Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
HP8565 hacking/mods
|
swingbyte
Hi fellow hp'ers,
I'm the new owner of a HP8565 SA and think its pretty good. However, the hacker in me is already pondering modding it. In particular, I have been thinking of replacing the storage pcb with a new one, only operating the tube in conventional mode and using the rest of the space to add a digitiser/playback system. Has anyone else attempted something similar? Although digital systems and ADCs are now fast enough to do this - I chose the 8565 because it YIG RF systems still seem current state of art - and I could always digitise the rear outputs with something to avoid having to use the long persistence. I was prompted to write this after reading John Mills email about trying to increase the stability of his 8566 YIG and stated thinking maybe the 8565 could benefit from this? ( by the way - great web site you have!) The 8565 was at the price I could afford, I would have gone for an 8569 and would LOVE an 8566 - but here in Australia they're expensive - even through ebay - the shipping cost kills any bargains! But this is a hobby and I like hacking and modding. Another thought was to route the if signal to a DSO and FFT for greater resolution - would the 8565 LO be stable enough? Thanks for any thoughts Tim ( no call sign yet) |
|
John Miles
I was promptedThanks! You should probably check out the Nov-Dec '07 QEX article by Roland Cordesses: The 8565 is similar enough to the 141T (basically a newer, more compact version) that the same techniques should be usable with it. Another thought was to route the if signal to a DSO and FFT for greaterPossibly. I'm not familiar with this particular unit but if you do this, you will quickly learn whether or not it's stable enough. :) Then the real hacking can begin. -- john, KE5FX |
|
swingbyte
Hi John,
toggle quoted message
Show quoted text
Yes I have seen that unit - I was pondering something more sophisticated. Along the lines of 100MSa/s to FPGA to digitize x,y or if and an embedded web-server to get the data out. I figure I might have enough space on that board and use the control inputs for other purposes such as selected storage mode or other function since they wouldn't be controlling the tube any more. Still early ponderances - I only just got the SA and need to play with it a bit more!! I've decided its time to move into the exciting world of RF and high speed digital - spent enough of time on slow - medium speed digital and software - besides - that's what I do at work. I was interested in your hybrid 2GHz vco and possible use as a yig stabilizer. I just realised that we looked at using your pc sound system where I work. We're now moving from SGIs into pcs and may be looking at it again. :-) Tim John Miles wrote: I was promptedThanks! You should probably check out the Nov-Dec '07 QEX article by Roland Another thought was to route the if signal to a DSO and FFT for greaterPossibly. I'm not familiar with this particular unit but if you do this, |
|
lothar baier
ok , you got me confused here, the 8565 does not have a storage pcb since it uses a storage CRT , as far as connecting a digitizer that should be fairly easy, as far as i remember the 8565 has a interface connector to connect a storage normalizer (8750) on the back panel, this connector gives you acess to all the signals you need for this task.
However if you put in the time and effort to go for digital storage why dont you ponder the idea of discarding the CRT at all and replace it with a LCD panel, there is a bunch of embedded controllers that offer LCD/TFT interface build in, all you need is a microcontroller, A/D converter and a Surplus TFT panel and youre ready to go swingbyte <swingbyte@...> wrote: Hi fellow hp'ers, I'm the new owner of a HP8565 SA and think its pretty good. However, the hacker in me is already pondering modding it. In particular, I have been thinking of replacing the storage pcb with a new one, only operating the tube in conventional mode and using the rest of the space to add a digitiser/playback system. Has anyone else attempted something similar? Although digital systems and ADCs are now fast enough to do this - I chose the 8565 because it YIG RF systems still seem current state of art - and I could always digitise the rear outputs with something to avoid having to use the long persistence. I was prompted to write this after reading John Mills email about trying to increase the stability of his 8566 YIG and stated thinking maybe the 8565 could benefit from this? ( by the way - great web site you have!) The 8565 was at the price I could afford, I would have gone for an 8569 and would LOVE an 8566 - but here in Australia they're expensive - even through ebay - the shipping cost kills any bargains! But this is a hobby and I like hacking and modding. Another thought was to route the if signal to a DSO and FFT for greater resolution - would the 8565 LO be stable enough? Thanks for any thoughts Tim ( no call sign yet) --------------------------------- Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. |
