Electronics-Related.com
Forums

The HP 3561A - still useful in 2022?

Started by bitrex January 7, 2022
It's a cheap and cheerful (relatively compact) boat anchor, I'm curious 
if it could be good for doing any useful low-frequency control system 
loop analysis a la:

<https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf>

Unfortunately it only seems to include some kind of shaped noise as a 
source for gain/phase measurement but might still be useful for some 
stuff.. The beefier 3562 included a curve fitting algorithm to fit an 
s-domain transfer function to the gain/phase response automatically but 
nowadays I think you could just dump the data to Matlab/Simulink and do 
a better job of it anyway

Here's the specs for 3561A:

<https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf>
bitrex wrote:
> It's a cheap and cheerful (relatively compact) boat anchor, I'm curious > if it could be good for doing any useful low-frequency control system > loop analysis a la: > > <https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf> > > Unfortunately it only seems to include some kind of shaped noise as a > source for gain/phase measurement but might still be useful for some > stuff.. The beefier 3562 included a curve fitting algorithm to fit an > s-domain transfer function to the gain/phase response automatically but > nowadays I think you could just dump the data to Matlab/Simulink and do > a better job of it anyway > > Here's the specs for 3561A: > > <https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf> >
Haven't used one of those since about 1988. The two-channel ones are much better. I used to have a 3562A, which I liked a lot, and a 36660 that I didn't like as well because it used one digitizer for both channels, so that it only went up to 50 kHz in two-channel mode. I now have a 35665A, which I paid $300 for. It has the same problem but has extra functionality. BTW there used to be an outfit called GLK Instruments that would sell you a customized ROM that turned on all the options. (I have one--it was $50 IIRC.) Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
On 1/7/2022 10:52 AM, Phil Hobbs wrote:
> bitrex wrote: >> It's a cheap and cheerful (relatively compact) boat anchor, I'm >> curious if it could be good for doing any useful low-frequency control >> system loop analysis a la: >> >> <https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf> >> >> Unfortunately it only seems to include some kind of shaped noise as a >> source for gain/phase measurement but might still be useful for some >> stuff.. The beefier 3562 included a curve fitting algorithm to fit an >> s-domain transfer function to the gain/phase response automatically >> but nowadays I think you could just dump the data to Matlab/Simulink >> and do a better job of it anyway >> >> Here's the specs for 3561A: >> >> <https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf> >> > > Haven't used one of those since about 1988.&nbsp; The two-channel ones are > much better.&nbsp; I used to have a 3562A, which I liked a lot, and a 36660 > that I didn't like as well because it used one digitizer for both > channels, so that it only went up to 50 kHz in two-channel mode. > > I now have a 35665A, which I paid $300 for.&nbsp; It has the same problem but > has extra functionality.&nbsp; BTW there used to be an outfit called GLK > Instruments that would sell you a customized ROM that turned on all the > options.&nbsp; (I have one--it was $50 IIRC.) > > Cheers > > Phil Hobbs >
Thanks, I have a USB/PC-based VNA that works well enough for the small amount of RF stuff I do. But for low-frequency analyzers the options seem more limited. The 3577A seems like the most general-purpose tool but a certified refurb runs about 2k and then probably another 2k for the S parameter test set which seems nice to have but I don't know how often I'd use it...I may just keep my eye out for a working 3562 for around 1k that doesn't seem unreasonable for what you get. My short list for lab upgrade this year is a low-frequency analyzer and a nice bench LCR meter/component analyzer but I tend to get deep in a hole looking at the options, I don't really want to spend even $300 on this kind of Shenzen-special stuff: <https://www.amazon.com/East-Tester-Capacitance-Resistance-100hz-10KHZ/dp/B08FR2GB92/> But then pretty quickly you're pushing several grand or more for units from BK, etc.
On 1/7/2022 12:57 PM, bitrex wrote:
> On 1/7/2022 10:52 AM, Phil Hobbs wrote: >> bitrex wrote: >>> It's a cheap and cheerful (relatively compact) boat anchor, I'm >>> curious if it could be good for doing any useful low-frequency >>> control system loop analysis a la: >>> >>> <https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf> >>> >>> Unfortunately it only seems to include some kind of shaped noise as a >>> source for gain/phase measurement but might still be useful for some >>> stuff.. The beefier 3562 included a curve fitting algorithm to fit an >>> s-domain transfer function to the gain/phase response automatically >>> but nowadays I think you could just dump the data to Matlab/Simulink >>> and do a better job of it anyway >>> >>> Here's the specs for 3561A: >>> >>> <https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf> >>> >> >> Haven't used one of those since about 1988.&nbsp; The two-channel ones are >> much better.&nbsp; I used to have a 3562A, which I liked a lot, and a 36660 >> that I didn't like as well because it used one digitizer for both >> channels, so that it only went up to 50 kHz in two-channel mode. >> >> I now have a 35665A, which I paid $300 for.&nbsp; It has the same problem >> but has extra functionality.&nbsp; BTW there used to be an outfit called >> GLK Instruments that would sell you a customized ROM that turned on >> all the options.&nbsp; (I have one--it was $50 IIRC.) >> >> Cheers >> >> Phil Hobbs >> > > > Thanks, I have a USB/PC-based VNA that works well enough for the small > amount of RF stuff I do. But for low-frequency analyzers the options > seem more limited. The 3577A seems like the most general-purpose tool > but a certified refurb runs about 2k and then probably another 2k for > the S parameter test set which seems nice to have but I don't know how > often I'd use it...I may just keep my eye out for a working 3562 for > around 1k that doesn't seem unreasonable for what you get. > > My short list for lab upgrade this year is a low-frequency analyzer and > a nice bench LCR meter/component analyzer but I tend to get deep in a > hole looking at the options, I don't really want to spend even $300 on > this kind of Shenzen-special stuff: > > <https://www.amazon.com/East-Tester-Capacitance-Resistance-100hz-10KHZ/dp/B08FR2GB92/> > > > But then pretty quickly you're pushing several grand or more for units > from BK, etc.
There's also the HP 3563A "Control System Analyzer" which seems like a pretty useful piece for LF gain/phase also. HP stuff from this era is weird it seems like there's a lot of cross-over and you could turn some tools into any other tool just through options and ROM packages...
On Fri, 7 Jan 2022 04:13:06 -0500, bitrex <user@example.net> wrote:

