It's not that we don't want to sell you boards. It just that our designs are still in the making. I'm not really speaking for everyone else.
We can't really take Bob's design and make boards for it and sell them.
Although if you look around the forum you will find that some have offered boards for there interpretations of Bob's oscillator. Maybe they still have some boards kicking around.
This is a DIY forum. A place to exchange knowledge, ideas and engage in technical discussion. It's really about doing it for yourself.
If you want help, you have been offered plenty. If you want boards you'll just have to wait until someone offers them.
You seem to reject everything offered to you so why would we trust you would buy our design if it were offered.
What we put up here is up for grabs by anyone. The next thing we know is someone slapping a patent on or taking credit for something novel we've done.
It has happened.
Last edited:
I know Victors modules are a bit unconventional but mounting them really is intended to be a few round holes (rca jacks, level control and mounting screw (ground). I have made several suggestions for the 35V issue and at least one guy has successfully mounted a module is a cookie tin (good excuse to buy some) and is running it off of 4 9V batteries.
I think you are looking more for a kit like a Velleman where most of engineering is done and its kitted and just waiting to go into a chassis. This may happen here but it could take a while.
I'm here for the intellectual exercise. I see the interesting and big challenge is an ultra low distortion oscillator that doesn't need exotic stuff to calibrate or optimize. Samuel's targets are quite reasonable except I would ratchet up the flatness spec. Its really nice to just know that the error is essentially zero.
I think the targets can be met and without adjustments if the AGC element can be made to be really transparent. The State Variable has the best chance because the distortion from the AGC is low pass filtered twice. I think the Intersil sample and hold is a good option even at $8.00 since its still current production and available. The opto AGC can work but its time constants are long and it has a high voltage co-efficient making for distortion proportional to signal.
Tuning is a big challenge. It requires either a lot of parts or precision variable parts (pricing a dual gang precision pot can be shocking). Switches and relays are also expensive. This is why reclaiming a chassis with suitable components can be very attractive. The Heathkit would not be my first choice but its not bad. A board that could drop in and work in one would be pretty slick.
Failing that option a lot of decisions about what is needed are required. Continuous tuning is a big issue to address. A set of cardinal frequencies could make a lot of sense. Match those to a twin t filter with some signal conditioning and it could be very useful front end to a soundcard + fft.
Dirk, where do you get off speaking to Dick like this? He is suposed to supply you with what you need because you are to lazy to learn or unable to learn about this topic? Dick has shared considerable knowledge here and on his website.
Go buy a used device if you can't or won't do the work yourself, but, don't blame Dick, or anyone else.
Ken
Because you continue to tremendously underestimate the difficulty of such a project. Anyone with basic experience and knowledge can put together a few transistors and call it The Subjectively Ultimate Power Amp. However getting reliable distortion figures at or even below the -120 dB level is an entirely different world--the notch filter is far less forgiving than the hum ear. I've spent now several years researching the design of high-performance oscillators and invested money for which you could by an entire pile of working HP 8903Bs (and I'm not doing this for profit!). Still I don't have a design at hand which I can offer as fully tested and well documented project. I'll share it once I'm there, but your attitude makes me feel like my efforts are not appreciated and I shouldn't.
Samuel
Last edited:
@Samuel Groner -- My build of Bob Cordell's oscillator runs around 0.6ppm THD at 1kHz at 6.3VRMS output, when tweaked for best overall operation across all ranges. I was able to optimize at 1kHz with a bit less than 0.3ppm THD. That's pretty good, but not ground-breaking. It would be interesting to try out some of the SG-505's tweaks on the build; but now I'm using different amps in it, not 5534s.
I would love to look at an SG-505 -- When they were affordable, i was doing other things -- now I'd like to have one but don't want to spend the ridiculous amount of money they bring, even in poor condition.
I would love to look at an SG-505 -- When they were affordable, i was doing other things -- now I'd like to have one but don't want to spend the ridiculous amount of money they bring, even in poor condition.
Just for the records the link:
http://assets.fluke.com/manuals/510A_AD_imeng0000.pdf
Also very interesting is the Fluke 5200A:
http://assets.fluke.com/manuals/5200A___imeng0300.pdf
Samuel
That's true. One has to be patient and wait until a good offer pops up. I was lucky to shoot mine in a TM503 rack together with an AM503+probe and a 1U TM500 prototyping/DIY kit (very rare) for less than EUR 500, supposedly from the estate of someone deceased and all in excellent condition. Not exactly cheap but well worth in the end. Applies to TM500 stuff in general, no way I could easily design and build a better (let alone cheaper) diff-amp than the AM502, for example...