|
John Miles
Hi John,Keep in mind that there's no benefit to digitizing post-detected trace signals (x and y) at rates beyond the bandwidth of the widest video filter, typically a few MHz at most. And if you digitize the final IF, the bandwidth may still be limited to a few MHz by other filters in the SA's signal chain. If you are that ambitious, I think the best approach -- meaning, the one that would benefit others the most, and require you to do the least original engineering work -- would be to launch a project using the USRP or other GNU Radio peripherals as a general-purpose spectrum analyzer back end. Keep it simple at first or you'll never get anywhere! Trust me on that... I figure I mightIt's definitely a good idea to spend time coming to grips with the limitations of what you have. The obvious specification deficiencies aren't always the ones that will end up limiting your measurements. I was interested in your hybrid 2GHz vco and possible use as a yigHeh... that will teach you all you (n)ever wanted to know about noise and stability. Building a clean YIG synthesizer is a challenging project. I still haven't managed to do as well as the old HP designs, even with much-newer opamps and other parts... which is one reason I chickened out when I realized I was faced with that sort of problem in my 8566 'redesign' project. It will have to wait until I can devote a lot more time to it. I just realised that we looked at using your pc sound system where IOh, you mean the sound system I'm supposed to be working on right now. Oh, yeah, THAT. :) -- john, KE5FX |
|
one problem i find is that the market for stuff like this is soo small that
toggle quoted message
Show quoted text
it isnt worth my time. i cant afford to spend alot of time on things that have very little return at oner time i had enough money to play for long periods of time that isnt the case any longer im not as fotunate as john miles. ----- Original Message -----
From: John Miles To: hp_agilent_equipment@... Sent: Saturday, November 17, 2007 3:16 PM Subject: RE: [hp_agilent_equipment] HP8565 hacking/mods > Hi John, > Yes I have seen that unit - I was pondering something more > sophisticated. Along the lines of 100MSa/s to FPGA to digitize x,y or > if and an embedded web-server to get the data out. Keep in mind that there's no benefit to digitizing post-detected trace signals (x and y) at rates beyond the bandwidth of the widest video filter, typically a few MHz at most. And if you digitize the final IF, the bandwidth may still be limited to a few MHz by other filters in the SA's signal chain. If you are that ambitious, I think the best approach -- meaning, the one that would benefit others the most, and require you to do the least original engineering work -- would be to launch a project using the USRP or other GNU Radio peripherals as a general-purpose spectrum analyzer back end. Keep it simple at first or you'll never get anywhere! Trust me on that... > I figure I might > have enough space on that board and use the control inputs for other > purposes such as selected storage mode or other function since they > wouldn't be controlling the tube any more. Still early ponderances - > I only just got the SA and need to play with it a bit more!! I've > decided its time to move into the exciting world of RF and high speed > digital - spent enough of time on slow - medium speed digital and > software - besides - that's what I do at work. It's definitely a good idea to spend time coming to grips with the limitations of what you have. The obvious specification deficiencies aren't always the ones that will end up limiting your measurements. > I was interested in your hybrid 2GHz vco and possible use as a yig > stabilizer. Heh... that will teach you all you (n)ever wanted to know about noise and stability. Building a clean YIG synthesizer is a challenging project. I still haven't managed to do as well as the old HP designs, even with much-newer opamps and other parts... which is one reason I chickened out when I realized I was faced with that sort of problem in my 8566 'redesign' project. It will have to wait until I can devote a lot more time to it. > I just realised that we looked at using your pc sound system where I > work. We're now moving from SGIs into pcs and may be looking at it > again. :-) Oh, you mean the sound system I'm supposed to be working on right now. Oh, yeah, THAT. :) -- john, KE5FX |
|
John Miles
Heh... I tend to go for a couple of years at a time without doing much
toggle quoted message
Show quoted text