>It's a cheap and cheerful (relatively compact) boat anchor, I'm curious >if it could be good for doing any useful low-frequency control system >loop analysis a la: > ><https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf> > >Unfortunately it only seems to include some kind of shaped noise as a >source for gain/phase measurement but might still be useful for some >stuff.. The beefier 3562 included a curve fitting algorithm to fit an >s-domain transfer function to the gain/phase response automatically but >nowadays I think you could just dump the data to Matlab/Simulink and do >a better job of it anyway > >Here's the specs for 3561A: > ><https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf>
These were awesome instruments in their time ! I would love to have one just for nastalgic reasons cuz we used to use these in the 80s. boB
On 1/8/2022 3:23 AM, boB wrote:
> On Fri, 7 Jan 2022 04:13:06 -0500, bitrex <user@example.net> wrote: > >> It's a cheap and cheerful (relatively compact) boat anchor, I'm curious >> if it could be good for doing any useful low-frequency control system >> loop analysis a la: >> >> <https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf> >> >> Unfortunately it only seems to include some kind of shaped noise as a >> source for gain/phase measurement but might still be useful for some >> stuff.. The beefier 3562 included a curve fitting algorithm to fit an >> s-domain transfer function to the gain/phase response automatically but >> nowadays I think you could just dump the data to Matlab/Simulink and do >> a better job of it anyway >> >> Here's the specs for 3561A: >> >> <https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf> > > > These were awesome instruments in their time ! > > I would love to have one just for nastalgic reasons cuz we used to use > these in the 80s. > > boB > > >
Mainly looking for a tool for profiling e.g. power inductors and audio transformers, plus some control loop analysis. I'm not sure the 3561 is the best tool perhaps, the lack of swept-sine and other signal injection types is kind of a bummer. The 3562 and 3563 seem like more appropriate tools for that job. There's a spot down the road from me that carries a lot of this stuff surplus and not paying shipping would be great, I'm planning on picking up a 3478A and a power supply or two there shortly: <https://www.bmisurplus.com/product/hewlett-packard-hp-3478a-digital-multimeter/> I'll ask if they ever get some of those two in stock, a G seems a fair price for one in as-is working condition. There was a 3561A on eBay for $200 plus shipping recently but someone grabbed it first. It was functional but throwing a level 1 error code and I looked it up, like "DMA writeback time violation" or somesuch. That doesn't sound necessarily trivial, a bad memory board or IC to hunt down perhaps. Don't really want to fuss with it...
On 1/8/2022 12:51 PM, bitrex wrote:
> On 1/8/2022 3:23 AM, boB wrote: >> On Fri, 7 Jan 2022 04:13:06 -0500, bitrex <user@example.net> wrote: >> >>> It's a cheap and cheerful (relatively compact) boat anchor, I'm curious >>> if it could be good for doing any useful low-frequency control system >>> loop analysis a la: >>> >>> <https://www.hpmemoryproject.org/an/pdf/an_243-6.pdf> >>> >>> Unfortunately it only seems to include some kind of shaped noise as a >>> source for gain/phase measurement but might still be useful for some >>> stuff.. The beefier 3562 included a curve fitting algorithm to fit an >>> s-domain transfer function to the gain/phase response automatically but >>> nowadays I think you could just dump the data to Matlab/Simulink and do >>> a better job of it anyway >>> >>> Here's the specs for 3561A: >>> >>> <https://82fai3h5wg8220a0l3ilfiz1-wpengine.netdna-ssl.com/wp-content/uploads//2015/09/HP3561A.pdf> >>> >> >> >> These were awesome instruments in their time ! >> >> I would love to have one just for nastalgic reasons cuz we used to use >> these in the 80s. >> >> boB >> >> >> > > Mainly looking for a tool for profiling e.g. power inductors and audio > transformers, plus some control loop analysis. I'm not sure the 3561 is > the best tool perhaps, the lack of swept-sine and other signal injection > types is kind of a bummer. > > The 3562 and 3563 seem like more appropriate tools for that job. There's > a spot down the road from me that carries a lot of this stuff surplus > and not paying shipping would be great, I'm planning on picking up a > 3478A and a power supply or two there shortly: > > <https://www.bmisurplus.com/product/hewlett-packard-hp-3478a-digital-multimeter/>
Getting these calibrated to a high standard is pretty cheap, like 50 bucks per unit.