Canon T90: Non-functional and revived through systematic troubleshooting, DC/DC converter replaced, one LED digit missing, sticky shutter

[Andreas Thaler]

Checking a Canon T90 according to the troubleshooting instructions by Larry Lyells in the SPT Journal.

Canon T90 Repair Guide

From the Society of Photo-Technologists Journal

learncamerarepair.com

No sign of life.

Battery compartment removed, measuring the input voltage on the DC/DC converter.

Front and top cover removed for further measurements.

A double diode needs to be tested. To do this, I'm desoldering it with hot air.

The diodes are ok.

Soldered in and back in place.

The DC/DC converter, which generates different voltages from a 6-volt input (4 x AA batteries), has incorrect voltage readings.

There are obviously cold solder joints on the circuit board that I can't repair.

The converter is defective; I have a replacement in my stash.

 

[koraks]

here are obviously cold solder joints

It's hard to make out on the photo, but it looks like that part of the PCB overheated. Something downstream might have shorted out, causing overload on the DC/DC converter. That's just a guess though.
Either way, I'd be cautious replacing it since you might just end up frying the next one, too.

 

[Andreas Thaler]

DC/DC converter. Front and back

After the replacement, the T90 is back to life. All functions are working.

One of the seven-segment displays in the finder is missing a digit. I would have a replacement LED.

As with all T90s, the sticky shutter is also present here.

There is still some work to be done

 

[Andreas Thaler]

It's hard to make out on the photo, but it looks like that part of the PCB overheated. Something downstream might have shorted out, causing overload on the DC/DC converter. That's just a guess though.
Either way, I'd be cautious replacing it since you might just end up frying the next one, too.

Yes, a transistor may have been overloaded; it doesn't look good. However, I desoldered it at high power. It's safest to replace the entire assembly.

 

[koraks]

However, I desoldered it at high power.

Not sure what that means.

It's safest to replace the entire assembly.

What I meant is that the DC converter supplies power to other parts of the camera. The failure of the DC converter may be due to failure elsewhere. Monitor power draw on the converter closely.

 

[Andreas Thaler]

I'll monitor this and measure the current flowing through the converter. Voltages too, which are measured via the back of the circuit board.

 

[Andreas Thaler]

Not sure what that means.

My specialist for desoldering with desoldering wick.

High power

I held onto it for some time, which must have contributed to the damage.

 

[Andreas Thaler]

The three motors draw the most power. Let's see what the measurements show.

 

[koraks]

I now see you added a photo to #4 that I didn't see before. That switching transistor is fried, obviously, and it's not due to desoldering. The question is why it broke.

 

[Andreas Thaler]

Current and temperature measurements

I measured the following currents in the T90 (without film, connected to the laboratory power supply; in brackets the reference values according to the SPT Journal):

• Main switch OFF: 7,8 uA (7,0 uA)
• Main switch (LCD) ON: 44,5 uA (430 uA)
• Trigger switch (lightmeter, LEDs and LCD illumination) ON: 109 mA (190 mA)
• Serial shooting in motor mode HIGH: 600 mA (650 - 700 mA during film advance after exposure)

Additionally, I measured the surface temperature of the connectors on the DC/DC converter board with an IR thermometer.

The temperature remained constant even during serial operation.

I conclude that the problem was with the DC/DC converter and that replacing the board resolved it.


Possibilities to replace the viewfinder LEDs

A digit in the viewfinder display (a bar on one of the seven-segment LEDs) has failed (see the thread). This means that the displayed values can still be read.

I've looked into possible replacement options.

Position of the board with the LEDs.

Unit diassembled (on a spare parts mirror box).

The LED unit is attached to the FD board, which is soldered to the mainboard (blue arrow). It can be removed.

Alternatively, only the LED array could be separated from its circuit board. This can be done using hot air.

However, soldering it back on is a delicate process; the connections would have to be tinned and then re-soldered with hot air, out of sight.


Conclusion

• Both options are risky.
• Replacing the FD board with the LED array also involves replacing an IC that presumably controls the LEDs. I don't know the board versions or whether they're compatible, so it would be an experiment.
• Soldering and re-soldering the LED array is tricky, and if it's not done correctly, it could render the connection unusable.
• I'll therefore leave everything as it is. The failure of one digit doesn't justify the risk.
• It remains to be seen whether the failure is related to the previously defective DC/DC converter (possible overload). I'll continue to monitor this.

 

[Andreas Thaler]

Contacts on command dial

(Picture from another T90 repair project.)


The adjustment wheel (command dial) on the grip near the shutter release does not change the values on the LCD in click increments.

This means that the corresponding gold-plated contacts on a disc in the grip must be cleaned.

This is not a big effort; see

Post in thread 'Canon T90: Address unresolved problems; broken plastic parts in Copal S shutter, epoxy, EEE/HELP/bc errors, troubleshooting partially successful'
https://www.photrio.com/forum/threa...ting-partially-successful.213404/post-2892278

 

[Andreas Thaler]

Unfortunately, this project was also unsuccessful

I was able to disassemble the shutter, clean it, and adjust it on the camera tester.

After assembling the T90, the charge motor started briefly as soon as I applied voltage, followed by an EEE on the display.

By lifting the mirror box and operating the mirror switch (SW13) I was able to start the camera for a moment.

The SPT Journal has a dedicated troubleshooting instructions for this.

And so I was able to locate the error. It appears to be a contact problem in the front flex connector under a metal pressure plate.

Here, three circuit boards are connected one above the other with gold-plated contacts.

Elastic connectors with gold-colored inclusions are located between them.

By tightening the two pressure plate screws differently, I was able to get the T90 working again.

Only during continuous operation did the mirror occasionally stop working. Inserting two elastic connectors one above the other improved performance.

I also cleaned and polished the externally accessible contact surfaces of the release magnet. But as I touched the release magnet with the screwdriver, the camera jammed and EEE was permanently on the display.

This problem now occurred with two T90s after assembly. The cameras worked before.

I suspect that after parting the boards, the connection via the elastic connectors no longer works reliably.

I couldn't figure out how these work. They connect all rows of contacts without short-circuiting them, but must still be conductive because they connect two boards each.

---

I've now ruled out any mistakes during disassembly and reassembly as it already worked and I strictly followed the instructions by Larry Lyells in the SPT Journal.

It's a shame, but I'll keep at it.

A few defective T90s are still waiting in the box