Test Equipment Anonymous (TEA) group therapy thread

[tverbeure]

3 weeks ago, I had one TDS xxx scope, a 684B. Since then, I've added a TDS 684C (free!), last weekend a TDS 754D (not so free), and today this TDS 794D fell into my lap after unexpectedly 'winning' one at a massive online test equipment auction. I don't quite get how it happened: a broken TDS784D sold for $450. I got the TDS794D, also broken, for $290.



A TDS 7404 4GHz/16Gsps sold for $1800 and a DPO71604B 16GHz/50Gsps sold for $6200. Based on the photos on the auction, the 7404 didn't boot at all. The display of the 71604 has pics of booting up, but no waveform display. It probably has issues as well. But I disgress.

So what was broken about the TDS 794D:

• the pictures showed that all green LEDs were on. That usually means the DS1486 RTC is broken which is easy to fix. But that was not the case here: the PCB between the CPU and ACQ was turned upside down. I found out after unplugging it and accidentally not plugging it back in before powering up.
• the CRT doesn't work at all. Not a case of being dim, no image at all. But VGA is working fine.
• a sine wave showed as 1/4th sine wave, 1/4th ground, 1/4th sine wave, 1/4th ground. A simple SPC run fixed that.
• a push on the run/stop button doesn't always work. Should be a matter of cleaning.

That's it! A 2GHz input signal comes through nicely. I'm super excited about finally having the final 3-digit TDS boss and may even consider selling some of the others. Maybe.

One thing I don't understand: it has the 2M option, but with only 1 channel enabled I can only record 500k points, not the expected 2M. Any idea what's up with that?

I also bid $60 for a monstrous 60lb beast of a HP 8782B vector signal generator and unfortunately won it. Had to smuggle it inside the house through the garage.

[Vince]

• the pictures showed that all green LEDs were on. That usually means the DS1486 RTC is broken which is easy to fix. But that was not the case here: the PCB between the CPU and ACQ was turned upside down. I found out after unplugging it and accidentally not plugging it back in before powering up.

Oh yeah I can relate, happened to me when I was fixing my TDS 544A a few years ago.
I had it working, in pieces on the bench, was so happy, then I put it all back in one piece and it failed to boot, I was disgusted !
... then figured out what you just did.. that bloody interconnect PCB can happily be plugged the wrong way around and keep the scope from booting...

One thing I don't understand: it has the 2M option, but with only 1 channel enabled I can only record 500k points, not the expected 2M. Any idea what's up with that?

That's because 'M' doesn't mean 'MB'. It just stands for 'Memory'. So 1M or 2M simply means "Memory option #1 or #2". So that can mean any size of memory depending on the scope model and marketing decisions at a given time.

[TERRA Operative]

EVERYONE who has worked on one of these scopes has flipped that interconnect upside down and had a mini-heart attack... It's a rite of passage. hahaha.


Looks like you have a good candidate for an LCD conversion in that TDS794D. I might have some spare CRT boards, if you can't find one locally for a decent price....



Also, if you don't mind, can you dump the firmware and NVRAM contents out from all your scopes? I've been collecting firmwares to upload to the Tekwiki, so if you have versions I don't have, it would add to the collection.

Instructions here: https://www.eevblog.com/forum/repair/unified-tektool-released-!-(firmware-flash-tools-for-old-tds-series)/msg3758918/#msg3758918

[Robert763]

I've done it again. Bought ANOTHER AC DC current clamp. Worse it's ANOTHER HEME 1000. That's 5 (plus a couple of spares units.
<SNIP>
Robert.

Well the current clamp turned up and it's totally wrecked. Looks like severe battery leakage corrosion. All visible metal is green, the jaws won't move and the screws are seized 
Waiting for a response from the ebay seller. If I don't hear from them today I will raise a case wih ebay.

[AVGresponding]

I also bid $60 for a monstrous 60lb beast of a HP 8782B vector signal generator and unfortunately won it. Had to smuggle it inside the house through the garage.

That's great value; $1/lb. Quite an impressive beast!

[factory]

I've done it again. Bought ANOTHER AC DC current clamp. Worse it's ANOTHER HEME 1000. That's 5 (plus a couple of spares units.
<SNIP>
Robert.

Well the current clamp turned up and it's totally wrecked. Looks like severe battery leakage corrosion. All visible metal is green, the jaws won't move and the screws are seized 
Waiting for a response from the ebay seller. If I don't hear from them today I will raise a case wih ebay.

The extras no good either? Was it advertised as parts? The low price you paid makes me think it might have been.
The high possibly of battery leakage, from excessively incontinent batteries, really puts me off buying more handheld meters/DMMs.

David

[tverbeure]

That's because 'M' doesn't mean 'MB'. It just stands for 'Memory'. So 1M or 2M simply means "Memory option #1 or #2". So that can mean any size of memory depending on the scope model and marketing decisions at a given time.

