Why would the DC voltage on my 2215A be exactly 1/2 what the voltage really is? I've verified it with my fluke volt meter and my Simson 260. Any Ideas? For example the setting says 1v/division so that means to me that a deflection of 4 small divisions would be a reading of 4 volts. But the other meters tell me it's 8 volts.
That is really odd. I would not think it possible to calibrate it that far off.
I see the 2215a has a 500mV p-p calibration signal. Does it read correctly?
One thing that confuses me is you say 4 SMALL divisions equal 4 volts. But it is 4 large divisions that equal 4 volts. Maybe it is just a difference in what we call large and small divisions. The screen has 8 large divisions vertically (10 in the horizontal direction). Each large division is divided into 5 smaller ones. So on the 1 volt setting, a large division is 1 volt, and each small division is 0.2 volts.
Terry
I see the 2215a has a 500mV p-p calibration signal. Does it read correctly?
One thing that confuses me is you say 4 SMALL divisions equal 4 volts. But it is 4 large divisions that equal 4 volts. Maybe it is just a difference in what we call large and small divisions. The screen has 8 large divisions vertically (10 in the horizontal direction). Each large division is divided into 5 smaller ones. So on the 1 volt setting, a large division is 1 volt, and each small division is 0.2 volts.
Terry
Well you cleared up one problem. I was reading tht 2215 wrong. I get it. Big squares would be one volt if the scale is set to 1v. The other problem cleared is that my fluke was stuck reading in 1/2 mode. It read 3 volts as 1.5. How it got that way I have no idea. I got it reset and things are making a lot more sense now. Thanks for the help. I've been going round and round.
I have an audio signal generator that has 600 ohm output load requirement and a spectrum analyzer that has a 50 ohm input. How do I do that? Use 2 bnc's? One with 600 and another 50 ohm? Looks like the 50 ohm side to be seen by the spectrum analyzer and on the other side a 600 ohm bnc seen by the signal generator. Am I understanding this right?
That is not right. The generator has a 600 Ohm source impedance. The spectrum analyzer is 50 Ohms. You will have a voltage divider with the 600 Ohms in series with the 50 Ohms of the spectrum analyzer. It would be around 1/12 of the source voltage at the spectrum analyzer.
50 Ohm spectrum analyzers are usually RF instruments and don't work down to audio frequencies. They are also easily damaged with too much signal or static discharge to the inputs.
50 Ohm spectrum analyzers are usually RF instruments and don't work down to audio frequencies. They are also easily damaged with too much signal or static discharge to the inputs.
In addition to what 1audio said, let me try to explain a bit more...
1.) The generator output impedance is probably not really 600 Ohms. Instead, it is rated to deliver the rated performance (voltage output, distortion, bandwidth) into a load of 600 Ohms. How it reacts to a smaller load resistance (like 50 Ohms) depends on the generator. Probably the output level will be reduced and the distortion will increase. But in fact, the output impedance is probably near zero Ohms if it is an audio generator. Here is how to tell: Set it up for a sine wave at some convenient level and frequency (like 1 V at 1kHz). Measure the output level with no load (like into an oscilloscope or dmm). Then load it with 600 ohms in parallel with the scope or meter input. If the generator output impedance is 600 Ohms, then the level will drop to half of what it was with no load. If it stays about the same, then the generator output impedance is actually near zero.
2.) Like 1audio says, be careful about the input to the spectrum analyzer. They are made for looking at small signals. Check the maximum input level spec for it and don't exceed it. Also, most spectrum analyzers cannot handle ANY DC voltage at the input without damage, so be careful.
3.) Like 1audio, I suspect that since the spectrum analyzer input is 50 Ohms, then it probably does not work at audio frequencies. It probably won’t work at frequencies below a few MHz.
4.) The ‘correct’ way to do what you want to do is to use some sort of matching network between the generator and the analyzer. But matching doesn’t matter much at audio frequencies unless the line lengths get really, really long. So an easy way to do what you want to do is to just use a 550 Ohm resistor in series with the generator output (or analyzer input). This means the generator would see a load of 550 Ohms plus the 50 Ohms to ground that the analyzer input will have. So it gets the 600 Ohm load it is rated to drive. The analyzer will not have the proper source impedance, but it won’t matter at audio frequencies. You just have to deal with the amplitude decrease caused by the voltage divider. It will reduce the voltage at the generator output by 50/600 or 0.083 times, which is -21.6dB (voltage) or -10.8 dB (power).
5.) If you want to get the generator load impedance and the analyzer source impedance both correct (just for the sake of being technically correct), then you need to build a divider with a 570 Ohm resistor in series with the generator output and a 53 Ohm resistor in parallel with the analyzer input. This will give you about 0.043 times or -27.3 dB (voltage) or -13.6 dB (power).
6.) Since it is inconvenient to work with amplitude changes like above, usually you would make a slightly more complicated attenuator/matching network that requires three resistors, called a pi attenuator. This page has an easy to use calculator: Matching Pi Attenuator Calculator
Using it, you can see that a 20 dB attenuator that matches 600 Ohms to 50 Ohms requires
Shunt in = 1573 Ohms
Shunt out = 62.7 Ohms
Series = 936 Ohms
The calculator lets you play around with ‘real’ values to see how close you get.
7.) The easiest way is just to plug the output of the generator directly into the spectrum analyzer. But be careful of the levels! Even the DC! And don’t be surprised if the generator doesn’t meet all of its specs when loaded with 50 Ohms. Expect the distortion to be a bit higher than spec and the output level to be lower.
