Adding Video Trigger to TDS744A

[rochoz]

Actually Mouser does have the UCD series in stock in both the 33uF/10V and the 10uF/35V values. I personally don't think using a part that is over-rated will help much, but its up to you.

I don't know what format your EPROM programmer saves its file in, but there is an easy way to tell. After you save the file, go look at it in Windows Explorer. A binary file will be exactly 512K bytes in size, a HEX file will be much larger because it is an ASCII format. You can also open a HEX file and view the data using a text editor such as Notepad, whereas a binary file will either not open at all or just display as a bunch of unrecognizable characters.

[nctnico]

On the topic of replacing capacitors:
How many capacitors are required? I have a TDS510A in which the capacitors still look okay. Removing SMD electrolytics has proved to be tricky when using one iron so I choose to leave the TDS510A as it is (if it ain't broken don't wreck it). Because the TDS510A turns out to be a very nice scope to use I also bought a TDS644A (*). I have no idea what to expect since the display doesn't work. Maybe it needs the capacitors replaced in which case I'll probably take care of both scopes. I recently bought a cheap SMD tweezer soldering station from Aoyue (model 950) so removing the capacitors without damaging the PCB should be possible.

(*) It's this junker:
http://d3d71ba2asa5oz.cloudfront.net/23 ... 64p-02.jpg
I wonder what happened to the screen... Not that it matters much to me. The replacement LCD display arrived today

BTW the price of ceramics is roughly half of the price of electrolytics (especially if you use 22uf instead of 33uf) but the extremely low ESR may cause other problems. In theory higher voltage electrolytics should have a longer life. The ECU in my car uses 105degr. 63V capacitors in the 5V circuitry even though it is mounted in a cool place in the passenger compartment.

[hkp7]

I copied the NVRAM over the weekend and installed a socket. I copied the information that was on it as a hex file and tested the new NVRAM for functionality on the scope. It booted up, recognized all of the options ( Opt 05,13,1F,1M,2F) and passed the self test. I have discovered that the flyback has a crack in it and the display is 'breathing'. If I can't find a replacement flyback, I will try encapsulating the cracked area with silicone to see if that might work. I wont be powering it up again until I change out the caps. As far as capacitor count, there are 85 33uF 10V, 25 10uF 35V and if you have a video trigger, it has 15 3.3uF 50V caps on it.

Rochoz, did you use hot air to get the old caps off or some adaptation of a soldering iron? I have a hot air/soldering station and I'm resourceful enough to make an end for a soldering iron that could be used to heat both terminals of the cap simultaneously. Just wondering which you used. It would seem hot air might be easier but might have drawbacks as well. I haven't ordered the caps yet because I was still undecided which to use; I should have that finalized and the order placed this week.

[tech8]

The number one reason for intermittent SPC failures on TDS744A's and similar models are the relays on the attenuator hybrids. This is especially likely on the seldom used channels. Also, the thermal grease between the hybrid and the metal casting has probably dried out. Remove the acquisition board from the unit, remove the casting from the acquisition board, remove the 2 cinch connectors over the attenuator hybrids, unsolder the hybrids from the BNC's, remove the attenuator hybrids from the casting and re-grease the spots on the casting (wipe off old grease first). Be careful not to drip solder down the pin and BNC when soldering the connections back on (it's easy to create a short) during re-assembly. Also be careful to properly align the hybrid and the cinch connectors on top of them when putting them back in the casting. If you don't, you'll crack the hybrid(s) when you screw the whole thing back into the acquisiton board. After everything is back together and re-installed, exersice the relays a lot and hope that fixes things. Good luck!

[rochoz]

My method of removing the electrolytic caps is a bit unconventional, as I don't desolder them at all. What I do is take a pair of wire cutters and cut the cap off the board right at the small indent right above the PCB. This leaves the cap pins still soldered on the card, with 2 stubs and the small plastic insulator remaining. Then I take my fingernail and snap the plastic insulator off, it usually comes off in two pieces. Now I can desolder the 2 pins one at a time, and this is much easier than doing both pins simultaneously. The other advantage is this: if the electrolyte has leaked onto the card, it tends to corrode the solder such that it takes more heat to melt the solder and remove parts. That corrosion also tends to foul up your desoldering tip if you're using a desoldering tweezer. By snipping the caps off, I have one tip for my iron that I use for all the desoldering and another tip I keep for "good" for doing the soldering work.

Good luck!

[hkp7]

Rochoz, your method sounds like the safest I've heard so far. I will definitely give that method a try. As I have said earlier, I will value your advice since you've had success and your advice is not based on guesswork. The method you describe sounds to be actually quicker and less risky than the other methods I have heard discussed. Even though there is the extra step of snipping, the step really wont consume much time compared to trying some of the other methods. I have used your method on tantalum capacitors in another project with great success. It didn't even cross my mind to try it with SMD capacitors due to the apparent lack of clearance for the snippers.

[tech8]

The TDS684A and TDS744/A came out after the time period of the notoriously leaky capacitors. Tek had discovered the problem and changed vendors by then. That's not to say that you couldn't have some that have dried out or started to leak after all these years. However, unless you actually see leakage or dull discolored looking solder connections on the caps, I wouldn't bother replacing them. I was still working working as a tech on the TDS manufacturing line at the time the cap rework occured and spent many years repairing TDS boards for customer service afterwards. So I speak with some years experience.

[hkp7]

Tech8, thanks for the advice. Mine is a 544A so unfortunately it definitely has the questionalble capacitors in it. I am just trying to get my game plan together to be as safe as possible when the time comes to change them. The only error my scope showed was the temperature compensation error other than the 'breathing' display caused by the flyback, so hopefully the work will not be fruitless. I suppose worst case scenario it might be 'therapeudic' to change the caps

[tech8]

Sorry, I thought we were still talking about a TDS744A. I remember the first time I replaced all the caps on all the boards on a TDS544A. It think it's like 96 of them? It took me 5 hours and I had all the proper tools for replacing and cleaning. In case it hasn't been mentioned already, there are 10uf and 33uf caps. Be sure to mark the board locations of one of the values, so you get the right values in the right places when installing the new ones. I know folks have mentioned the importance of thurough cleaning afterwards. If you have access to a sonic cleaner, don't use it. One of the versions of the 4 DEMUX's found on the board doesn't like it if you do. You'll get acquisition memory failures.

[hkp7]

Thanks for that heads up! My game plan is to remove all of either the 33uF or 10uF, clean the board using simple green, deionized water then alcohol, replace all of the removed caps and then go with the other value and do the same cleaning. I will end up cleaning twice (which might not be a bad thing) but I know I will get the caps correctly in their respective places. I did have access to an ultrasonic but was palnning to wash the boards by hand. Now the use of an ultrasonic cleaner wont even enter the equation for me. Thanks.