Considering that this is the first time you have shared this on this thread, then your offense is misplaced. If and when you offer some kind of kit for sale, I'll be one of the first to buy it. I had no idea you were considering this, so thank you for your efforts.
You misunderstand. I was not blaming anyone. I was saying that it is a shame that no one appeared to be taking the lead on this project in spite of the fact that there are many highly skilled designers on this thread. There are many other folks on this message board who are great leaders and provide designs and/or boards that are ready to go. Nelson Pass is one of them, for example. Why not some good DIY test gear also? That's all I'm saying.
Dick has designed an apparently outstanding notch filter with a lot of documentation and it's great he's done that but there are no boards that I know of that I can buy and assemble. Having a great design is one thing, making it into a real piece of gear is another. Dick has been extremely generous with his web site and his contributions here, but I'm not about to try to make my own board from his designs because I'm not confident that anything I would produce would work well. I've only done basic things in Pad2Pad for example, and I make mistakes often.
Well, you don't have to worry about me stealing any ideas because I can't get a patent. I work there, but I understand your concern. The simple solution is to file a patent application as soon as you think you have something unique, even if the application is incomplete.
I know people have offered help here and I am grateful for that. No one has stated that boards are forthcoming, except in a very vague way, so how would I know that something is happening? Why do you assume I reject other people's ideas? I'm in the process of making a board for Coluke's HP 239A clone, but I'm missing some information, for example. I bought Victor's oscillators on the recommendations from you guys, and I did it without thinking twice about it because I trusted you, but when I did think twice about it I realized how hard it is to use what he has produced.
If this forum was purely about doing it ourselves then there wouldn't be so many boards and kits for sale here, and the site itself sells Nelson Pass amplifiers in it's own store! So no, this isn't even "mostly" about DIY. There's a whole vendor's forum here with countless items for sale.
Anyone can ask Bob for permission to take his design and make a board for sale if the want to. It's been 30 years since it was designed. I would think he wouldn't have a problem with it since he doesn't offer boards for it either.
So, I hope this clears things up for you.
Well, as I said, I bought Victor's oscillator without really thinking it through. I know that I could get a faceplate made that would look super but for some reason I resist the idea. Drilling at intervals of 0.5mm is beyond my ability, since I can't even see that close. I know I can take a file and fiddle stuff, but then it looks ugly. When you put four 9V rechargeable batteries in series, you don't get 36V. It's not even close. It's outside of Victor's spec on input voltage. I did it, and I got 32 point something volts, which is apparently too low for his oscillator. I bought super hi zoot ($$) 9V Lithium ion batteries too.
Thanks for sharing all of the issues with making a good oscillator. I don't understand a lot of them, but I would add balanced outputs to the list of desired features. I know I can make a separate board myself that will do that so it's no big deal to me.
I'd be happy with a board and BOM, as I've said. It doesn't have to be a complete kit like what Velleman offers.
But, I think everyone's right. I am too stupid and too inexperienced to participate in this thread, so I'll just say goodbye and good luck for now. Sorry if anyone was offended by my remarks.
Last edited:
Actually Bobs sold quite a lot of boards for it after writing the article years ago.
I think one aspect of this thread that is causing friction are the different unstated goals of various participants. Rather than trying to figure out those goals we should describe the target use of the device, REMEMBER ITS JUST A TOOL!. While there is intellectual satisfaction in going further past the edge of the art if you have a wrench adequate to tightening or loosening a nut a better wrench won't do a better job of the task.
Ultra low distortion oscillators are relevant to measuring distortion. Typically these devices are less well suited to general purpose oscillator functions, like finding the resonance of a driver and because of frequency limits, the overall response of a circuit. For me the utility today of one of these is evaluating ADC circuits and to a lesser degree analog amplifiers. I have not encountered (even in ultra high end) amps and preamps that are within 30 dB of these distortion goals. Not even close. So my Boonton with its -100 dB THD+Noise is adequate to those tasks. For measurements of really low distortion chains you don't need to have continuous tuning (I can't think of a mechanism in a linear system that would have the distortion change a lot with small frequency changes). This suggests to me that a few cardinal frequencies with a matched twin t filter would be really useful. Reducing the complexity makes the device much more accessible. Making the source floating would make it useful for balanced or unbalanced applications.
That's my use case for these devices. Who has others? more input would help narrow the actual requirements, which in turn can lead to something that really can be made.
