Yep, the input bridge is good... I have a perfectly solid 120VDC.
However, it dissapears fairly quickly when the switcher turns off. If I
lift the horizontal oscillator so the HOT never turns on, I can verify
that I have 120VDC with a minimal load (support circuitry load only)
What I was describing in that very first email is that I can lift the
base of Q5 and the B+ rail (emitter of Q5) falls to a nice steady 36VDC
(instead of 120V) and I can run without tripping the switcher into
overload. With the B+ at 36VDC and the HOT running Im only drawing 105ma
from the switcher. If I put the 120V B+ back up, the switcher shuts off
right around 1.1A load (which is about right given the main fuse (F1) is
a 1A SB).
Im going to put square wave generator onto 1-4 now and see what I see
with the flyback out of circuit. I just pulled it.
Im pretty sure that the transformers are bad after thinking through all
of this. When Im running it with 36V on the B+, my output voltages
aren't the correct ratios which points to something inside the
transformer affecting the Q and hence messing with the charge-discharge
behavior.
jess
> Have you checked your input bridge yet (4x 1N4004) as well as the 12V
> regulator U4(78L12) (not too likely, but if shorted then Q7 is also
> bad) for damage from the un-isolated 120VAC power test you did? PTC1
> should be fine - no connection to ground there. Pg 54 of schematic PDF.
>
> I would separate the flyback from the circuit and then hook up pins 1
> (common) and 4 (pulsing DC) to some AC supply using a bridge rectifier
> to get pulsing DC running at the correct frequency (otherwise the
> transformer is not efficient - results will be skewed lower). Hook a
> scope up to pins 1 & 4 to see the picture, it should be fairly clean.
>
> Then scope pin 5 to see if you have a bit of a boost on the pulsing
> DC, next scope across pins 7 (probe) & 9 (common) should show AC about
> 15-20% of the original input voltage (120/24), pins 6 & 9 showing
> around 5% (filament of 6.3VAC). You can also use a HV probe (if you
> have one) for the HV output, again it should be the ratio of your test
> voltage to 120V.
>
> John :-#)#
On 1/17/2011 12:36 PM, John Robertson wrote:
> Jess Askey wrote:
>> From what I read, I figure that if there was a short the Q would be
>> lowered enough that I wouldn't get any ringing... even in circuit.
>> However, with that said, since my ringing is drastically *lower* than
>> I expect (I thought there should be a 300V oscillation on the
>> collector with a 36V supply), something is dragging this down.... so
>> I think it is simply ringing at the expense of excessive current
>> because the Q *is* lower.
>>
>> I will try the low voltage out of circuit ring test today and report
>> back... with the low amounts of current, Im guessing that I will not
>> get a *good* ring out of it.
>>
>> jess
>>
>>
>> On 1/17/2011 12:56 AM, Rodger Boots wrote:
>>>
>>> John, it's a raster monitor, not vector.
>>>
>>> So why is it ringing in the first place? Is the damper diode
>>> connected? Is the horizonal output transistor the correct part? Is
>>> the correct yoke winding connected? Is the filter capacitor on the
>>> power supply to the output transformer working (should be DC to the
>>> transformer, very little AC).
>>>
>> ---------
>
>
>
>
---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Mon Jan 17 14:50:03 2011
This archive was generated by hypermail 2.1.8 : Mon Jan 17 2011 - 15:50:00 EST