[FIXED] Tektronix 1502 TDR repair

[sorenkir]

Hi,

I have acquired a quite cheap Tektronix 1502 TDR, 1986 vintage, in very good condition, but sold as not working (not powering up).

A common fault is that the unit does not power-up when there is no battery, or battery completly dead.

Indeed the battery was there but it's small 2A fuse was open.
So I tried to power the instrument from a bench supply (set to 12V 0.8A) with banana plugs inside the battery compartment: no sign of life, 20mA drawn.

Does not turn on? Take it apart!

On visual examination of the power supply board, I immediately found the suspect, there is even a red arrow on the side pointing to it  :



The missing piece of the blown transistor was not in the unit, so somebody already went there...
Did he/she tried to fix with no success? Hum hum...

This power transistor Q6435, part number 151-0352-00 is a GE X44C282.
According to the "Tektronix Semiconductor Cross Reference" (http://w140.com/tekwiki/wiki/Reference_material), it is from D44C6 family: 45V 4A 30W NPN transistor.

I found a D44H8 (60V 8A 60W NPN, same pinout) in my "stock" and managed to solder it from the component side, without removing the board!
I also reinstalled one of the spare fuses in place, instead of the small piece of wire somebody had soldered before as a "replacement".



I checked the power rails for shorts as there are some tantalum capacitors on the boards: no short.
First try, powered from the bench supply: the unit works!
Here with a 3.70m piece of coaxial cable:



Powered from mains, with a new battery fuse, the battery even takes some charge!

On first visual inspection, I was a bit worried by the small piece of unattached cable at the rear of the strip line:



I removed the shield to have a look:



According to figure 4-26 in the service manual, it seems normal, although I have not found any indications about its purpose.

Indeed this unit has been tested by real people:



Michel.

[texaspyro]

Another very common faults in these units is bad high voltage caps on the CRT power supply board...  these are the large-ish brown (usually) rectangular caps (with dipped plastic coating).  I think that they are 0.33 (or 0.033) uF 1600V... it's been a while since I worked on one.

[tggzzz]

Congratulations, and I hope you have fun playing with it.

The pill fuses are available, but cost ~£8 each, IIRC.

I've had Q6246/Q6341 spew a corrosive material across the PCB, destroying tracks. Not a major problem since it only a 2-layer board. I suspect it is worth leaving them unless you see any warning signs.

You might like to try and work out why that transistor failed so spectacularly.

If you do replace the batteries, be aware that even with the 1502 turned off, there is a 1-2mA current drain. That can reverse-charge weaker battery cells. Hence the manual stating that they must be recharged every month. I prefer to leave the battery pack out, or remove the fuse on the battery pack.

I've found battery fakers to be unreliable, NiMH to have questionable capacity and need a low charge current (change R6131); hence I prefer sub-C NiCd cells (and have a small stock of them ).

Don't forget to check that the BNC shorting bar operates; it helps prevent damage to the unobtanium tunnel diode.

[texaspyro]

Instead of pill fuses, just use regular radial lead picofuses (50 cents or so), fold one lead down the fuse body,  trim the lead to match the pill fuse,  add some heat shrink around the fuse body and lead,  profit!