G.Kleinschmidt said:
The HP 3581 uses crystal filters, 3 to 300 Hz BW. I think that it has a 100kHz IF and a VTO of 1.0 to 1.5 MHz. Time was when you could buy an HP3581 for $20, now they go for hundreds of dollars.
I have not yet worked with their chips, but I would imagine that the clock-through is going to require some post-filtering. MSI has some white-papers on their website which might be a bit more revealing.
You know, you can always use some programmable potentiometers for very quiet bandpass filtering -- use some JFETs for switches for the range (decade) capacitors. Analog Devices has an application note on the potentiometer part of this up on their website.
jackinnj said:
OK, but that HP is a sweeping frequency analyzer, like my SA200K, which also uses crystal filters. The VGA unit is a different kettle of fish, however.
I like the idea of a 60 band unit for the MkII VGA version, so that MSI chip with 6 channels per DIP is looking like the best option so far. Still don't know what they cost though and still haven't received a reply from MSI. Time so send another email.
Cheers,
Glen
G.Kleinschmidt said:
Apparently, $5.95 in 1000's. Which probably translates to $10-12/pc.
http://eepn.com/Locator/Products/ArticleID/34624/Action/Issue/34624.html
Great project!
Hi Glenn,
What a great idea!
I'm using an HP 3585A and have a broken 3580A to fix. I'm thinking that if you can output the data to a computer, you can then measure frequencies and levels very accurately. This also gives you the ability to document your other projects in great detail. That is one of the more difficult aspects of using the older gear.
Your first idea was really cool. I'll watch the second project you have decided to go with (I liked your first idea better). Also, it gives some practice with programming PIC devices. Nothing wrong there.
The newer DSO products do not do FFT very well unless you get into the more expensive products (I'd be extremely happy with an Agilent DSO-6000 series). Both noise and resolution are problems. Also, they are almost all 8 bit and I suspect you could get higher resolution with your focused design.
Good going!!
Best, Chris
Hi Glenn,
What a great idea!
I'm using an HP 3585A and have a broken 3580A to fix. I'm thinking that if you can output the data to a computer, you can then measure frequencies and levels very accurately. This also gives you the ability to document your other projects in great detail. That is one of the more difficult aspects of using the older gear.
Your first idea was really cool. I'll watch the second project you have decided to go with (I liked your first idea better). Also, it gives some practice with programming PIC devices. Nothing wrong there.
The newer DSO products do not do FFT very well unless you get into the more expensive products (I'd be extremely happy with an Agilent DSO-6000 series). Both noise and resolution are problems. Also, they are almost all 8 bit and I suspect you could get higher resolution with your focused design.
Good going!!
Best, Chris
Re: Great project!
Hmm... OK. That's not too bad, so long as I don't have to buy 1000......
Thanks Chris.
I'm actually going with and building all the designs detailed here. All will go up on my webpage eventually. It's just taking me ages because it's such a tedious process.
Cheers,
Glen
syn08 said:
Hmm... OK. That's not too bad, so long as I don't have to buy 1000......
anatech said:
Thanks Chris.
I'm actually going with and building all the designs detailed here. All will go up on my webpage eventually. It's just taking me ages because it's such a tedious process.
Cheers,
Glen
jackinnj said:
Neat, though I think my VGA display is going to look much cooler than that
Anyway, was occupied mainly with home maintenance stuff this weekend, but I did manage to knock off the complete band pass filter and peak detector board layout in the evenings.
I have revised the design slightly from 20 bands to 21 bands, as 21 half octave bands is a better mathematical fit into the compete audio (20Hz-20kHz) frequency spectrum.
Top layer:
Attachments
.....and the bottom layer.
One BC548, one BC558, 16 quad opamps (TLO74), 23 schotkey diodes, 143 capacitors and 300 resistors.
Bandpass Fc's are:
20Hz
28Hz
40Hz
57Hz
80Hz
113Hz
160Hz
226Hz
320Hz
453Hz
640Hz
905Hz
1.28kHz
1.81kHz
2.56kHz
3.62kHz
5.12kHz
7.24kHz
10.24kHz
14.48kHz
20.48kHz
Cheers,
Glen
One BC548, one BC558, 16 quad opamps (TLO74), 23 schotkey diodes, 143 capacitors and 300 resistors.