electronics work at all, selling all my gear on eBay and going back to the real world to rebuild my finances. Eventually I feel compelled to return to the workbench, where I spend a lot of time and money answering questions that could, frankly, be better resolved for free in a library. I don't recommend this approach to others, especially those with families and other real-world responsibilities. :) -- john, KE5FX one problem i find is that the market for stuff like |
|
i have the same problem john
toggle quoted message
Show quoted text
deling with freeloaders that dont want to crank their heads or geet b their hands dirty maybe we should go on strike for them ----- Original Message -----
From: John Miles To: hp_agilent_equipment@... Sent: Saturday, November 17, 2007 3:47 PM Subject: RE: [hp_agilent_equipment] HP8565 hacking/mods Heh... I tend to go for a couple of years at a time without doing much electronics work at all, selling all my gear on eBay and going back to the real world to rebuild my finances. Eventually I feel compelled to return to the workbench, where I spend a lot of time and money answering questions that could, frankly, be better resolved for free in a library. I don't recommend this approach to others, especially those with families and other real-world responsibilities. :) -- john, KE5FX > one problem i find is that the market for stuff like > this is soo small that > it isnt worth my time. > i cant afford to spend alot of time on things that have > very little return > at oner time i had enough money to play for long > periods of time > that isnt the case any longer im not as fotunate as john miles. > ----- Original Message ----- > From: John Miles > To: hp_agilent_equipment@... > Sent: Saturday, November 17, 2007 3:16 PM > Subject: RE: [hp_agilent_equipment] HP8565 hacking/mods > > > > Hi John, > > Yes I have seen that unit - I was pondering something more > > sophisticated. Along the lines of 100MSa/s to FPGA to digitize x,y or > > if and an embedded web-server to get the data out. > > |
|
John Miles
Hmm, I see I left some room for that to be taken the wrong way. When I
toggle quoted message
Show quoted text
said, "Eventually I feel compelled to return to the workbench, where I spend a lot of time and money answering questions that could, frankly, be better resolved for free in a library," I was referring to MY OWN questions, not anyone else's. Meaning, if I had any sense, I'd spend a couple of years in an EE classroom rather than wandering the empirical path of Paracelsus, Ramanujan, and that nutcase down the street with the tinfoil on his windows. :) -- john, KE5FX -----Original Message----- |
|
lothar baier
Hmmm sounds kinda familiar, im going trough this cycle every few years, usually when im employed i packrat stuff especially microwave parts, machinery and test gear to build up my lab until my wife throws out divorce threats, then i lose my job and start selling stuff off left and right to make ends meet, usually when i reduced my stuff a bunch i find another job and the cycle starts again, its really easy to be a packrat when you are a design engineer especially for a large player, most distributors and manufacturers are very generous with free samples and tools, usually once a month you get a email with offers for some samples or a free development kit and usually i reply :) so stuff gathers up fast !
The last really useful thing was a PSOC kit from cypress, i just love those little mixed signal microcontrollers, they are very easy to programm, the C compiler is cheap and the devices are cheap and versatile not to mention they have devices with USB, the last project i did was a USB switch controller, well im still kinda working on the firmware for that :) John Miles <jmiles@...> wrote: Heh... I tend to go for a couple of years at a time without doing much electronics work at all, selling all my gear on eBay and going back to the real world to rebuild my finances. Eventually I feel compelled to return to the workbench, where I spend a lot of time and money answering questions that could, frankly, be better resolved for free in a library. I don't recommend this approach to others, especially those with families and other real-world responsibilities. :) -- john, KE5FX one problem i find is that the market for stuff like --------------------------------- Get easy, one-click access to your favorites. Make Yahoo! your homepage. |
|
lothar baier
Building a good YIG based synthesizer has its challenges i admitt, first problem is that usually YIGs are not as good in phasenoise to begin with, next thing is that you have to design your drivers carefully and use really low noise powersupplies and filtering, one issue is that with the tuning range of the YIG beein much wider than of a VCO noise and interference on the psu and tuning lines will take its toll on the phasenoise so even the driver and associated lines need to be shielded and filtered.
