Standford Research DS345 Error

[bostonman]

This is unfortunate.

Silly question, are you sure it's actually 9? I feel the designers didn't do a good job making the errors legible because deciphering mine took a few minutes.

The designers should have just used whole incremental numbers instead of  trying to display words on an LCD display.

[SoundTech-LG]

I hear what you're saying... it should have been error 0 through error xxx. It sure looks like a 9 to me. The confusing part is they say all possible errors are listed in the manual.
Obviously others can be generated.

[bostonman]

Can it be this:

Offset O Err 9 Output has excessive DC offset when set to 0 (>±100mV). Can be a problem
with offset control or output amplifier.

This is on page 76 (assuming you have the same revision PDF manual I have).

Maybe the O is really a space holder for the 9?

[SoundTech-LG]

Thanks to you I took another look. They also mix uppercase and lower case letters which adds even more confusion. There is a self test error in the manual...
Shown as...

Offset G Error 9 The DC offset function gain is more than ±10% from nominal. Can be a
problem with DAC or output amplifier.

See how it is showing uppercase G ? Well the display is actually showing a lower case g...  which also really looks like a 9.

So I'm going with g.

[SoundTech-LG]

Disregard the 9 I put in from the manual...   that is only the response code when using commands from GPIB. (also confusing)

[bostonman]

Too funny, I actually just saw that G error message in the manual and realized maybe I was wrong to assume O was a place holder, but didn't think to consider little g versus capital G.

You posted the reply just before I had time to gather my thoughts and post an update.

The funny thing: the programmers must have went through an extensive amount of effort to display an actual 'word' on the screen. I can see if they didn't have intentions of creating a manual with a list of error codes, so they they tried making it an obvious error to the user, but they listed all error codes in the manual.

It would have saved lots of time just having (as you stated) error 1 - xxx.

[bostonman]

Thanks to you I took another look.

By the way, I appreciate the compliment. Many on this site have helped me, and, when I see posts by others on questions I can possibly help with, they already have numerous responses.

So it's nice to see that my contribution made a difference.

[SoundTech-LG]

Yes, you prompted me to delve deeper into this and we both agree the error codes are a source of much confusion!!!

[SoundTech-LG]

The programmers in their quest to simplify for the user, complicated it instead!

[SoundTech-LG]

So my original error, and your existing error as shown in the manual is...

Offset O Err   Output has excessive DC offset when set to 0 (>±100mV). Can be a problem
with offset control or output amplifier.

Shows the uppercase O  (for output?) but shows it on the display as lowercase o. Originally I thought it was a Zero being displayed!!! 

[bostonman]

I'll admit, I'm happy the manual has the error code list and schematics. I'll take having to figure out the hieroglyphics over not having access to such material.

I just can't imagine the additional programming time and time to think of a 'word' to display on the screen that went into the code list. Not to mention others on the project approved such craziness.

[SoundTech-LG]

CERTAINLY!  The crazy thing too, they did it all for the USER! 

[bostonman]

I think it was for job security so the software engineer(s) could have a job for life and everyone need to ask him/her for assistance with interpreting error codes.

Although, a decoder wheel could have been used where 9 = g, o = O, I = 1, oh, and maybe you need to hold the unit in front of the mirror upside down for other, more complex error codes.

[SoundTech-LG]

Ah...  you're on to something now!

My plan is back to poking in the output amp where the original error was generated. Hopefully it doesn't just start working again like last time.

[SoundTech-LG]

I wonder if SRS still has it's User Friendly Programming Dept. ? They produce so few new products it's surprising they are still around at all. They are just over the hiil from me in Sunnyvale, CA. (Silicon Valley)

[bostonman]

Not to deviate away from the main topic in this thread (as I tend to get frustrated when I'm reading through a thread that is on key to an issue I'm trying to fix and it suddenly takes a sharp left), but I am big into the Commodore 64 and see old software company advertisements in PDFs and wonder what occupies that space now where such a large company once resided.