I get that. But the tekwiki says that the 2M option on the 794D corresponds to 2M points. 500k points is for the 1M option. Both the startup screen and the sticker in the back say 2M present.

My 754D has the 1M option. I’ll open up the 794D again and check if it has the AS7C256 or AS7C1024 memory chips on the acq board…

[tverbeure]

Also, if you don't mind, can you dump the firmware and NVRAM contents out from all your scopes? I've been collecting firmwares to upload to the Tekwiki, so if you have versions I don't have, it would add to the collection.

Will do for this one. I’ve already sent you those files for the 3 other ones in the past 2 weeks.

The 794D has 6.0e.

Is there a page somewhere that clarifies the following:

- is, say, 6.0e on a 794 the same as 6.0e on a different t model?
- what can I upgrade to? I know that a 6.0e can be upgraded to a higher 6 version, but what about 7 or 8 versions? I assume others have a tried it a d that you can just revert if it doesn’t work?

[tverbeure]

That's great value; $1/lb. Quite an impressive beast!

The unit has some 10811E VCXO variant in it. It could turn a profit if I sell just that on eBay. (But who am I kidding… I’ve already told my daughter that “one day this will all be yours.”)

[TERRA Operative]

Also, if you don't mind, can you dump the firmware and NVRAM contents out from all your scopes? I've been collecting firmwares to upload to the Tekwiki, so if you have versions I don't have, it would add to the collection.

Will do for this one. I’ve already sent you those files for the 3 other ones in the past 2 weeks.

The 794D has 6.0e.

Is there a page somewhere that clarifies the following:

- is, say, 6.0e on a 794 the same as 6.0e on a different t model?
- what can I upgrade to? I know that a 6.0e can be upgraded to a higher 6 version, but what about 7 or 8 versions? I assume others have a tried it a d that you can just revert if it doesn’t work?

I'm not sure about swapping the TDS794D firmware for other models, as it has a completely different acquisition board that is pretty specific to this model (no input hybrids for example) but maybe it could work, maybe? I never tried as mine is on v8.0 which is where the firmware on Tekwiki came from, and I haven't seen another scope with such a late revision yet.
But for scopes that are equivalent like the TDS754D and TDS784D etc, those firmwares are interchangeable, something that I have tested myself with those specific models.

The main revision number is tied to the revision of the hardware itself, so you can't upgrade from say v6.x to v8.x AFAIK (well, not without hacks, but you're on your own there...)
If you wanted to try it to see, then as long as you have a copy of your original working firmware and NVRAM for your instrument, you can always flash it back and there will be no issues. I tried this myself just for fun, I flashed a completely different firmware into a scope (something like a TDS640 firmware into a TDS754 or something) and it freaked out on boot, but once I reverted it, the scope was all happy again with no problems at all.

[Vince]

That's because 'M' doesn't mean 'MB'. It just stands for 'Memory'. So 1M or 2M simply means "Memory option #1 or #2". So that can mean any size of memory depending on the scope model and marketing decisions at a given time.

I get that. But the tekwiki says that the 2M option on the 794D corresponds to 2M points. 500k points is for the 1M option. Both the startup screen and the sticker in the back say 2M present.

My 754D has the 1M option. I’ll open up the 794D again and check if it has the AS7C256 or AS7C1024 memory chips on the acq board…

The articles in Tekwiki aren't always 100% truthful in my experience. 99,9% of the time they are, but sometimes not.
Better to take info from official Tek literature of the day, like service manuals or catalogues, or datasheet.

See attached the datasheet of back in the day, for the TDS 700D series.
As you can see, the option "2M" for a 794D does not mean 2Mpoints, but 8Mpoints, so Tekwiki was misleading you there.
500K for one channel however, does exist and corresponds to the "1M" option.

That still means there is something inconsistent in your scope. Maybe the F/W was messed with in the past and is displaying "2M" regardless of the actual H/W present in the scope.

Next thing to do is just open up the scope and see for yourself how many memory chips are actually soldered to the board, this can't lie 

[wkb]

I could not help dragging this Advantest SA one home...

It works, but smelled suspicious. And lo and behold: literally all but the 2 bulk mains input electrolytics in the SMPSU were spilling their guts onto the PCB. Not what I expected from Nichicon, but still they did.

In the meantime busy with searching in my stock for replacement caps.

[TERRA Operative]

Also, if you don't mind, can you dump the firmware and NVRAM contents out from all your scopes? I've been collecting firmwares to upload to the Tekwiki, so if you have versions I don't have, it would add to the collection.

Will do for this one. I’ve already sent you those files for the 3 other ones in the past 2 weeks.

The 794D has 6.0e.

Thank you!

[Runco990]

I could not help dragging this Advantest SA one home...