Next is the approach on how to reduce the Frequency, a frequency divider is nice however there are not too many out there covering 2-18GHz that offer low noise, another approach is to use a harmonic mixer to down convert the Signal to a lower IF range and then prescale and go to a PLL. Generally it never hurts to look at what HP did , study the design and learn from it, one thing about the 8566 is that it uses a 3-6GHz YIG and for some reason 3-6GHZ seems to be a sweet spot for low phasenoise, at least you get the idea if you look at datasheets, most low phasenoise YIGS are 3-6GHz, dont ask me why John Miles <jmiles@...> wrote: > Hi John, Yes I have seen that unit - I was pondering something moreKeep in mind that there's no benefit to digitizing post-detected trace signals (x and y) at rates beyond the bandwidth of the widest video filter, typically a few MHz at most. And if you digitize the final IF, the bandwidth may still be limited to a few MHz by other filters in the SA's signal chain. If you are that ambitious, I think the best approach -- meaning, the one that would benefit others the most, and require you to do the least original engineering work -- would be to launch a project using the USRP or other GNU Radio peripherals as a general-purpose spectrum analyzer back end. Keep it simple at first or you'll never get anywhere! Trust me on that... I figure I mightIt's definitely a good idea to spend time coming to grips with the limitations of what you have. The obvious specification deficiencies aren't always the ones that will end up limiting your measurements. I was interested in your hybrid 2GHz vco and possible use as a yigHeh... that will teach you all you (n)ever wanted to know about noise and stability. Building a clean YIG synthesizer is a challenging project. I still haven't managed to do as well as the old HP designs, even with much-newer opamps and other parts... which is one reason I chickened out when I realized I was faced with that sort of problem in my 8566 'redesign' project. It will have to wait until I can devote a lot more time to it. I just realised that we looked at using your pc sound system where IOh, you mean the sound system I'm supposed to be working on right now. Oh, yeah, THAT. :) -- john, KE5FX --------------------------------- Be a better sports nut! Let your teams follow you with Yahoo Mobile. Try it now. |
|
Peter Gottlieb
I have a degree, but the truth be known I learned at least 95% of what I know by playing on the bench, taking things apart, modifying radio gear, building my own test gear when I was a kid, and then working my way up in jobs in engineering before I decided to go and get an actual degree. Yes, I spent a lot of time in libraries, and had a couple of uncles in engineering who answered questions, gave me books, and generally pointed me in the right direction. When you love something and are truly driven, it is amazing what you can do!!
toggle quoted message
Show quoted text
Peter John Miles wrote:
|
|
swingbyte
HP call the board that controls the storage tube persistence, erase etc the storage assembly. I have been thinking of replacing the tube with and LCD - but that would cost more than the SA!! Unless I can get a VGA LCD the right size for ~$400 the cost is prohibitive. The tube is fine, and there are many CRTs still floating around that could be swapped in if no persistence was required. I was planning for a minimal modification and swapping one assembly board with a new one seems good. Yes I can just stick the outputs into my DSO and I have an 8569 effectively and that's what I will do to start with - but then I don't see the trace for a long time so I may do something for that.
toggle quoted message
Show quoted text
I have used some LCD panels at work for MFD simulations - but cost isn't a problem there! More suggestions welcome. Tim lothar baier wrote:
|
|
lothar baier
there are several companies that sell surplus LCD for about $100 or less, earthlcd is one example, changing to LCD would also have the advantage that you get some additional function like printer interface and USB for no additonal cost !