In some cases the building appears to be present (using Google maps), but it's amazing how many companies come and go. Relatively recent I came across a company in MA when browsing an old C64 magazine; or I'll see people who wrote into magazines who are MA based. You have to wonder where those people are now and/or if they realize they can obtain (in PDF) the question they wrote to a magazine as a kid.

Hopefully the problem with your SRS doesn't vanish again because it can be frustrating when you want to use that for testing and have to hold your breath on whether it will work (or is producing an accurate output).

[SoundTech-LG]

One source I use sometimes in desperation is the wayback machine https://archive.org/web/web.php for information on the past. Not always 100% on everything, but a good resource.

The other possibility of course is it will revert back to the error you are seeing, (my original error).

[SoundTech-LG]

So without boring detail, but after many hours of troubleshooting, replacement of suspect ICs (put on sockets), and resoldering many areas (specifically the output  diff to SE amp), the unit now passes all tests, and the AutoCal procedure. It also seems to meet specs.
However, I do not trust it, and fully expect something will pop up again, although it has been running trouble free for several days now. No idea exactly what I did that actually got it to this point. More luck, than skill.

[bostonman]

I know you're not comfortable with the "repair", but hopefully this is good news and it remains working.

Most likely you would have been happier seeing something wrong like a cold solder joint or a bad component.

Unfortunately I haven't returned to analyzing mine yet due to my soldering iron breaking. Thankfully I have a junk iron that somewhat handles the components on the Weller PCB. Hopefully I can get my iron repaired soon because, as you know, without an iron, all repair projects get placed on the back burner.

[bostonman]

I'm resurrecting an old thread as I've been working on my broken unit with a 'Offset 0 error code'.

I'm measuring U600 and getting 5.6v on pin 2, 5.4v on pin 3, and the output is pegged at -12.5v.

Obviously pins 2 and 3 should be almost equal which is why the output is pegged. From what I can tell, it looks like I need to trace OUT_OFF (label on the schematic on page 147), but I haven't located it.

Has anyone worked on this section of the circuit, and has anyone traced the OUT_OFF label to the connecting point?

[SoundTech-LG]

Unfortunately there are quite a few reasons why U600 gets an undesirable offset DC. This is exactly where I have been spending time looking at all the possible ways. If you want to look at one, OUT_OFF, that's the actual voltage you can punch in at the front panel for offsets you desire. That comes from the top board at U107A pin1, via pin 25 on the 60 pin cable (look at ckt. page  DDS ribbon, a/d's and d/a's for System) but it is not labeled at all on that page! So, then, you have to go to ckt. page  DDS Ribbon Cable, Trig, and Sync Select, to see the other end of the cable at JP401 pin 25 and THEN you will see it marked!

Not sure of your unit, but mine is quite early on, and they (SRS) made a lot of changes to the boards, and do not document anything or give boards any Rev numbers. The only way I discovered this was from purchasing a couple other units with later S/N's to help TS this beast.
In my case I found a few things, and blew up a few things in the output amp along the way. All those were cleared up. what I am left with is probably why they made some changes in the ckt. that drives U600. Oh, and by the way...  SRS has given us two U600s in this unit!!!  the other is the AD834 on ckt. page DDS Filters and Doubler. But, my issue seems to be the Gain Adjust ckt. that precedes Doubler AD834 U702 on ckt. page DDS Sync and Gain Adjust. There are two signals here. One is the main function Gen Signal, the other, DC, in Gain Adjust. Combined up in the AD834, then sent through two 50 ohm coax to the lower board amp into U600. You'll see a red arrow for the DC levels. My issue seems to be DC DRIFT over time...  a LONG time. Days. But eventually causes enough offset to give errors in self test. This seems to be where they made board changes before the AD834 in the DC Gain ckt., using different op amps than shown in the schematic. I'm not convinced it is worth modding the board to match the later ones. The curious thing is they did not change IC's, they just re-routed traces to use different ICs.
Hope this might save you time. Check your S/N, and compare with other photos of units on Ebay.
I did find a way to reset this offset, and it will again last for days on end, and eventually trip with errors again. The AD834s are kind of strange with this. Maybe there is a fix for it.

