I have almost the same question, but from two different different angles...considering a multi-winding transformer's specs for intended rectifier configurations, when seeking to use for something else.
Long post, hopefully of interest to someone besides myself, and I promise no discussion of what color my socks or underwear are today.
I have many transformers that have specs, and I would like some opinions, corrections or sanity checking on thoughts about repurposing. The specs were from drawings, with a 'test circuit' they were expected to satisfy. The test circuit, for obvious reasons, is simpler than the product(s) the transformer was installed into.
I often refer to Hammond's (and others) charts for different rectifier configurations and loads. They answer most questions that come up when considering 'translating' AC VA rating to various DC rectification applications. My initial 'comfort level' adjustment for FWB+filter cap repurposing of a small isolation transformer is VA/1.8. = W. Basically as a first check, paying attention to the size/capability of the transformer.
Second consideration is to pay attention to voltage and current components of VA or W and look at what rectifier topology was or will be used.
The specs often include a maximum temperature rise. When abusing/repurposing a transformer the 'by-design' test circuit, which can comprise multiple windings and loads, is often no longer applicable. Common sense indicates conservative use would be a good idea if one doesn't want to replace a transformer shortly after building something. So it's not possible to come up with a simple conversion or compensation ratio in some cases.
I have a couple scenarios that I think about because I don't find (or recognize) answers for certain rectifier 'transformations'.
I may be overthinking, but often use a FWB where an application originally used FWCT, or even HW (typically for very low currents).
I think it's worth contemplating these issues simply because using something for a purpose it wasn't made for should be considered.
Here's a specific example:
A (NOS) transformer with three two-wire secondaries (no taps).
Sec. #1: 248 vrms, FW voltage doubler (two axial rectifiers + 2 x 40 uF). Used to provide B+ for a RF transmitter. 700 VDC unloaded, 600 VDC minimum at 150 mA DC. Call that 90W max. DC load
Edit: corrected ac & DC V values - typed wrong from memory.
Sec. #2: 105 vrms, HWR+ a few 10's of uF. Produces 135 VDC @ 30 mA DC (says the drawing), about 4 W DC
Sec. #3: 6.3 vrms, 5.5 a ac. 34.7 secondary VA.
Repurposing thoughts:
Use Sec. #3 as intended for tube/valve filaments. (No00 question to pose).
Sec. #2: (Question at end) First thought - I rarely want to use HWR because 120 Hz (for me, 100 Hz for some of you) from FW rectification is easier to filter, and I give barely a passing second thought to some DC flux in the transformer core resulting from HWR. If designed for HWR, that is likely taken into consideration, if it's even significant. (For toroidal power transformers, this is sometimes listed as not to be done). Sometimes EI laminations are interleaved in different patterns which affects (slightly) stacking factor and thus distributed air gap. I found a paper giving numbers for this, but one transformer engineer I talked to said it's true, but insignificant. He typically was thinking about things like gaps in SE valve OPT's, and distributed gap in a PP OPT don't amount to anything useful (cannot use for SE).
Avg. DC voltage for HWR should be lower than FWB, but I often do not see as large a difference as expected. It took me a long time to figure out an LTspice simulation of a HWR bias supply including voltage divider (R) vs. hand calculations because I had pushed HWR so far to the back of my mind, my spreadsheet didn't match my simulation. I figured it out eventually.
So, if one uses a FWB instead of HWR, elimination of some small amount of DC flux would be a tiny benefit of flux density headroom in the core. I don't think this benefits the VA or DC W 'rating'. But keeping in mind transformers are 'VA-limited', if a higher filtered DC voltage (ignoring an extra diode drop for FWR) results from FWB+filter cap, then the available DC current from that winding might be lower for the same DC W product. :O(. Is there any known benefit from using FWB in place of HWR other than easier to filter 2x ripple frequency and a trivial option of 'selecting' one of two DC voltage options by choosing the alternative rectifier?
Sec.#1: Main question here is what to consider what happens from using a winding specified for a voltage doubler, instead with a FWB. SOME increase in DC load current should result, but probably not quite double, keeping copper losses in mind. The 248 vrms with FWB would produce about 350VDC unloaded compared to the 700 VDC NL with doubler. The copper losses (I^2R) would be higher (4x for 2x current) when hypothetically considering 300 mA DC for FWB vs. 150 mA for FWVD, and the CM/A number for the unknown wire gauge of that secondary would double. CM/A seems to be a looser parameter, discussed as being usable over a wide range depending on the design. I assume that would still be a copper losses phenomenon and not a magnetic phenomenon. I've talked myself into rationalizing that using a FWB instead of a FW voltage doubler with the potentially available current only multiplied by 2^0.5. (x2, /1.414). That winding was intended to deliver the largest proportion of secondary power by the 3 secondaries. Anyone else have personal 'rules' they apply to repurpose transformers along these lines?
I have cartons of another NOS transformer with a 154 vrms @ 0.45 A DC (FWB), and 6.3 vrms @ 14 a rms. That one I want to use with a doubler instead of FWB, and will simply halve the DC load current rating.
Thank you
Murray
