ad: Radclub22-1

HP-23A transformer core losses?

Discussion in '"Boat Anchor" & Classic Equipment' started by K1ZJH, Oct 7, 2017.

ad: L-HROutlet
ad: l-rl
ad: Subscribe
ad: K3QNTad-1
ad: l-gcopper
ad: L-rfparts
  1. K1ZJH

    K1ZJH Ham Member QRZ Page

    I'm upgrading two of the shop's HP-23 power supplies to the K8GNZ PCB modification.

    I noticed that the transformer idle current on my HP-23A supply is over 450 mA at 125 VAC
    primary voltage, and the idling current drops rapidly below 120 VAC. I suspect there are
    some core losses involved, and I am curious if anyone else has ever bothered to measure the
    no-load AC current on the dual primary 120/240) versions of these supplies. The earlier 120 volt
    only HP-23 is fine at 125 volts. I'm not sure if this is normal for that transformer, or if it
    has a defect. I'm letting it cook on the bench on a fused AC supply overnight to see what happens.

    Pete
     
  2. K1ZJH

    K1ZJH Ham Member QRZ Page

    As a note, adding 4 or 5 uF across the primary does drop the idle current.
     
  3. K9STH

    K9STH Ham Member QRZ Page

    Pretty typical for the HP-23- series. Remember, there are resistors across the electrolytic capacitors and there is a load from both the high B+ and the low B+ as well as from the bias supply.

    I "assume" that you are turning on the power supply using a jumper in the 11-pin connector and not with it connected to any unit.

    Glen, K9STH
     
  4. K1ZJH

    K1ZJH Ham Member QRZ Page

    No, there is no load--the secondary windings were lifted for verification. Also, as noted, the rate in which the idle current climbs above 110 VAC is more than expected. The core is apparently going into saturation.

    This abnormality is not evident in my HP-23 but is present in the HP-23A which uses a different transformer. I am curious if this a problem with my supply, or characteristic to the later transformers in general.

    Pete
     
    Last edited: Oct 7, 2017
  5. WA6MCL

    WA6MCL Ham Member QRZ Page

    Worth mentioning how much the B+ (not HV) drops with the choke in line. I have rebuilt several with the kits around. You know, the SB and HW radios could run on 275-325 VDC quite well and lower output. Then again the 800V HV might a bit high over the 750V HV. How do they look loaded and unloaded?

    Vic WA6MCL
     
  6. KM1H

    KM1H Subscriber QRZ Page

    That sounds rather steep for magnetizing current Pete and reminds me of some Hallicrafters and National iron that are into partial saturation at 120-121V.
     
  7. N2EY

    N2EY Ham Member QRZ Page

    How warm does the transformer get with no load?
     
  8. K1ZJH

    K1ZJH Ham Member QRZ Page

    Comfortably warm after several hours... it has been suggested that the transformer might more efficient under
    load.

    I did measure the idling current with a thermal-converter based meter (HP 3403A) and noticed that adding a
    7 uF motor run cap across the winding for the low voltage B+ drops the current by over 50 percent. Weird.
     
  9. SM0AOM

    SM0AOM Ham Member QRZ Page

    By definition, the efficiency of an unloaded transformer is 0 %.
    It appears that the core area is marginal for the full primary voltage.

    In idle, the magnitude of the primary current is determined by the ratio between primary voltage and
    primary reactance+the equivalent copper and iron losses reflected to the primary. Ideally, the
    phase angle should be 90 degrees inductive, but practical transformers present an angle less than 90 degrees,
    as the losses introduce a real part in the impedance shown by the winding.

    It is only the I^2 R part of the apparent power that will actually heat the transformer during idle.

    The reason that a capacitor across the windings reduces the idle current is that a parallel resonance
    condition is created that reflects a higher impedance towards the primary terminals in the unloaded conditions.

    Equivalent-Circuit-of-Transformer-referred-to-Primary.gif

    What happens is that parallel equivalent inductance Xo is partially tuned out by the capacitor which reduces Io, assuming Ro is large compared to Xo.

    The core peak flux will be reduced somewhat when the transformer becomes fully loaded, reducing the core loss fraction of the total,
    but instead introducing more I^2 R losses in both the primary and secondaries.

    73/
    Karl-Arne
    SM0AOM
     
    N2EY and KM1H like this.

Share This Page