Topics

Help Troubleshooting 466, No +15VDC


Sparky
 

This 466 worked great for years, then suddenly I smelled something burning, and the fan stopped. All low voltages good on interface board, *except* for +15VDC. Reads about 4VDC, (and about 4 ohms to ground with power off....). If I leave power on past about 15 sec, I can smell something burning again, so I only power up long enough the check the voltages.
I'm thinking tantalum cap, but wondering a good way to isolate the offender? Assume the the +15VDC goes all over the scope, to other boards, so would be a bear to lift all tant caps one at a time. Thinking of breaking the trace somewhere near the PS source. Any other tips or likely suspects?

Thanks in advance!


Harvey White
 

There are a number of tricks.  One is to put a current limited supply on the +15.  Adjust the supply to +15 volts and the current limit to roughly between 50 ma to 100 ma.  Look for the lowest voltage across each capacitor.  Suspect the ones with the lowest voltage drop.  Bear in mind that in a parallel string fed from one end, the bad one and all the ones in the string after that would have the same voltage.

Another would be to take a current trace probe (HP made one), feed the line with a pulse train, and then see where all the current is going.

The third method is the LED test.  In a dark room, look to see what capacitor lights up like a dim LED, the dull red glow may be something you could observe.  Ditto if you had an infrared camera.  If there are any filter resistors (Tektronix frequently used a 10 ohm series resistor for isolation), check for damage.

Oh, and disconnecting boards to try to isolate to a particular board isn't a bad idea.

Harvey

On 9/7/2020 3:18 PM, Sparky wrote:
This 466 worked great for years, then suddenly I smelled something burning, and the fan stopped. All low voltages good on interface board, *except* for +15VDC. Reads about 4VDC, (and about 4 ohms to ground with power off....). If I leave power on past about 15 sec, I can smell something burning again, so I only power up long enough the check the voltages.
I'm thinking tantalum cap, but wondering a good way to isolate the offender? Assume the the +15VDC goes all over the scope, to other boards, so would be a bear to lift all tant caps one at a time. Thinking of breaking the trace somewhere near the PS source. Any other tips or likely suspects?

Thanks in advance!



 

Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin


Stephen Hanselman
 

Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,

Stephen Hanselman
Datagate Systems, LLC

On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:

Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin



Harvey White
 

So what do you do to repair craters in the PC board?

Harvey

On 9/9/2020 6:18 PM, Stephen Hanselman wrote:
Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,
Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:

Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin



Jim Ford
 

Maybe it does for shorted tantalum capacitors. Not guaranteed to work on shorts inside the PCB. Oh, it will show you where the short is alright, but that doesn't mean the short will clear or the PCB will be usable afterward. Don't ask....

Jim Ford

------ Original Message ------
From: "Stephen Hanselman" <kc4sw.io@...>
To: TekScopes@groups.io
Sent: 9/9/2020 3:18:34 PM
Subject: Re: [TekScopes] Help Troubleshooting 466, No +15VDC

Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,

Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:

Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin




Harvey White
 

I'd suspect serious consequences on anything other than 2 layer PC boards, and even then....

I think I'd prefer a, perhaps slower, method.

Harvey

On 9/9/2020 11:50 PM, Jim Ford wrote:
Maybe it does for shorted tantalum capacitors.  Not guaranteed to work on shorts inside the PCB.  Oh, it will show you where the short is alright, but that doesn't mean the short will clear or the PCB will be usable afterward.  Don't ask....

Jim Ford

------ Original Message ------
From: "Stephen Hanselman" <kc4sw.io@...>
To: TekScopes@groups.io
Sent: 9/9/2020 3:18:34 PM
Subject: Re: [TekScopes] Help Troubleshooting 466, No +15VDC

Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,

Stephen Hanselman
Datagate Systems, LLC
 On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:

 Hello Sparky,

 my preferred method for instant finding of shorts in these caps is with my IR-Cam.
 You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

 IR-cams are getting more and more accessible, perhaps you can borrow one?

 cheers
 Martin






Stephen Hanselman
 

Actually my partner used that procedure regularly, I think it’s a bit drastic but he swears by it.

Regards,

Stephen Hanselman
Datagate Systems, LLC

On Sep 9, 2020, at 20:50, Jim Ford <james.ford@...> wrote:

Maybe it does for shorted tantalum capacitors. Not guaranteed to work on shorts inside the PCB. Oh, it will show you where the short is alright, but that doesn't mean the short will clear or the PCB will be usable afterward. Don't ask....

