Scott LK150 - Eating Filter Caps

[pf...@aol.com]

Here is the Schematic: http://komkris4000.webs.com/lk150%20schematic.gif

It is eating the 150uF @ 75V filter caps - I replaced the OEMs first with 75V, 105F caps, they lasted about 20 hours. Then 100V, they lasted about 30 hours. The manifestation is increased HUMMmm, and the caps start to swell.

I am contemplating using 180V caps, but before I do so, any suggestions from the assembled multitude?

NOTE: All other voltages are correct and clean. When these caps are new, the amp is hum-free. As they age, a pronounced *60* cycle hum begins that appears to be volume dependent - that is, the hum goes down as volume increases.

The hum also remains with shorted inputs, and whichever way it is plugged in. US, nominal 120V/60hz single-phase (actual is 117V at the wallplate).

Thanks in advance!

Peter Wieck
Melrose Park, PA

[jf...@my-deja.com]

Is it possible that you are putting them in backwards?

[dansabr...@yahoo.com]

On Thursday, March 3, 2016 at 11:31:20 AM UTC-5, jf...@my-deja.com wrote:
> Is it possible that you are putting them in backwards?

Check the diode, I would guess that there is AC leaking through and perhaps the caps are the only filter of AC with no rectification at all.

Dan

[pf...@aol.com]

On Thursday, March 3, 2016 at 11:31:20 AM UTC-5, jf...@my-deja.com wrote:

> Is it possible that you are putting them in backwards?

No, that was the _first_ thing I checked. Also the second, third and last. But, as stranger things have happened, and I am by no means perfect, I will check again!

Dan:

I will check the diode. That was replaced pre-me, so a valid concern.

Thank you both.

[dansabr...@yahoo.com]

Verify that the diode is installed correctly too. If reversed, it would affect the caps too.

Dan

[tom]

<pf...@aol.com> wrote in message
news:eab85b47-2f5c-441d...@googlegroups.com...

And measure the 1.5K 2W. Is the transformer CT actually at ground? What is
the voltage across the caps? Across the 220 ohm R? What do you see (with a
scope) across the caps?

[Phil Allison]

pf...@aol.com wrote:

> Here is the Schematic: http://komkris4000.webs.com/lk150%20schematic.gif
>
> It is eating the 150uF @ 75V filter caps - I replaced the OEMs first with 75V, 105F caps, they lasted about 20 hours. Then 100V, they lasted about 30 hours. The manifestation is increased HUMMmm, and the caps start to swell.

** That schem looks wrong.

No way 75V electros should be fed from a 100V AC winding.

75uF @ 150V is more like it.

Is the secondary tapping really 100V rms ?

The hand written " -31 " is wrong too, should be more like -45V.


.... Phil

[legg]

It's possible that the original rectifier wasn't a silicon diode, and
that an inaccurate replacement has been used.

Should be easy enough to measure the voltage on the caps to determine
stress levels,

RL

[pf...@aol.com]

Phil:

I think you are right. In any case, the diode *was* in backwards - it is in a silly little brass shield grounded to the chassis - looks like a bit of brass tubing with a soldered on lead that covers the diode (1N4007 now, replacing a wrong-way 1N4004). I will check voltages with the existing slightly swollen caps for now, then replace them with appropriate voltage and capacity units - but no less than 180V at this point. As for testing purposes, I will be right there and with an eye on the current meter, I have no worries about a few minutes with bad caps.

[legg]

On Thu, 3 Mar 2016 08:01:41 -0800 (PST), "pf...@aol.com" <pf...@aol.com>
wrote:

Later rev schematics that employ silicon rectifiers in the -65V bias
have a 6K8 2W bleeder across the first filter cap.

RL

[jf...@my-deja.com]

On Friday, March 4, 2016 at 5:13:42 AM UTC-8, pf...@aol.com wrote:
> ...the diode *was* in backwards...
If you had positive bias on the control grids, you may have damaged the output tubes. Some of the cathode coating could have evaporated and gotten deposited onto the grids, leading to grid emission.

[Phil Allison]

pf...@aol.com wrote:
>
>
> NOTE: All other voltages are correct and clean.
> When these caps are new, the amp is hum-free.
> As they age, a pronounced *60* cycle hum begins
> that appears to be volume dependent - that is,
> the hum goes down as volume increases.
>
>

** The electros are leaking badly as they heat up - cos they are being over-voltaged. Even 10mA ( about 1W ) would get them damn hot.