Doing the whole thing is fairly easy since all you have to do is process XY Data, the propellor microcontroller from parallax is perfect for that and you can get A/D conveters with 2 channels and a few K samples very cheap as well. You can start with a Propellor eval board which has a VGA connector and a pc monitor and go from there, take a look at the datasheet ! Of course you can also go down the complicated route and use a FPGA to do the same functionality, i build a XY Storage display on a Stratix board years ago just for the heck of it and it was quite a experience. 8565A are fairly cheap but keep in mind that the basic RF Functionality in between the 65 and 69 is about the same but the 8569 has digital storage and plot capabilities. swingbyte <swingbyte@...> wrote: HP call the board that controls the storage tube persistence, erase etc the storage assembly. I have been thinking of replacing the tube with and LCD - but that would cost more than the SA!! Unless I can get a VGA LCD the right size for ~$400 the cost is prohibitive. The tube is fine, and there are many CRTs still floating around that could be swapped in if no persistence was required. I was planning for a minimal modification and swapping one assembly board with a new one seems good. Yes I can just stick the outputs into my DSO and I have an 8569 effectively and that's what I will do to start with - but then I don't see the trace for a long time so I may do something for that. I have used some LCD panels at work for MFD simulations - but cost isn't a problem there! More suggestions welcome. Tim lothar baier wrote:
--------------------------------- Never miss a thing. Make Yahoo your homepage. |
|
swingbyte
HP must have been an amazing place to work in in the 1970s. It seems they did a lot of R&D in those years. I'm extremely grateful for their old service notes - more information and explanation than one sees these days. there's no great electronics industry in Australia - and from what I have read about HP and TEK it seems to be over in the USA as well. I have spent alot of time reading HP journals etc. I am always impressed with their work and I think I learnt more from them than some of my lectures at university. I still think all EEs should have to work as a tech for at least a couple of months.
toggle quoted message
Show quoted text
The 8566 yig controller is mighty impressive hybrid digital control/analogue signal system - something I would use as a starting point for anything I try to do. This is not intended to be a commercially viable solution - just buy a better more $$ SA if you want that - I was looking for the journey rather than the destination. At the same time - I'm not going to throw too much money into it - if I wanted to waste money I'd have a boat!! You have some very interesting points in your post (below) thanks. The 3-6 GHz sweet spot must have taken a bit of work to discover. Tim lothar baier wrote:
|
|
J Forster
First off, I have no dog in this fight, but it seems to me used laptops, especially lower performance ones, are dirt cheap (>$100) these days. Why not a simple interface that uses the parallel or USB port and let the OS and laptop do
toggle quoted message
Show quoted text
everything else. I would stay away from surplus LCDs because a continuing supply is problematic. -John lothar baier wrote: there are several companies that sell surplus LCD for about $100 or less, earthlcd is one example, changing to LCD would also have the advantage that you get some additional function like printer interface and USB for no additonal cost ! |
|
On Saturday 17 November 2007, lothar baier wrote:
Building a good YIG based synthesizer has its challenges i admitt, firstI think that Bob Pease wrote a good article or two about ripple rejection for power supplies. Google for "what's all this ripple rejection stuff anyway". He shows how to do it with high voltage supplies too. Cheers, Kuba |
|
Alexander Whiplash
--- Kuba Ober <ober.14@...> wrote:
...There are many sources of noise that can get onto the YIG tuning current. Power supply ripple will have already been knocked down to a very low level in any piece of HP gear that is working properly. Any instrument will have a regulated power supply to power the YIG driver circuits, and those supplies have very little ripple on their output(s). The YIG driver itself will have additional power supply rejection, perhaps 80 dB or more. The opamps and transistors and even some passive components in the YIG driver will impose their noise on the YIG tuning current and cause non-line related noise. This is usually dealt with by careful design and choice of components. Often an additional series RC is placed across the coil to shunt noise currents away. One can only take that so far before the tuning performance or loop bandwidth is too much affected. Units like the 8568 already have these "tricks" and I don't think that there is any low hanging fruit to be had in improving phase noise on it or comparable instruments. AW ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. |
|
lothar baier
designing for low phasenoise never has been a trivial task, this especially goes for wideband oscilators such as YIGs, you also have to drive the shielding of the tuning lines almost to a excessive level.
J Forster <jfor@...> wrote: Agreed, assuming the caqpacitors around the YIG tuning circuits are all up to spec. I've seen one or two (non Ta) go low C. Best, -John Alexander Whiplash wrote: --- Kuba Ober <ober.14@...> wrote:...There are many sources of noise that can get onto the --------------------------------- Get easy, one-click access to your favorites. Make Yahoo! your homepage. |
|
J Forster
Agreed, assuming the caqpacitors around the YIG tuning circuits are all up to spec. I've seen
toggle quoted message
Show quoted text
one or two (non Ta) go low C. Best, -John Alexander Whiplash wrote: --- Kuba Ober <ober.14@...> wrote:...There are many sources of noise that can get onto the |
to navigate to use esc to dismiss