Jim Ford

------ Original Message ------
From: "Stephen Hanselman" <kc4sw.io@...>
To: TekScopes@groups.io
Sent: 9/9/2020 3:18:34 PM
Subject: Re: [TekScopes] Help Troubleshooting 466, No +15VDC

Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,

Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:
Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin






Chuck Harris
 

Physically, tantalum drop capacitors have a lot of
the characteristics of a fuse. The tantalum metal
melts at a reasonably low temperature, and pops the
capacitor's body, leaving a couple of bare wires
sticking up in the air.

IF, the circuit board traces can handle the current,
the old capacitor will have revealed itself, and be
ready for an easy top-side repair.

Capacitors on the logic supply planes of a multilayer
board clear most dramatically with this trick, but
when it goes wrong, the thermal breaks blow too, and
you may have to drill the short out of the board and
put the capacitor elsewhere... or do without.

I prefer a slower method, using an IR camera, and a
middling amount of current to warm the offending
traces, and the offending capacitor. They show up
like little streetlamps on the IR camera's screen.

-Chuck Harris

Stephen Hanselman wrote:

Actually my partner used that procedure regularly, I think it’s a bit drastic but he swears by it.

Regards,

Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 20:50, Jim Ford <james.ford@...> wrote:

Maybe it does for shorted tantalum capacitors. Not guaranteed to work on shorts inside the PCB. Oh, it will show you where the short is alright, but that doesn't mean the short will clear or the PCB will be usable afterward. Don't ask....

Jim Ford

------ Original Message ------
From: "Stephen Hanselman" <kc4sw.io@...>
To: TekScopes@groups.io
Sent: 9/9/2020 3:18:34 PM
Subject: Re: [TekScopes] Help Troubleshooting 466, No +15VDC

Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,

Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:
Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin








 

On 10. Sep 2020, at 15:14, Chuck Harris <cfharris@...> wrote:

...They show up like little streetlamps on the IR camera's screen.
Exactly!

I have such a camera for some years, it became my most efficient troubleshooting tool as soon as my nose rings the bell... and even beyond that.

Recently I was checking an HP function generator that blew the fuse.
Fuse replaced, cautious ramp-up on the variac with the IR-cam in the other hand, one of the rectifiers quickly made its "streetlamp"...

cheers
Martin


Vince Vielhaber
 

I've done that ... once. I had a main board for a Radio Shack Model III computer with ONE shorted 0.1uf bypass cap out of about 50 or so. 5V at 10 amps cleared it in less than a second.

Vince.

On 09/10/2020 02:16 AM, Stephen Hanselman wrote:
Actually my partner used that procedure regularly, I think it’s a bit drastic but he swears by it.

Regards,

Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 20:50, Jim Ford <james.ford@...> wrote:

Maybe it does for shorted tantalum capacitors. Not guaranteed to work on shorts inside the PCB. Oh, it will show you where the short is alright, but that doesn't mean the short will clear or the PCB will be usable afterward. Don't ask....

Jim Ford

------ Original Message ------
From: "Stephen Hanselman" <kc4sw.io@...>
To: TekScopes@groups.io
Sent: 9/9/2020 3:18:34 PM
Subject: Re: [TekScopes] Help Troubleshooting 466, No +15VDC

Our favorite way was to use a HP 2100 power supply 5VDC at 50Amps. Finds’um every time

Regards,

Stephen Hanselman
Datagate Systems, LLC
On Sep 9, 2020, at 07:00, Martin <@musaeum> wrote:
Hello Sparky,

my preferred method for instant finding of shorts in these caps is with my IR-Cam.
You may supply the +15V rail with a current limited supply, or shortly apply mains power - the component will instantly reveal itself.

IR-cams are getting more and more accessible, perhaps you can borrow one?

cheers
Martin






Sparky
 

Thanks, lots of good suggestions. The IR imaging camera sounds really helpful, have to look into getting one.
What I ended up doing was to separate the 15VPS from all the loads, and coincidently, or maybe not, there was one of those white-insulated jumpers proximate to the last PS component, C173x?, 15uF?EC. Pulled the jumper, and, voila, got 15.0 volts. Expected the other side to be still near shorted, but oddly it was now reading about 50 ohms. Huh? Reinstalled the jumper, with series ammeter, and voltage holding steady at 15.0 with 0.75A current. Buttoned everything back up.
How could pulling the jumper have "fixed" the problem? I'm stumped.
Note: my other 466 had a bridge of white powdery material, like the stuff on battery posts, between this same cap (C173x?) at the output of the 15VPS and the adjacent 65V test point. Thinking it likely came from the EC leaking over the years... Cleaned it up, and working fine now, too (both CH traces would get unfocused intermittently...).
Alls well that ends...so far. Not overly confident, as nothing was replaced.

Question: there was mention of multi-layer pc boards. Did TEK use more than a 2-layer board (top and bottom traces) in this generation of scopes?