[ARC5] PS Hum in BC-453-B (Was Converters for Q5-ers)

[Leslie Smith]

  Ken & Other list readers,
  This thread has taken a life of it's own.  I changed the subject to
  reflect that.
  I have a good supply of places to look, and a number of measurements I
  need to make.

  #1  False trail (deduction)  Listening to a speaker I heard no hum.
  Using (modern) headphones I hear hum.
  I assumed the hum was always in the set; I assumed I hadn't heard it
  before now because I used (1) old WW-II vintage 'cans' and (after
  that) a low-fi speaker (fed by a main power transformer to match the
  600 ohm "Z" to the 8 ohm speaker "Z".  
  Wrong.  The hum is NOT in the set.  The hum is in the P.S.  (see #2).

  (All this reminds me:  Why did the bee hum?  Answer:  It didn't know
  the words.  Just like the Aussie cricket team, when they sing "Advance
  Australia Fair".)
 
  #2.  Poor observation.  Recently I changed both the P.S. card AND the
  power transformer.
  Both are similar to the pair in use previously.  I will call the
  origin P.S. card and transformer PS-1 and T-1.
  I assumed the more recent PS-2 and T-2 were working the same as PS-1
  and T-1.  (They are similar, almost identical)
  Wrong.
  
  Measuring voltages around PS-1, the voltage fed to the zener chain was
  68V.
  (Matches Ken's calculation, as he wrote earlier.)
  Voltages in PS-2 are substantially DIFFERENT.  I can focus on the PS,
  not the set.

  Proof that this is so:  When I turn on the power, well before the
  filaments have had a chance to warm up, and no valve is amplifying  -
  I hear hum in my cans.
  (The hum is there using the speaker - I just don't hear it with the
  lo-fi gear.)
  
  #3.  Voltage to the zener string.  (See circuit is posted in the Yahoo
  group "ARC-5radio" files. )
  You will see I used x5 zeners, each 12V, in a string fed from a 180
  ohm resistor.
  I can report that the voltage at the 'top' of 180 ohm resistor should
  be 68V (as Ken calculated).
  I can report that the voltage on THIS card is much lower.  The fault
  lies in the PS, not the set.

  One benefit from this:  My set has an intermittent fault.  I reported
  originally that I had to replace the 12K8.
  Wrong.  I replaced the 12K8 and the intermittent disappeared.  Later
  the same fault re-appeared.
  After that I probed around the 12K8 valve socket (used multimeter
  probes).  
  The fault disappeared.   Now, the fault  appears more regularly.
  This means I'll be able to find it.  There is some good in every bad.
  

  73 de Les Smith
  vk2bcu at operamail.com


On Tue, Jan 20, 2015, at 16:06, Kenneth G. Gordon wrote:
> On 20 Jan 2015 at 12:59, Leslie Smith wrote:
> 
> > Filaments - 24V AC. B+ 60V, zener regulated, supplied by doubler
> > from 24V AC.
> 
> Hmmm...I tried operating a limiter using 24 VDC from the filament supply, 
> very well filtered, and had tons of hum. When I used a dropping resistor
> from 
> the B+ supply, I had no hum. A clue, perhaps?
> 
> Besides, your voltage doubler would output ~68 VDC: to attempt to
> regulate 
> that with Zeners to 60 VDC would not provide you with enough "overhead". 
> Try disconnecting the Zeners completely to see if the hum goes away.
> 
> If you are only using 470 MFD, you may not have enough filtering, either.
>  
> > I take the AF from the rear connector - J3, pin #2. This is (I
> > believe) 500 ohms.
> > 
> > The set is BC-453-B, not ARA. That's why I think - 600 ohms.
> 
> It may be. As I remember it, the output transformers of that model have
> two 
> output taps: one at something like 4K and the other at 600 ohms, but it
> would 
> depend on how it is connected as to whether or not your particular
> receiver is 
> actually set to 600 ohms. Normally, as I recall it, the BC-453B was
> always 
> set up for 600 ohms....except when it wasn't.
> 
> > (3) Until recently I fed the AF into a power transformer, 110V :
> >     11V. Small.
> > 
> > With this the AF lacks bass, and in the "middle" of the tuning is
> > muffled.
> 
> Hmmm...that sounds like an odd result to me...I don't know how to explain 
> that...
> 
> > (5) Problem. Heaps of hum. 50 or 100Hz hum (I can't tell the
> >     difference).
> > 
> > The bottom plate from the set is missing.
> 
> Doesn't matter.
> 
> > DC filament supply v AC filament supply. Does AC to the filaments
> > matter?
> 
> No. Not one whit. Absolutely not.
>  
> > Hi-Z signal path between the (diode) detector & the 12A6 grid.
> > How much hum is picked up here.
> 
> Next to none.
> 
> > B+ supply, from voltage doubler. 1/2 wave doubler, followed by x5 12V
> > zeners in parallel with 470uF filter cap. (always thought 470uF would
> > take out a lot of hum!)
> 
> At ~68 VDC, 470 mfd is not nearly enough filtering, IMHO.
> 
> I have seen a chart which listed the amount of filtering needed at what 
> voltage for so much ripple, but I don't know where it is at the moment.
> 
> I also just ran a program which calculates the minimum capacitance needed 
> for 68 mV ripple at 68 VDC at 50 mA: that value is ~3700 uFd.
> 
> As I said, in my opinion, 470 uFd is not NEARLY enough to keep the ripple 
> to a reasonable value.
> 
> At 470 uFd at 68 VDC at 50 mA, your ripple would be 3.5 volts, according
> to 
> my calculations. I am very curious how my calculations stack up against a 
> 'scope view of the output. In any case, 3.5 V PP ripple is too much. That
> is 
> about 5.8%, which in a receiver is too high.
> 
> > Where should I start looking for the hum?
> 
> In the output of your ~68 VDC B+ supply...
> 
> I would try the following: 
> 
> 1) increase your filter capacitor to at least 3700 uFd. and 
> 
> 2) for now, get rid of the Zeners completely.
> 
> I think that if you want to use Zeners to regulate the B+ (which, in my 
> opinion, is a very valuable and worthwhile effort), you may have to go to
> a 
> voltage tripler to get enough overhead....but the necessary overhead can
> be 
> calculated too.
> 
> > First thought: Get a 24V battery and build a 'clever' voltage doubling
> > circuit to supply 50V B+. That way I can cart the whole she-bang into the
> > middle of a paddock and see if the hum is still there ....
> 
> Or use a dynamotor... 
> 
> Ken W7EKB

-- 
http://www.fastmail.com - The way an email service should be