Ultra low distortion oscillators are relevant to measuring distortion. Typically these devices are less well suited to general purpose oscillator functions, like finding the resonance of a driver and because of frequency limits, the overall response of a circuit. For me the utility today of one of these is evaluating ADC circuits and to a lesser degree analog amplifiers. I have not encountered (even in ultra high end) amps and preamps that are within 30 dB of these distortion goals. Not even close. So my Boonton with its -100 dB THD+Noise is adequate to those tasks. For measurements of really low distortion chains you don't need to have continuous tuning (I can't think of a mechanism in a linear system that would have the distortion change a lot with small frequency changes). This suggests to me that a few cardinal frequencies with a matched twin t filter would be really useful. Reducing the complexity makes the device much more accessible. Making the source floating would make it useful for balanced or unbalanced applications.
That's my use case for these devices. Who has others? more input would help narrow the actual requirements, which in turn can lead to something that really can be made.
I think your perspective is useful even if sometimes inelegantly expressed. Also some of us presume a level of understanding that eludes many and makes it difficult to catch up if you aren't already there. (I barely understand PID stabilization and would not presume to explain details so you are not alone in bits of this going over your head.)
As Samuel said making ultra low distortion oscillators is really challenging. Its all in the details. I hope we will have a basic PCB that provides an oscillator and provisions for tuning at some point. In threads like this that can take months.
I hope Dennis posts on his experience with the Victor oscillator soon.
I'd like to note that it is not that difficult to make a point-to-point wiring version of Bob Cordell's PCB layout on perf board, and Steve Lafferty Modifications: Heathkit IG-18 Audio Generator did it nicely the same way for his HP 339A clone -- it is not magic. Steve's layout would fit nicely into many different enclosure types, and switching is just not that complicated.
That said, getting something that looks like a Tek SG-505 is no small task. I'm content with tuna cans and cookie tins, and knobs scavenged from old gear or from Radio Shack. But I'm a geezer who started breadboarding when breadboards were actually breadboards -- I used bare thumbtacks stuck in the board to solder parts to. Sometimes, if your standards are high, you just have to buy instrument-grade stuff.
That said, getting something that looks like a Tek SG-505 is no small task. I'm content with tuna cans and cookie tins, and knobs scavenged from old gear or from Radio Shack. But I'm a geezer who started breadboarding when breadboards were actually breadboards -- I used bare thumbtacks stuck in the board to solder parts to. Sometimes, if your standards are high, you just have to buy instrument-grade stuff.
Hello Rick,
Thanks for your kind words. I didn't think that the zip file had uploaded correctly. KiCad is worth investigating because it is open source and it works though it is a little different to other CAD systems which seem to annoy some people. I think that the file I uploaded was for version 0.1 of the PCB. I had to make some modifications to the schematic to make it work with some crude alterations to the PCB so beware. I also made a case and have a Google SketchUp drawing of that. I will try to upload the modified version of the project.
Personally I find these comments ridiculous. It may look ugly if you file the faceplate, but who cares as long as it is functional? This is a tool, not art design. And for the supply issue, Victor kindly published the schematics and it can be seen it runs off of two 15V shunt regs, so you can easily modify this 35V thing to be fed from any split or single rail power supply, say a simple PSU with two 18V regs. Or use one more 9V block and slightly increase R32&R33 (if even necessary at all)...
EDIT: Or try decrease shunt regs to +-13V and use your 4 9V blocks, I don't think it will give any penalty in performance and the range of R16 should still sufficient.
Last edited:
Victor's oscillator is the bomb, nuff said. Ridiculously low distortion, and I managed to easily put it in a Hammond enclosure, looks nice in my opinion.
As others have noted, if you want <0.00003% distortion, continuously variable freq from 10Hz to 100 kHz shows you don't really understand what you are asking for, nor are you going to properly apply it. For making ultra-precision measurements with ultra-accurate notch filters, there is only a need for a few (less than 4) fixed frequencies. When you NEED continuously variable, you shouldn't need ultra-low distortion.
Dirk, I don't think you have been fair to the contributors on this site. No one needs to dedicate their time to your pet project; few people need what you are asking for; they understand the applications don't require continuous adjustability.
Thank you, Victor, for the most impressive yet practical oscillator I have ever come across, for such a low price! I may come back for another frequency some day.
As others have noted, if you want <0.00003% distortion, continuously variable freq from 10Hz to 100 kHz shows you don't really understand what you are asking for, nor are you going to properly apply it. For making ultra-precision measurements with ultra-accurate notch filters, there is only a need for a few (less than 4) fixed frequencies. When you NEED continuously variable, you shouldn't need ultra-low distortion.
Dirk, I don't think you have been fair to the contributors on this site. No one needs to dedicate their time to your pet project; few people need what you are asking for; they understand the applications don't require continuous adjustability.
Thank you, Victor, for the most impressive yet practical oscillator I have ever come across, for such a low price! I may come back for another frequency some day.