So the negative bias voltage applied to the output tube grids has a high level of 60Hz ripple. If the 6550s were perfectly matched, this would not matter - but they aren't so some of this voltage get amplified.

I don't buy your story of the 1N4004 diode being reversed, 6550s draw over 500mA each with positive grid bias and the amp would have quickly blown up if that were the case.


... Phil

[pf...@aol.com]

On Friday, March 4, 2016 at 10:24:44 PM UTC-5, Phil Allison wrote:
> pf...@aol.com wrote:

> I don't buy your story of the 1N4004 diode being reversed, 6550s draw over 500mA each with positive grid bias and the amp would have quickly blown up if that were the case.
>
>
> ... Phil

Well, when I put the diode in correctly (per the schematic), the hum reduced greatly and the amp draws about 30 watt less than before. When new caps arrive, I will try them out. Could also be that the diode was bad in the first place, not sure as this is the first time I checked it.

[Boris Mohar]

It could be that the original diode was Selenium which has much larger Vf.
--
Boris

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[legg]

On Sat, 05 Mar 2016 09:11:39 -0500, Boris Mohar
<borism...@sympatico.ca> wrote:

>On Sat, 5 Mar 2016 05:01:45 -0800 (PST), "pf...@aol.com" <pf...@aol.com> wrote:
>
>>On Friday, March 4, 2016 at 10:24:44 PM UTC-5, Phil Allison wrote:
>>> pf...@aol.com wrote:
>>
>>> I don't buy your story of the 1N4004 diode being reversed, 6550s draw
> over 500mA each with positive grid bias and the amp would have quickly
> blown up if that were the case.
>>>
>>>
>>> ... Phil
>>
>>Well, when I put the diode in correctly (per the schematic), the hum
> reduced greatly and the amp draws about 30 watt less than before. When
> new caps arrive, I will try them out. Could also be that the diode was
> bad in the first place, not sure as this is the first time I checked it.
>>
>>Peter Wieck
>>Melrose Park, PA
>
>
>It could be that the original diode was Selenium which has much larger Vf.

This is clearly the case in images of this model and later schematics
also show a 2W 6K8 shunt resistor across the first filter cap.

RL

[Mike Tomlinson]

En el artículo <b04ec4db-a196-4f77...@googlegroups.com>,
pf...@aol.com <pf...@aol.com> escribió:

>Well, when I put the diode in correctly (per the schematic), the hum reduced
>greatly

Ignore Phil, he's a thick Australian* cunt. You're doing fine, carry on
as you are.

* I do apologise for the Australians, they're the result of criminals,
murderers and rapists transported to van Diemen's land and the product
of enthusiastic interbreeding ever since.

--
(\_/)
(='.'=) Bunny says: Windows 10? Nein danke!
(")_(")

[Gareth Magennis]

"Mike Tomlinson" wrote in message news:tUb4h9Ee...@jasper.org.uk...

En el artículo <b04ec4db-a196-4f77...@googlegroups.com>,
pf...@aol.com <pf...@aol.com> escribió:

>Well, when I put the diode in correctly (per the schematic), the hum
>reduced
>greatly

Ignore Phil, he's a thick Australian* cunt. You're doing fine, carry on
as you are.

* I do apologise for the Australians, they're the result of criminals,
murderers and rapists transported to van Diemen's land and the product
of enthusiastic interbreeding ever since.

Bottom line, Phil called this correctly IMHO.

No way a reversed bias diode would not have destroyed the output stage in a
few seconds.




Gareth.

[Phil Allison]

Gareth Magennis wrote:

>
>
>
> Bottom line, Phil called this correctly IMHO.
>
> No way a reversed bias diode would not have destroyed the output stage in a
> few seconds.
>

** Of course.

If the original 4004 was leaky, that would explain a lot.


.... Phil

>
>
> Gareth.

[Gareth Magennis]

"Phil Allison" wrote in message
news:e8877c67-598c-4734...@googlegroups.com...

And yes, the schematic doesn't make sense, you don't get an (apparently
expected?) -65v bias from a 100v tap and a (faulty) diode.

You would, however, get rather a large amount of crossover distortion from
such a tap, for as long as the caps held up.



Gareth.