tek 7603 repair help!

[AWPJudy]

today i picked up a tektronix 7603 mainframe with a 7d02 logic analyzer module off craigslist for the princely sum of $25 usd. the owner said it hadn't been powered on in at least a decade. after cleaning, i booted it up and she ran just fine - for the most part. a couple 7d02 firmware ROMs were throwing some errors, so i went off to track down ROM images and some pin-compatable chips. when i came back the screen had died. occasionally it flashed a bit of display, a few characters for a second. it hasn't since.

checking all the psu rails, they're all in +1% tolerance except one: the 5v rail. it's showing around 4.2v, and the datasheet minimum for a lot of the 74xx ic's on the display board are 4.75. now, i've only just dabbled in electronics, and mostly digital logic at that, so a power supply of this complexity is way above my head, and after reading the service manual a couple times i'm even more lost. going off the schematics every single other voltage is correct, and i cannot make heads or tails of how the feedback circuit that's supposed to be regulating the +5 output works. Q988 appears to be where the +5 is created, and it's getting the proper unregulated +8v from the power supply. in fact, every single voltage i've checked is correct except for the +5 rail and it's sense circuit. highlighted is the only place on the entire board i've found an actual reading of +5v. all of the other +5v (and +5 sense feedback) read +4.2v.

[capt bullshot]

First you should remove all plugins and measure the voltages again (just in case you didn't do so already). To be sure it's not a faulty plugin.

The next thing that's easy to check is the B-E and C-E voltages of the transistors. The B-E voltage shall be around 0.5...0.7V (or -0.5...-0.7V for a PNP), the collector should be more positive (negative for a PNP) than the Emitter. Often this gives you a first indication of a defective transistor or open resistor feeding the base.
The regulator circuits of theses scopes are rather tricky, I don't understand them at a first glance either. Look for the circuit explanations in the service manual, many concepts are explained there.

Edit: I've just discovered the current limiting circuit around R989, U973A and U973B. Check the voltage across R989 to see if there's excessive ouput current causing the voltage to drop.
The voltage regulating loop is made using the other transistors of U973. The ratio of R970 / R971 determines the output voltage (-1/10 of -50V).

[texaspyro]

Most failed 7603's can be traced to shorted dipped tantalum caps.  The failure appearing after a few minutes of power-on after long storage is a classic sign.

[David Hess]

So *you're* the one who picked up that $25 7603 with 7D02.  We were discussing that listing over on TekScopes@yahoogroups.com.  Had I been close, I would have tried to get it.

The 7603 will not unblank the beam without a plug-in installed into the horizontal slot so unless you have 7000 series amplifier or timebase plug-in handy, the 7D02 will be your only option.

Check that there is not excessive ripple at the collector of Q988.  The regulator may be dropping out due to a worn out input capacitor.

The voltage across R989 will indicate the output current.

As I recall, the 7D02 plug-in draws a lot of power so if the 7603 power supply was weak, there could be problems.

[AWPJudy]

after poking about at it all day the 5 volt rail sometimes drifts back up to around +4.5v and a little garbage appears on the screen, so at least i think i'm heading the right direction. it also probably would have saved a LOT of trouble if i had started with the correct schematics for this exact board revision 

i just noticed something extremely odd: reading across r995 (8.06kohms) and r994 (145kohms) the multimeter goes totally batty and eventually settles around 407 ohms and 9.4 kohms respectively. since it's some sort of divider and the ratio is roughly the same, i thought they might have changed the values, but sure enough the codes ON the resistors show the 8k/145k values specified on the schematics. why would two resistors both read massively less resistance, in the same direction, roughly in proportion? they're both reading the proper -50v/0v otherwise.

i've tested every other input voltage, every other resistor, all the B-E voltages were around .5-.7v range capt bullshot mentioned specified.

So *you're* the one who picked up that $25 7603 with 7D02

yep! don't worry, it has found a loving home, even if i have no idea what i'm doing. it just looked far too cool to pass up. the unregulated 8v on the collector is showing a steady 12v p-p for an rms of around 8.3vdc.

[capt bullshot]

the unregulated 8v on the collector is showing a steady 12v p-p for an rms of around 8.3vdc.

Did you measure this with a VM or a scope? 12Vpp superposed on 8.3Vdc smells like a dried out bulk capacitor. Look for the raw 8V bulk cap and check this one for capacity, ESR, or just replace it for a try.
If you have a scope, check the 5V output for ripple, that would indicate the same.

[David Hess]

So *you're* the one who picked up that $25 7603 with 7D02

yep! don't worry, it has found a loving home, even if i have no idea what i'm doing. it just looked far too cool to pass up. the unregulated 8v on the collector is showing a steady 12v p-p for an rms of around 8.3vdc.

I am dubious that you will get a lot of use out of the 7D02 but a set of standard plug-ins will turn the 7603 into a fantastic oscilloscope.  The 7603 is known for its unusually large CRT.

Those voltage measurements are consistent with an open or otherwise bad input capacitor.  The voltage at the collector should be DC with a ripple of a couple volts at most.  The peaks should never be lower than about 8 volts.

[AWPJudy]

I am dubious that you will get a lot of use out of the 7D02

...

Those voltage measurements are consistent with an open or otherwise bad input capacitor.  The voltage at the collector should be DC with a ripple of a couple volts at most.  The peaks should never be lower than about 8 volts.

i'd love to find the z80 or 68k personality modules. my dad and i love tinkering with old computers and even the general purpose probes might just help with debugging his old PERQ io board.

included are some shots of the unregulated 8v and the 5v output. looks like c821, a giant 18000uf 15v electrolytic can, is our chief suspect? it's the bulk cap for the 8v rail as far as i can tell.

[capt bullshot]

included are some shots of the unregulated 8v and the 5v output. looks like c821, a giant 18000uf 15v electrolytic can, is our chief suspect? it's the bulk cap for the 8v rail as far as i can tell.

Yes, it is. You can replace it with a modern type, you may want to read this thread https://www.eevblog.com/forum/testgear/tektronix-76037613-repairs/ for a very professional repair, or bogde a new one in, or even just put a new one of similar rating (e.g .15000uF / 16V) in parallel with the old one.
Thinking long term, it could be wise to replace all bulk capacitors on this particular board.

[AWPJudy]

thanks for the help! i'll post pictures once i've tested the rest of the bulk caps, sourced some replacements, and gotten this beauty up and running. it's early 80's from the dates on the logic. the case is dinged up, but the internals are in absolutely stunning condition. best $25 i ever spent!

[David Hess]

Yep, C821 is open.  This happens when the rubber seal leaks allowing oxygen and water to get into the capacitor and corrode the aluminum terminal connections.