Stanford Research Systems DS345, really loud [fix inside]

[bayjelly]

I got this awesomely well preserved SRS DS345 from eBay, so pristine on the outside and even the inside (no dust at all!) that I am wondering whether it was used much at all.

There is only one problem with the unit: It is really, really loud. Loud enough that I am relieved when I turn it off, which is not optimal.

Does anyone know whether those units are known to be that loud, and, more importantly, has anyone had good results with replacing the fan? If so, what fan did you use?

[lowimpedance]

Yes the fans are rather loud !, more so if you use it in a quiet setting like your home lab otherwise less noticeable in standard work labs.
I have not replaced the fan ,and not checked but it may be a mains powered one and those are rather noisy when a decent amount of air movement is required.

[edpalmer42]

Mine isn't particularly loud, but it has an annoying whine that is rather unpleasant.  I think mine is rather old (date codes of 1991-1992) so it's probably the bearings.  I'll change it eventually and probably try to use rubber fan mounts to help isolate any bearing noise.

Lowimpedance, the fan is 24V and doesn't seem to be moving a lot of air.  When I change mine, I'll consider a 12V fan because they're so much more common and add a dropping resistor.

Ed

[edpalmer42]

I've got a question for you guys.  What firmware version do you have?  It displays the version right after you turn it on.  I've got 1.03 which is the same as the one Shahriar fixed on his video.  I'm wondering if there's a more recent version.

Ed

[bayjelly]

I've got 1.04! Unfortunately I didn't remember looking at date codes when I had the unit open for cleaning (which wasn't necessary at all), but the S/N is 31426. Stamped on the back, and also displayed when powering on.

Wonder what's different between 1.03 and 1.04?

[lowimpedance]

Double checked a pair here at work and both are version 1.03 and both have quite a whine to their fans and they sound the same, (I recall that this whine has always been the same and has not got louder with time). approx. 1991 vintage.

 Okay I took the cover off for a look and for reference the fan details are:

Sanyo Denki model number 109P0624H402
24V @ 0.07A

 Powered up with the cover and PCB out of the way, looks to be quite a high RPM and seems to be moving a fair bit of air, and whines like a ........

[edpalmer42]

I can't find any release notes so I have no idea whether there's a significant difference between the versions.  I looked through all the photos I could find as well as ebay listings.  On ebay there were a couple of units that had V1.04, but most didn't have a photo of that info.

I'm working through a couple of minor quirks with my unit and wondered whether they're fixed in a newer version.  I have the probably unreasonable expectation that a synthesized function generator should be able to output the frequency that you key in with a tolerance comparable to the unit's resolution.  I realize that where DDS is involved, things can get 'interesting'.

First, I set the output for 1 Hz and sync both my counter (Fluke PM6681) and the DS345 from my Rubidium house standard (HP 5065A).  Then, I measure the period of the sync output and get a standard deviation in the range of 100 us. (which is also 100 uHz).  But, the DS345 claims to have a resolution of 1 uHz.  Seems a little silly to claim a resolution that's buried under noise that's 100 times larger!  I'm still exploring that to see if there are some bad components that need replacing.

But I figured out how to use the arb capability to bypass the analog processing on both the main and sync outputs and access the digital core.  Now, when I look at the sync output, the standard deviation of the period is a more reasonable 37 ps., an improvement of about 2.5e6 times (!!) and is stable from test to test.  Unfortunately, there's a tiny period error in the range of 50 ns.  Yes, it's tiny.  No, it won't bother 99.999% of the users.  I'm a time-nut and we love to obsess over these things!  This is the kind of thing that a firmware upgrade might affect.

Lowimpedance:  That's the same fan as mine.  The closest datasheet I could find is for a 109P0624H602 which is 24V@0.07A, 14.8 cfm, 31 dBA, 4200 RPM.

Ed

P.S.  Don't forget to go to their web site ( www.thinksrs.com ) and get their AWC software to create arbitrary waveforms.

[bayjelly]

Thanks for the fan model, that makes things much easier.

Should I attempt to read out the firmware for you?

[edpalmer42]

Thanks for the fan model, that makes things much easier.

Should I attempt to read out the firmware for you?

If it wouldn't be too much trouble, that would be great.  You should upload it to a site so that everyone has access to it.  I recommend KO4BB.com ( http://www.ko4bb.com/getsimple/index.php?id=manuals ).  Even though it says 'manuals', there are lots of firmware images there, too.

Ed

[Magiciaen]

Hi all,

The best fan I have found to replace the noisy one in the DS345 is the following:

NMB Technologies
2410SB-05W-B30-B00

The Digikey part number is P14744-ND.

It is whisper quiet, and I've also installed one in my Stanford Research SR650 (which uses the same fan).

If you want to change out any old crusty BNC connectors on the front, these are the best ones (with Teflon insulation):

Amphenol RF Division
112431

Digikey: ACX1048-ND

When I replaced the old caps in mine, they were an absolute pain to get off. They are bonded to the board with ATV. The best strategy I found to get them off was first to slide various exacto knife blades between the caps and board to slice the silicone, before desoldering. Just thought I'd mention that so no one damages their board.

[LapTop006]

It is whisper quiet, and I've also installed one in my Stanford Research SR650 (which uses the same fan).

Huh, my DS345 isn't too bad, but louder than my SR620.

The one that was problematically loud was the CG635 which takes a 40mm fan, luckily 12v. For now I have the best Digikey offered, but eventually probably a Noctua or one of the other super-quiet ones.

[bayjelly]

The best fan I have found to replace the noisy one in the DS345 is the following:

NMB Technologies
2410SB-05W-B30-B00

Thanks for finding this. I finally replaced the fan in my DS345 with exactly this part and it makes a world of a difference.

As an anecdote, my fiancée, who is usually very tolerant of my antics, carefully asked me how much a different, quieter unit would cost just a short time after I turned it on. So I kept it off when she was around, but did not use the device much even when she wasn't, because the noise was grating on my own nerves as well. After the fan replacement she did not notice anything anymore when I asked her!

It's still far from quiet, but definitely quieter and the noise is a bit more white with much less of that previous whine.

Maybe the previous fan was aged or defective? The unit was otherwise in absolutely astonishing condition, but it could be that some grease hardened up. The descriptions of the sound here pretty matched what I was hearing, though.

[Magiciaen]

Glad I could help, bayjelly!

One thing I've also noticed is there are two versions of the NMB fan; one made in Japan, and one made in China. Unfortunately, they have the same part number. The Japanese one seems to be a bit quieter for some reason.

Does your DS345 have the OCXO?

[quint]

Thanks for the fan recommendation.  The noise on my DS345 drives me a bit crazy so I am going to try the replacement.  For reference, mine has firmware version 1.04.
Has anyone managed to get the SRS arbitrary waveform composer software running on Windows 10 (with serial connection to the DS345)?  It seems to hang up on trying to establish a connection, with the message "Finding VISA resources... Please wait!" forever.

[quint]

Has anyone managed to get the SRS arbitrary waveform composer software running on Windows 10 (with serial connection to the DS345)?  It seems to hang up on trying to establish a connection, with the message "Finding VISA resources... Please wait!" forever.

To answer my own question, for anyone else who runs into this, you need to install the latest VISA software from NI and then the AWC software from SRS seems to work fine, at least over the serial interface connection.  I don't have the appropriate GPIB hardware to try that.