty construction.

[KM1H]

Well, after a beautiful weather 540 mile round trip I wound up with a lot more than planned

I loaded up the wifes beautiful, well kept, 2007 Volvo S-260 Turbo AWD....I left her home . The rest included all the other transformers and chokes except 866 and 8008 filaments. SS replacements for those, the 833A caps, the 1000W dummy load, and a bunch of little stuff he tossed in the trunk when I wasnt looking. Extras were as cheap as what I went there for. The mod reactor is a replacement and much heavier than the mod xfmr. Its 60H @ 600ma, 10KV Hi-Pot and has a Gates label and part #.

Everything was unloaded yesterday eve and she hasnt noticed the windshield chip and slowly forming crack....yet....  Damn 18 wheeler in front of me slewed onto the shoulder during a huge wind gust on the NY Thruway and was peppering everone with sand and debris. No fun at 80mph.

Carl

[KM1H]

Those are well-known problems.  Mercury vapour tubes don't like to work into capacitive input filters.  It's not so much the PIV as the peak current that eventually destroys them, although they don't tolerate excessive PIV any better than do solid state diodes.

Excessive peak current plus high winding resistance equals burnt wire insulation which equals smoke.

By running the power supply with the UTC unloaded with no bleeder, you effectively ran it as capacitor input, even with the  choke in the circuit.  A choke input filter requires a reactor with at least a minimum critical inductance.

I didnt say anything about a choke since there wasnt one. As far as the minimum L those 30's QST's covered that in depth with graphs and formulas. Too bad the HB's didnt follow suit.

In actual practice, the static plate current helps the bleeder along, so a 30K resistor or higher might appear to work.

 

There is none in a ham RF amp which is biased beyond cutoff in receive or standby. In fact both the RF and mod tubes will be cutoff in receive; I guess some just kill the HV.

The Gates swinger appears to run at 27ma, 2700V @ 100K bleed in the BC-1T with the RF tubes providing a full load at a minimum. In a ham TX that bleed will have to be increased to ~25K to obtain the minimum 100 ma specified on the swinger label.  Its not a simple plug and play obviously when using a BC TX for parts. I havent decided which way Im going with the PS yet; maybe its better to use the Gates iron

[k4kyv]

Those are well-known problems.  Mercury vapour tubes don't like to work into capacitive input filters.  It's not so much the PIV as the peak current that eventually destroys them, although they don't tolerate excessive PIV any better than do solid state diodes.

Excessive peak current plus high winding resistance equals burnt wire insulation which equals smoke.

Peak current is not what overheats the transformer.  Solid state rectifiers will withstand a certain amount of excessive peak current but that will shorten the life of MV rectifiers.  It is the rms current pulled through the winding that causes the temperature rise in a transformer. Since the Gates iron is designed to be used with a choke, it may not work satisfactorily with capacitive input (excessive voltage sag) even if the rms current pulled through it is not enough to cause it to overheat.

I didn't say anything about a choke since there wasn't one. As far as the minimum L those 30's QST's covered that in depth with graphs and formulas. Too bad the HB's didnt follow suit.

There were some good articles in G-E Ham News and Radio on that subject. The older ARRL handbooks say little or nothing about dynamic voltage regulation either.

In actual practice, the static plate current helps the bleeder along, so a 30K resistor or higher might appear to work.

 

There is none in a ham RF amp which is biased beyond cutoff in receive or standby. In fact both the RF and mod tubes will be cutoff in receive; I guess some just kill the HV.

That's what I do.  I don't like the idea of the unused HV lurking around when I'm not transmitting.  Beside the hazards it poses to human safety and transmitter component life, it may also generate noise in the receiver that can interfere with weak signals. In any case you should always have sufficient bleeder current with just the resistor. Never rely on the tubes to maintain the minimum load on the power supply.  Loss of load = soaring plate voltage = crapped out filter cap and other components, possibly including iron.

The Gates swinger appears to run at 27ma, 2700V @ 100K bleed in the BC-1T with the RF tubes providing a full load at a minimum. In a ham TX that bleed will have to be increased to ~25K to obtain the minimum 100 ma specified on the swinger label.  Its not a simple plug and play obviously when using a BC TX for parts. I havent decided which way Im going with the PS yet; maybe its better to use the Gates iron

If you plan to use the rig on AM-only with common power supply, the RF final plate current should keep the PS constantly loaded down sufficiently with the stock Gates components just as in the original transmitter.  If you plan to use it on CW or SSB, then you will need to use a much lower resistance bleeder.  As I found out with my BC1-T, you will also need to increase the filter cap from 8 mfd to a minimum of about 25 mfd to keep the plate voltage from bouncing all over the place with the varying load. I would use as much capacitance as possible without having to resort to step-start.

The stock Gates doesn't even use a protective bias supply for the PA.  They rely on the overload relay to kick out in case of loss of rf drive to the final or loss of modulator bias. Over the years after several catastrophic failures, I have finally equipped both homebrew rigs with reliable overload relay protection as well. I added protective bias to the RF final so I could use it on CW.  To maintain critical inductance in the power supply during key-up, I switch in the modulation reactor for double duty to provide additional filter choke inductance in CW mode. Using only that 5-15 henry swinger, I would need to drop to a 15K bleeder to maintain critical inductance during key-up.

My HF-300 uses a pair of identical power supplies for mod and final, each with a pair of 872-A or 4B32 tube type rectifiers. The chokes are approximately 8-30 henry swingers, with 28 mfd of filter.  One uses a 25K bleeder and other a 30K. Static voltage regulation (per KV meters) in both supplies shows < 10% variation from full load to no load.  The class-B modulator supply voltage hardly budges with voice modulation because of the short duty cycle of the load pulled by the class B modulator combined with the amount of filter capacitance, just as in the case of a slopbucket leen-yar. If I maintain a sustained whistle into the mic or inject a tone from the signal generator, I can see the voltage sag.

[KM1H]

Thanks for all Don.

This will be AM only, I have enough vintage and modern SSB/CW gear and amps.

I suppose the 32V2 will be the driver after I go thru it and clean up the poor Collins style audio.

Carl

[The Slab Bacon]

I have the plate iron from a BC1-x (I dont know which variant it was) in my big HB rig. It would make around 2800v into the bleeder stack in choke inpoot. It was just not enough to get what I wanted out of the 4X1. I changed it to cap inpoot (C-L-C). This got the HV up to somewhere around 3500v+ into the bleeder stack. and it only drops down to around 3300 + under the 400mA resting carrier mode. It keepz the 4X1 singing for itz suppa!!

It is one of those binoclear stiled transfomas somewhere right around 90-100 lbs.
It is pretty beefy and doesnt sag very much unda a load!!

It also comes right up to the T.I.T. standards. (Timtron Institute of Technology) At 10w per pound of weight. It can run all night and never get very warm. (I grid block the 4X1 in standby and leave the HV up and running.) I just feel that that is easier on everything than constantly nailing the plate iron on and off.

[KM1H]

That iron is about 40 Ohms either side of CT which isnt bad as old iron goes. The primary is on a par with a Dahl for a similar CCS kVA which agrees with your good regulation results I have some el cheapo hammy grade iron that is less voltage and twice the R. Now that crap will sag.