[bostonman]

After posting about trying to locate where OUT_OFF is on the schematic, I manually traced the circuit and discovered it goes to pin 1 and then got confirmation from you.

I must have spent two-hours looking at the schematic trying to trace OUT_OFF and can't believe it's not on the schematic You're also correct, there are two U600 chips which threw me off for a bit.

From what I can tell, and comparing to a good one, it looks like U600, pin 3, is the culprit causing the output of the OP-AMP to go to the rail. It's my opinion the board is built well and a cold solder joint wouldn't be the issue. I tried removing the square chip in the socket and cleaning it just in case corrosion is causing an open connection, I applied pressure on the ribbon cable for the same reason (and cleaned it), but can't find anything related to a bad connection.

A thought I had was to remove the battery. Maybe this would cause a hard reset. Have you tried this?

[SoundTech-LG]

Hi,

I don't believe it is a bad connection problem. Nothing leads me (in my unit) to that conclusion. Yours may be different.

Well, changing the battery for a new one might be good (it's old for sure), but somehow doubt that will resolve any offset problems. I think the recall-able factory CAL settings are stored in a non-volatile memory, so battery swaps only destroy user settings. Pretty sure. It will be interesting if you try it to know the results. May as well put a new battery in at the same time if you do it.

One thing I tried is diconnecting, and isolating U600 pin 6 from the transistor buffers on the output. Check to see if the trace has zero volts, or close to that, and put zero volts on the trace if it's close. Those 2n3904, and 2n3906 transistors blow easily, and short, check them. Also, one of the traces there is extremely close to the transistor pin pad holes. Then look at those inputs on pin2, and pin3 again. U600 is pretty hardy, not seeing issues with it. The output amp should also have close to zero DC offset now.

Ultimately (in my case), the real issue causing offset was not U600 circuit, or the output amp, but the offset coming from U702 AD834 over the coax from the upper board. The open collector outputs on it, require a bias voltage higher than it's main supply bias. That voltage comes from the lower board in the output amp over the two coax. R615, and R616 supply the voltage from U602 6.8V Reg. to U702 AD834 collector outputs. I found that lowering that voltage on either collector output (at the coax) with a resistor to ground of about 600 ohms, and then removing it, would reset the offsets, and then all good for days on end. Ultimately creeping back again, only to trip into error once more.

I also swapped out several op amps, and U100 bi-lateral switch in that OUT_OFF circuit to no avail. Also suspecting, but never changing the D/A converter U109 in that ckt. Very expensive, obsolete, and ebay only replacements.

Good luck in your TS efforts!

[bostonman]

I have a SRS DS344 (?) that doesn't have a reference. That has a dead battery (it measures about 0.3v) and gives an error on power up, but then doesn't return providing I keep AC plugged in during reboots. This should just be a battery issue.

While I was at it, I measured the voltage in my bad DS345, and it was fine. The battery isn't very expensive, but it's not cheap either. I may hold off buying a replacement battery since I don't really use the non-referenced version often; nor is the error an issue since I can re-power it.

As for the bad DS345, I'm uncertain how much more time I'll devote to it. Unfortunately I have many projects that have popped up lately and lingering ones. Last week I was looking at my equipment trying to figure out which ones were broken, and, since I dislike keeping broken stuff, thought to give it another look in hopes the error went away or a simple wiggle of something would fix it. I even tried moving the little pots one way and then back into the same position hoping they were dirty. With the exception of the other unit with a dead battery, this is my only piece of test equipment that remains broken.

[SoundTech-LG]

We all hope for, and look for the easy fix, but sometimes it does not happen. This has been one of, if not the most, challenging pieces of equipment I have troubleshot. The more challenging a problem is, the more you stand to learn. It is of course necessary to weigh the worth of your time, vs. the worth of what you may learn, and of course the dollar value of the equipment as well. I have not given up yet, but I may set aside, and go back to it again. Sometimes that helps too. Brain refresh.