It works, but smelled suspicious. And lo and behold: literally all but the 2 bulk mains input electrolytics in the SMPSU were spilling their guts onto the PCB. Not what I expected from Nichicon, but still they did.

In the meantime busy with searching in my stock for replacement caps.

Were those by chance the dark brown Nichicons?  If so, I have had a batch that began leaking still in the parts bin!  I'm very suspicious of them now anywhere I see them, as they like to leak quietly underneath long before you SEE the damage.

[squadchannel]

Nichicon's low ESR capacitor from the later 80's to early 90's have a problem with leaking the electrolyte.
many computers(NEC PC-98 etc....) have been affected.

[factory]

The peeing capacitors from the 1980s & 90s, this happens with Panasonic, Rubycon and other brands too, consider yourself lucky if the board traces/vias didn't get eaten and/or parts damaged elsewhere due to the electrolyte covering the PCB being conductive.
It doesn't happen with every capacitor, but usually all of the same value will be leaking and yes I've also found new old stock capacitors peeing.

David

[wkb]

Yes brown Nichicons. Nasty gunk they produce.

[wkb]

Interestingly enough only the electrolytics in the SMPSU were leaking. None of the caps on the various other PCBs are affected, thank goodness.
The unit is ~ 1989 vintage.

[tverbeure]

As you can see, the option "2M" for a 794D does not mean 2Mpoints, but 8Mpoints, so Tekwiki was misleading you there.
500K for one channel however, does exist and corresponds to the "1M" option.

That still means there is something inconsistent in your scope. Maybe the F/W was messed with in the past and is displaying "2M" regardless of the actual H/W present in the scope.

Next thing to do is just open up the scope and see for yourself how many memory chips are actually soldered to the board, this can't lie 

Updated the FW from v6.0e to v6.6e. Re-enabled option 2M. Still only 500k points per channel when only 1 channel is enabled.



Acq PCB has the IDT71024 SRAMs installed, which are the RAMs that are installed for 2M points. Mystery.

[TERRA Operative]

Can you post a photo of the boot screen banner, showing options and firmware revision?


Also, can you post a fresh NVRAM dump of v6.6e for the archives?

[Vince]

Updated the FW from v6.0e to v6.6e. Re-enabled option 2M. Still only 500k points per channel when only 1 channel is enabled.
Acq PCB has the IDT71024 SRAMs installed, which are the RAMs that are installed for 2M points. Mystery.

According to the datasheet, these chips hold 128KB. So how many of those chips does your board carry, what's the actual total amount of memory in that scope then ?

[tverbeure]

According to the datasheet, these chips hold 128KB. So how many of those chips does your board carry, what's the actual total amount of memory in that scope then ?

There's 32 of them on the top side of the board, which is good for 4M samples. I've been told here that there's another 32 on the backside, but I haven't removed the board yet to check that. My TDS 754D has AS7C256 SRAMs which is good for Option 1M and 500k samples. See my discussion here: https://www.eevblog.com/forum/repair/tds-754d-swapping-as7c256-srams-by-as7c1024-to-enable-1m-option/.

[Vince]

OK OK.....so plenty of RAM in this scope indeed... a strange scope you have indeed... I am curious to see what the problem is, if/once you find out.

I see Terra said the options are set in the NVRAM, so I guess you could look into  that as well.
Maybe there is something wrong with that NVRAM. Maybe back it up, then clear it to reset it, and see what that does.

You should at the very least try to go into whatever menu can't remember, and activate the "Reset to Factory defaults"  or something like that.
Helped me once, I remember...

Also check the error log, maybe there is a H/W problem that keeps the scope from using part of your RAM chips. Bad chips somewhere.

Maybe check the GPIB instruction set, maybe there are commands that could tell you what memory options is there, and what is the actual amount of physical RAM it can detect. See if that's consistent with the boot screen.

There has to be an explanation to all this.... 

[TERRA Operative]

Photo of boot screen banner please! (Does it show 1M or 2M?)

Did you flick the write protect switch, and send the command PASSWORD PITBULL before the command to enable the 2M option?

[tverbeure]

Photo of boot screen banner please! (Does it show 1M or 2M?)

Relax, relax! This scope's CRT doesn't work so I'd have to play Tetris with my lab bench-with-VGA-monitor which currently has a Very Complicated setup on it that I don't want to mess up! (That's also why you'll have to wait a bit for an NVRAM dump after the v6.6e update.)



But here's the banner before I did the v6.0e to v6.6e update.



After the update, and after re-enabling Option 2M with the pit bull, it shows the same banner but with v6.6e, of course. I don't have a photo of that.

Did you flick the write protect switch, and send the command PASSWORD PITBULL before the command to enable the 2M option?

Of course! Over the past 3 weeks, I've become very skilled in the art of handling the pitt bull and the write protect switch.