Tektronix TDS420a No Display on CRT, Fan & 10 LEDs light (hangs).

[slbender]

Hi Techs and Tek lovers,

This set just arrived yesterday. Nothing looks burned, but upon close inspection, the Centronics and RS232 board was inserted in such a way that its back was shorting against a metal chassis brace that runs from front to back in the chassis. The CRT filament does not appear to be glowing, I suspect the -6.3 volt supply may be blown, causing the turn on diagnostic to hang with the front panel ten LEDs on.  It also smells of "bad caps" but nothing can be seen, no corrosion, or damage.  First, I will pull that board and see if the turn-on diagnostics completes. And if the CRT comes on... But I won't hold my breath.

Suggestions appreciated.

-Steven

[borghese]

The first step is to check the power supply both for voltage and ripple and replace the electrolytic capacitors

[slbender]

Thanks Borghese,

Update - without the centronics / serial board, it acts the same, but the CPU board alternates between "3" and ".8". Putting the board back, still the same, I can now see the CRT's filament glow, I didn't see it before, but maybe I was looking in the wrong areas of the neck of the CRT.  The CPU board stops on the ". 8" again.  Still nothing shows up on the CRT.  This unit is a later set, with lots of 1995 date codes visible, next I will try to pull the PS and check those output caps, ESR, the voltages, and ripple.

Steven

[borghese]

You can also say thanks with the "Say Thanks" button in the upper right corner
If I remember correctly the scope has a VGA output, you could connect a monitor and observe the boot phase

[slbender]

Trying the VGA is a good idea, too bad I just removed the 420a set's PS.  That hidden behind and sort of under the transformer Torx #10 screw drove me crazy for over an hour, as I could not see it, except at a certain angle. Another case of being up all night on projects. So the VGA port will have to wait.

Thanks,

Steven

[rf+tech]

Hi Steven,

The TDS460 and TDS420 both have a known no-boot failure mode, easily identifiable by the MC68020 µP /RESET pin held constantly low.

Culprit is an open-circuit 27.4 kOhm SMT resistor, at the top of a voltage divider connected to the +5 Volt bus, on the PFI pin (9) of MAX695 µP supervisor.

PFI remains at essentially zero Volts, /RESET never goes high.

The repeat offender is visible next to MAX695, under clear potting. I *think* it's on the CPU card, or adjacent card.

RF+ Tech

[andy2000]

If it smells of bad caps, I wouldn't do anything else until you find out where that smell is coming from.  Check under the caps in the power supply for signs of leakage.  Also, check around any SMD electrolytics.  Try probing the boards near caps with an ohm meter looking for leakage.  Often there will be an almost invisible film of electrolyte near the leaking caps.

[slbender]

Hi Steven,

The TDS460 and TDS420 both have a known no-boot failure mode, easily identifiable by the MC68020 µP /RESET pin held constantly low.

Culprit is an open-circuit 27.4 kOhm SMT resistor, at the top of a voltage divider connected to the +5 Volt bus, on the PFI pin (9) of MAX695 µP supervisor.

PFI remains at essentially zero Volts, /RESET never goes high.

The repeat offender is visible next to MAX695, under clear potting. I *think* it's on the CPU card, or adjacent card.

RF+ Tech

Thanks, RF+ Tech, when I replace the PS I will look for that 27.4k smt resistor.

Andy2000 the cap smell seemed faint, not sure where it was coming from.  I ordered over $20 of various smt caps a few days back, the 33uF 10Volt ones just came in.  I have to sày the boards look clean, no corrosion.  Since this is the later "A" version, lots of 1994/1995 date codes, maybe the 'lytic caps are OK, but I will replace them anyway.

-Steven

[slbender]

Hi Steven,

The TDS460 and TDS420 both have a known no-boot failure mode, easily identifiable by the MC68020 µP /RESET pin held constantly low.

Culprit is an open-circuit 27.4 kOhm SMT resistor, at the top of a voltage divider connected to the +5 Volt bus, on the PFI pin (9) of MAX695 µP supervisor.

PFI remains at essentially zero Volts, /RESET never goes high.

The repeat offender is visible next to MAX695, under clear potting. I *think* it's on the CPU card, or adjacent card.

RF+ Tech

Hi, RF+,

The PS is still removed, but I took out the MC68020 CPU Board and checked that R marked 2742, after scraping away much of that clear goo.  Then measuring the R while it is still in the circuit, it measured 10.33 K ohms and the closer edge is connected to the Maxim MAX695 chip pin #9.  Hmm. I don't think that is the problem, unless that is what the circuit itself measures when that R is open. You did not state whether I should be able to measure "open" across that R while it is in the circuit, but that was implied.  So now it is still just not resolved if that R is open or just too low in value... or way too high in value.  So can you provide any additional information, or any ideas?

Steven

[rf+tech]

Hi Steven,

As I recall, in-circuit measurement of the resistance was an unreliable indicator. You mention 10.33 k, that seems to be about what I measured, something around 10 kOhms in-circuit.

The only reliable diagnostic is to measure the actual voltage present on MAX695 pin 9. As the datasheet explains, the supervisor holds /RESET low for nominally 200 msec, *after* PFI input pin 9 exceeds 4.65 Vdc nominal. If pin 9 never exceeds 4.65 Volts, /RESET remains low indefinitely resulting in a no-boot condition.

The /RESET output does not go directly to the MC68020 µP, rather through a couple of gates. Regardless, any non-booting test gear with a µP should have its /RESET pin checked before proceeding with disassembly. This is one of the basics, just like checking fuses and PS voltages.

RF+ Tech

[slbender]

Hi Steven,

The TDS460 and TDS420 both have a known no-boot failure mode, easily identifiable by the MC68020 µP /RESET pin held constantly low.

Culprit is an open-circuit 27.4 kOhm SMT resistor, at the top of a voltage divider connected to the +5 Volt bus, on the PFI pin (9) of MAX695 µP supervisor.

PFI remains at essentially zero Volts, /RESET never goes high.

The repeat offender is visible next to MAX695, under clear potting. I *think* it's on the CPU card, or adjacent card.

RF+ Tech


Thank you, I checked over the PS, it looked OK, I did not change any caps in the PS, nor have all the set caps come in yet, so I pulled off the one from the CPU Board, since this set has 1995 date codes, I think it is likely beyond most of the "cap" problems.  So I checked the 10uF set cap, it measured 9.9uF. basically on spec for 10uF., no leakage, so I soldered it back in.

I then soldered a new 1/4 Watt 25.4K across the suspect 27.4k set part near the Maxim chip, replaced the PS, and started it up.

It blinked and blinked for a while, the CRT came to life, it displayed a few more leds on, leds off, and was taking a bit longer than expected.  I turned it off, then tried it a second time.  Again the CRT came to life, it seemed to be going faster now, with what are usually for the starting sequence and then indicated "test passed". and went to the graticule and all four channels showed background noise when pressed too become active.  However, only channels 2,3,4 showed a random noise when I put my pinky in close contact with the center of the BNC.  So I think the Attenuator for channel 1 may have a problem...

Others have found blown parts on the ceramic substrate of channel 1 and repaired that, now I just have to locate that info again.

Next time I fire it up I'll do more extensive testing with a probe and the CAL signal.

Looking good at this point, at least 3 out of four working.

-Steven