[Amps] AL1500 PSU design

[pat_f6irf]

Tom W8JI a écrit :

>
> I don't understand how you could charge the capacitors through 1M ohm 
> 6 watts when the bleeder resistance is 400k ohms 56 watts. This would 
> only charge the caps to 1/3 voltage.

Hi Tom
The new cap. assembly  bleeder is 1.2 Mohms (12x 100k) so I was 
expecting the condensers to charge roughly at half the voltage - just 
for a first try... But I did not get the opportunity for a second attempt !

> Also what was the voltage rating of the 2W resistors? You would have 
> 1200V across each. Are they rated to handle that?

Don't know the ratings as they are in a drawer for 25 years or so... but 
they did not blow up... that's for sure...

>> I guess that the rectifier bridge has been damaged by the first 
>> "bang" (noticed
>> some yellow marks on a few ceramic disk capacitors), and that everything
>> has to be replaced on this board before trying again...
>
>
> That is a VERY rare failure. The reverse voltage is not an issue since 
> it is essentially the dc output of the supply, and current reserves 
> are good. Perhaps someone is uing the wrong fuse. Maybe a low voltage 
> rated fuse, a slow blow fuse, or too much current It is also possible 
> but rare there is a HV winding to earth failure, but that alone would 
> almost never damage the rectifiers unless the fuse was the wrong type.

Don't know the original specs of the fuses - and it is probably 
impossible to find them here, I just have 16A "domestic type" on the 
mains... anyway the 16A fuse is never burning, since the the resistor in 
the primary  is going away before the relay has the opportunity  to 
connect the primary directly to the mains...
Concerning a potential transformer HV winding to earth - not sure but I 
guess that it would show-up when testing the transformer alone... 
(tested alone the transformer delivers the expected AC voltage...)

> The original caps are 600V foil that is aged to 525V if they are still 
> using the same part. They are custom aged and tested for leakage at 
> the manufacturer, not bulk off-shelf parts. Of course it is possible 
> they have changed vendors, but voltage is a small influence on life 
> compared to temperature. The temperature is low in that area so the 
> life has always been acceptable. What will ruin a capacitor very fast 
> is an open equalizing resistor or a backwards capacitor.

The original caps are Philips 3186BE211T4450AMA3 - they are rated for 
450V service / 525V surge

>> - The original rectifier bridge is designed for 5000V PIV (5x1N5408 
>> in each
>> branch), which looks also unsufficient taking into account that a normal
>> rectifier design would require a PIV of twice the AC peak voltage 
>> (~6600V).
>
>
> Not true. It is a FW bridge with capacitor input.

OK - my mistake, 5000V PIV should handle the peak voltage in this bridge 
configuration.
but maybe "with capacitor input " is the key ??? what may happen if the 
rectifier is not connected directly to the capacitors ??? (IMHO nothing, 
but maybe I neglected a parameter ?)

>> As a first step, I am thinking about replacing the 1N5408 by BY255 
>> (1300V PIV -
>> 3A). Any comment / ideas / suggestions ?
>
 >Yes, find the real problem first. I do not understand how you tested 
or charged the electrolytics with a series 1M resistance. I do not 
understand how the diodes could blow running into 1 meg ohm, or even 
 >into a short. Perhaps it was earlier diode damage, but that is VERY rare.

Rare, maybe...  I am just trying to understand what happened in mine, 
and what I should replace, to get it running...  had tested the diodes 
one by one and they were "apparently" all OK before I applied the mains 
the 2nd time...

 >The diodes are one of the least likely failures in that 20 year old 
design unless the fuses are the wrong type. Most common is people use 
32V rated fuses, but sometimes they use slow blow or too much >current. 
Capacitors are only likely when a equalizing resistor opens.

Just facts: the rectifier bridge blew up twice with no load (tube and 
eveything disconnected on the second attempt).
Don't know what caused the first bang  (I guess a badly aging 
capacitor)- but the amp was in its original configuration... and the PSU 
blew up before I had the opportunity to apply a single watt at the Amp 
input... (as also mentionned one of the bleeder resistor was found 
"gone", but impossible to know if it has been the cause or the 
consequence...)

>
> I have an AL1500 I have been using since 1986. I have been running the 
> transformer on 250VAC into the 220V winding, and use 20A 250V fuses. I 
> haven't had a power supply problem in 21 years even though I use it 
> almost every day.
> 73 Tom

OK Tom , I have a homebrew  PA using a pair of 4CX250, that I built up 
some 20 years ago and that has thousands of hours of service using the 
same original pair of second-hand tubes. Just in 2005 I have done nearly 
20,000 QSO's in RTTY-contests  @500W with it... That's not the question; 
I am just trying to understand what caused the first failure, and what 
should I do since the first repair attempt ended up with a new rectifier 
failure... Rare or not I don't really care as it just happenned twice  
to me !

regards
Patrick