1
\$\begingroup\$

Here the UC3844B High Performance Current Mode Controller (at left, SO-8) is used with a custom transformer to step down bus voltage to 24V. The transformer is likely rated for about 40W, so estimating 1.6A:

Big "S" brand board, PSU section

On this working board, the transformer was removed and was measured with an LCR meter as:

  • Primary: 1.8mH, 143Ω
  • Auxiliary: 40µH, 7.4Ω
  • Secondary: 40µH, 6.2Ω

Now on another board, this transformer measures radically different than the first one:

  • Primary: 1.84mH, 12.8Ω
  • Auxiliary: 30µH, 10.0Ω
  • Secondary: 40µH, 9.1Ω

Now the inductances are very similar, but what is going on with the primary resistance?

Was the first transformer "marginal" or bad and just happened to appear working? Because that board seemed to work fine and output the required current without overheating.

THIS new board (with the much lower resistance primary) is happily outputting 68VDC (and popping downstream components.)

Is it possible that both transformers are perfectly fine (one just has a much higher Q-factor than the other), and the UC3844B was adjusted for both?

Edit: 700V bus and LCR meter at 10kHz for both sets of data.

\$\endgroup\$
3
  • 2
    \$\begingroup\$ LCR meter at what frequency? \$\endgroup\$ Jun 15, 2022 at 0:44
  • \$\begingroup\$ Thinking out loud, if 143Ω was the uni-filar resistance, deca-filar would be 14.3Ω, so not "original had most of the strands open." Perhaps the new transformer is using a copper ribbon primary or Litz wire instead of traditional stranded? They're potted of course. \$\endgroup\$
    – rdtsc
    Jun 15, 2022 at 13:18
  • 1
    \$\begingroup\$ Hm, those resistances don't make any sense, whether series or parallel equivalent, at that frequency. Well, 12.8 ohm primary seems plausible, but the rest, something's missing... \$\endgroup\$ Jun 15, 2022 at 14:34

1 Answer 1

0
\$\begingroup\$

The regulation issue of the new board ended up being an open 15V Zener in the 30V "limit" circuitry. With that Zener in place, this unit happily drives 24V under full load.

So to answer my question, yes, 143Ω and 12.8Ω primaries both can be allowable for the UC3844B. I'm not sure how the transformers differ internally, but both seem to work well. The only clue is that the new transformer had some ridges underneath, where the original one did not - so apparently there is "more" in the newer ones.

Incidentally, there are some different component values on this new board, so apparently the manufacturer knew of this transformer change and compensated for it. The primary pulses are definitely shorter and less frequent on this board as the primary current ramps up quicker and there is less resistive loss. The layout is identical, so they did this without revising the board itself. +1 to the designer for efficiency optimization.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.