HomeRoast Digest


Topic: Dimmer plus diode caution (18 msgs / 355 lines)
1) From: Ken Mary
I have been successfully using a diode across a dimmer to control fan speed
on a WBI. However, using the same setup to feed a transformer for poppers
with dc motors has given me a problem. The transformer would get very hot.
Today the transformer shorted out with a loud bang, of course in the middle
of a roast. Apparently the heat destroyed the insulation on the primary
side. I had been using this same transformer for almost 2 years, with
minimal heatup even running for 20 minutes. When I added the diode, the
transformer got hot within a minute. This could be coincidence, but I will
not know until I can buy another.
--
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

2) From: floyd burton
At least no fire and u did not loose 5# of beans like I guy I know who ran
out of gas during a roast.

3) From: dewardh
floyd:
<Snip>
minimal heatup even running for 20 minutes.
Had you been using it with a dimmer before?  Dimmers are generally not used 
with transformer loads because of overheating in the transformer caused by the 
high frequency components of the dimmer output . . .
Deward
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

4) From: Rick Farris
Ken wrote:
<Snip>
No coincidence, pal, you can't run DC through a transformer.
-- Rick
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

5) From: Rick Farris
Deward wrote:
<Snip>
But that's not the problem *he* is seeing.  He added a diode, which means
that he's running a large DC component through the xfmr, causing the core to
saturate, causing the impedance to drop, causing the current to rise,
causing the core to saturate even worse, lowering the impedance further,
causing a greater current flow....well, I think you get the idea.
-- Rick
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

6) From: dewardh
Rick:
Maybe.  That's why I asked if he'd used the dimmer with the transformer before, 
without a diode. Have you actually tried this and measured the effect?  If 
that's what indeed happens then a simple solution might be to put the diode in 
the transformer secondary.  But it's still generally not advised to drive 
transformers with dimmers anyway (I hadn't contemplated it when I suggested the 
"fix", but was only talking about 120V brush (AC/DC) motors, which have no 
problem at all with partially rectified drive).
Deward
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

7) From: Rick Farris
I thought this part made it perfectly clear:
-----
Ken wrote:
<Snip>
------
:-)
(I used to work in a transformer factory.  I don't remember all that much
about them, but I remember enough to know you can't push DC through a
transformer.)
-- Rick

8) From: dewardh
Rick:
<Snip>
about them, but I remember enough to know you can't push DC through a
transformer.)
Ah . . . I see . . . ever hear of a "spark coil" (like in your car?)?  DC in 
the primary.  Pulsed DC, to be sure (and "other things" going on), but things 
are not quite as simple as you make them out to be (which is not to say that 
"DC" in transformers is a good idea, but there are many cases where it is not a 
bad idea, or may even be a deliberate part of a design, as with audio output 
transformers in vacuum tube amplifiers).  But I've never worked in a 
transformer factory, and I've never tried pulsed (half cycle) DC in a low 
voltage power transformer, that I recall, anyway (just in little DC to DC 
converters, where it works fine if the time constant of the transformer core is 
correct).  That's why I asked for more details, and if you'd tried it to see . 
.. .
Deward
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

9) From: Jeffrey A. Bertoia
Deward
One thing is absolutely certain anytime that you put a DC bias, that 
exceeds it's ratings, on an iron transformer core
 it _will_ saturate, the primary and the secondary _will_ decouple and 
the current in the primary _will_ rise to values
 many times rated, the 'decoupled' energy is lost in the core therefore 
making it hot _very_ quickly.  This _will_
destroy the transformer if not externally limited.
On the other hand, the high frequency components of phase control will 
cause somewhat higher core losses and
therefore heating but many (even most) times it is not detrimental, 
especially if the application, of the transformer
in the design, is properly derated.  Most derating specifications 
suggest 20% derating on stacked core transformers
and an additional 10-20% in phase control applications.
jeff - who has fried too many 100+ KVA transformers
dewardh wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

10) From: Jeffrey A. Bertoia
OOPs  sent this off a little too quickly...
I forgot to mention that special purpose transformers (not 'standard' 
power transformers) use many methods to
increase the allowable dc bias on the core.  These techniques include 
gapping of the core and material selection
for less 'memory' of the bias.  In general these techniques reduce the 
efficiency of the power transfer and increase
the construction cost.
Also one must remember that DC offset as applied to transformer cores 
refers to voltage and not current.  That
is why Rick's suggestion of moving the diode to the secondary is a good one!
jeff
Jeffrey A. Bertoia wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

11) From: Ken Mary
<Snip>
Yes, two years with dimmer, no problem. Two weeks with the diode, kaboom.
I am going to check the waveform of the dimmer/diode on the WBI. The primary
coil showed overheating, the secondary was fine.
<Snip>
Please explain the benefits of putting a diode in the secondary circuit. The
25 volts ac passes through a bridge rectifier to run the dc fan motor.
--
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

12) From: Ken Mary
<Snip>
If diodes pass current in one direction (half the ac cycle), where does the
dc originate?
--
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

13) From: Jeffrey A. Bertoia
Ken
See my second post it's the unbalenced voltage that is the problem.  The 
DC current is OK.
Since the diode in the secondary allows the transformer core to 'see' AC 
voltage everything
should work fine.
jeff
Ken Mary wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

14) From: Ken Mary
<Snip>
There is nothing obviously wrong with the dimmer with diode bypass when used
on an ac brush motor like that in a WBI. There were no voltage spikes
visible anywhere on the oscilloscope. The diode-passed half cycle was
normal.
--
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

15) From: dewardh
Jeff:
All true . . . but does not answer the questions "what about pulsed (1/2 cycle 
AC) DC?", and "what is the "rating" of the typical small power transformer in 
this regard?".  In any case I don't regard it as a good idea to drive a 
transformer with a dimmer or a diode (unless it's specifically designed for it, 
like the pulse transformers in some DC to DC converters are designed for that 
service) . . . but I am curious what one can "get away with".
Deward
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

16) From: dewardh
Ken:
<Snip>
Then Rick (and jeff) are correct that the DC "bias" is saturating the 
transformer.
<Snip>
25 volts ac passes through a bridge rectifier to run the dc fan motor.
As I think about it . . . there are none.  It would not accomplish what the 
diode across the dimmer would acomplish.
Deward
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

17) From: Rick Farris
Things are *exactly* as simple as I say when one is discussing transformers
designed for sine-wave input.  Which one is.
-- Rick

18) From: Rick Farris
Ken asked:
<Snip>
In the half-wave rectification transformation applied by the diode.
-- Rick
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast


HomeRoast Digest