Bandpass Fc's are:
20Hz
28Hz
40Hz
57Hz
80Hz
113Hz
160Hz
226Hz
320Hz
453Hz
640Hz
905Hz
1.28kHz
1.81kHz
2.56kHz
3.62kHz
5.12kHz
7.24kHz
10.24kHz
14.48kHz
20.48kHz
Cheers,
Glen
Attachments
Glen,
At least one guy had similar ideas like you:
http://www.n2pk.com/
Too bad is doesn't extend down to audio freqs.
Jan Didden
At least one guy had similar ideas like you:
http://www.n2pk.com/
Too bad is doesn't extend down to audio freqs.
Jan Didden
Spent a couple of hours getting further with soldering bits into the prototype board for the digital logic video generator board this evening.
Spent most of my time beefing up the ground tracks on the underside of the board with heavy gauge wire and revising supply rail decoupling to get the 40 nS LM360 comparator working in its socket without oscillating.
The LM360 does the input voltage / exponential ramp comparison to generate the video signal for the amplitude level bars.
I was successful in the end, which really is amazing considering what a rats nest this prototype is. I almost finished the video blanking circuitry as well.
So far the design, which I completed entirely on paper, is working 100% as predicted
Here is a short 20-second video I made with my digital camera of the VGA display at this nearly completed stage (3 meg download):
http://users.picknowl.com.au/~glenk/VGAASA1.AVI
Note that I haven’t built the analogue band filter / peak detector board yet, so the display is so far still a little bit boring. None of the 21 amplitude bars will appear on the screen until their voltage inputs are stimulated.
For this demo I just wired in the potentiometer into one of the 21 inputs as I did before, to get one of the amplitude bars to appear on screen.
The yellow amplitude bar is dancing up and down in the video because I was twiddling the pot.
I still have to finish soldering in the video blanking logic for the graticule generator section. I currently have the vertical blanking operating on the horizontal graticule lines only. The vertical blanking for the vertical graticule lines still needs to be done, as does the horizontal blanking for both the horizontal and vertical graticule lines. When done, the graticule will consist of just the 21 by 10 square grid in the middle of the screen.
Cheers,
Glen
Spent most of my time beefing up the ground tracks on the underside of the board with heavy gauge wire and revising supply rail decoupling to get the 40 nS LM360 comparator working in its socket without oscillating.
The LM360 does the input voltage / exponential ramp comparison to generate the video signal for the amplitude level bars.
I was successful in the end, which really is amazing considering what a rats nest this prototype is. I almost finished the video blanking circuitry as well.
So far the design, which I completed entirely on paper, is working 100% as predicted
Here is a short 20-second video I made with my digital camera of the VGA display at this nearly completed stage (3 meg download):
http://users.picknowl.com.au/~glenk/VGAASA1.AVI
Note that I haven’t built the analogue band filter / peak detector board yet, so the display is so far still a little bit boring. None of the 21 amplitude bars will appear on the screen until their voltage inputs are stimulated.
For this demo I just wired in the potentiometer into one of the 21 inputs as I did before, to get one of the amplitude bars to appear on screen.
The yellow amplitude bar is dancing up and down in the video because I was twiddling the pot.
I still have to finish soldering in the video blanking logic for the graticule generator section. I currently have the vertical blanking operating on the horizontal graticule lines only. The vertical blanking for the vertical graticule lines still needs to be done, as does the horizontal blanking for both the horizontal and vertical graticule lines. When done, the graticule will consist of just the 21 by 10 square grid in the middle of the screen.
Cheers,
Glen
The Prologix adapter works just fine. I used the "configurator" from John Miles -- http://www.thegleam.com/ke5fx/gpib/readme.htm#prologix
and program using EZGPIB in Pascal (which my sons learned in high school !)
I put the 3577 instruction set on the following page:
http://www.tech-diy.com/TestEquipment/HP_Basic_Instr/HP_Basic.htm
maybe we should split this off since LittleSmith is doing so much work for us.
and program using EZGPIB in Pascal (which my sons learned in high school !)
I put the 3577 instruction set on the following page:
http://www.tech-diy.com/TestEquipment/HP_Basic_Instr/HP_Basic.htm
maybe we should split this off since LittleSmith is doing so much work for us.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.