8.) Impedance matching generally doesn’t matter at audio frequencies. The reason is the mismatch problems don’t usually show up until the length of the connection becomes some significant portion of the wavelength of the signal. Even at very high ‘audio’ frequencies, like 100kHz, the wavelength is about 3 kilometers. So unless you have very long cables, it won’t matter. If the cable becomes a significant portion of the wavelength (like can happen in MHz or above signals), then reflections can cause really strange amplitude distortions to the signal. This can make the signal amplitude higher or lower that the original signal, depending on the exact length of the cable.
9.) But you do have to deal with the fact that the amplitude will not read correctly on the analyzer. The analyzer probably reads in dBm. This is based on 1 milliWatt into 50 Ohms = 0 dBm. So let’s take an example:
If you set your generator to give 100 milliVolts peak to peak using your oscilloscope (you need to do this with a 50 Ohm resistor in parallel with the scope input if your scope doesn’t have a 50 Ohm input setting), then that is 35.4 mVrms. At 50 Ohms, this is 25 micro-Watts (0.000025 W) which is the same as 0.025 milliWatts. This is -16 dBm. So the analyzer should read about that level (assuming it can operate at such low frequencies).
Make sure you understand how to convert from voltage to power in dBm so you can be sure not to overload the spectrum analyzer and damage it.
Terry
1.) The generator output impedance is probably not really 600 Ohms. Instead, it is rated to deliver the rated performance (voltage output, distortion, bandwidth) into a load of 600 Ohms. How it reacts to a smaller load resistance (like 50 Ohms) depends on the generator. Probably the output level will be reduced and the distortion will increase. But in fact, the output impedance is probably near zero Ohms if it is an audio generator. Here is how to tell: Set it up for a sine wave at some convenient level and frequency (like 1 V at 1kHz). Measure the output level with no load (like into an oscilloscope or dmm). Then load it with 600 ohms in parallel with the scope or meter input. If the generator output impedance is 600 Ohms, then the level will drop to half of what it was with no load. If it stays about the same, then the generator output impedance is actually near zero.
2.) Like 1audio says, be careful about the input to the spectrum analyzer. They are made for looking at small signals. Check the maximum input level spec for it and don't exceed it. Also, most spectrum analyzers cannot handle ANY DC voltage at the input without damage, so be careful.
3.) Like 1audio, I suspect that since the spectrum analyzer input is 50 Ohms, then it probably does not work at audio frequencies. It probably won’t work at frequencies below a few MHz.
4.) The ‘correct’ way to do what you want to do is to use some sort of matching network between the generator and the analyzer. But matching doesn’t matter much at audio frequencies unless the line lengths get really, really long. So an easy way to do what you want to do is to just use a 550 Ohm resistor in series with the generator output (or analyzer input). This means the generator would see a load of 550 Ohms plus the 50 Ohms to ground that the analyzer input will have. So it gets the 600 Ohm load it is rated to drive. The analyzer will not have the proper source impedance, but it won’t matter at audio frequencies. You just have to deal with the amplitude decrease caused by the voltage divider. It will reduce the voltage at the generator output by 50/600 or 0.083 times, which is -21.6dB (voltage) or -10.8 dB (power).
5.) If you want to get the generator load impedance and the analyzer source impedance both correct (just for the sake of being technically correct), then you need to build a divider with a 570 Ohm resistor in series with the generator output and a 53 Ohm resistor in parallel with the analyzer input. This will give you about 0.043 times or -27.3 dB (voltage) or -13.6 dB (power).
6.) Since it is inconvenient to work with amplitude changes like above, usually you would make a slightly more complicated attenuator/matching network that requires three resistors, called a pi attenuator. This page has an easy to use calculator: Matching Pi Attenuator Calculator
Using it, you can see that a 20 dB attenuator that matches 600 Ohms to 50 Ohms requires
Shunt in = 1573 Ohms
Shunt out = 62.7 Ohms
Series = 936 Ohms
The calculator lets you play around with ‘real’ values to see how close you get.
7.) The easiest way is just to plug the output of the generator directly into the spectrum analyzer. But be careful of the levels! Even the DC! And don’t be surprised if the generator doesn’t meet all of its specs when loaded with 50 Ohms. Expect the distortion to be a bit higher than spec and the output level to be lower.
8.) Impedance matching generally doesn’t matter at audio frequencies. The reason is the mismatch problems don’t usually show up until the length of the connection becomes some significant portion of the wavelength of the signal. Even at very high ‘audio’ frequencies, like 100kHz, the wavelength is about 3 kilometers. So unless you have very long cables, it won’t matter. If the cable becomes a significant portion of the wavelength (like can happen in MHz or above signals), then reflections can cause really strange amplitude distortions to the signal. This can make the signal amplitude higher or lower that the original signal, depending on the exact length of the cable.
9.) But you do have to deal with the fact that the amplitude will not read correctly on the analyzer. The analyzer probably reads in dBm. This is based on 1 milliWatt into 50 Ohms = 0 dBm. So let’s take an example:
If you set your generator to give 100 milliVolts peak to peak using your oscilloscope (you need to do this with a 50 Ohm resistor in parallel with the scope input if your scope doesn’t have a 50 Ohm input setting), then that is 35.4 mVrms. At 50 Ohms, this is 25 micro-Watts (0.000025 W) which is the same as 0.025 milliWatts. This is -16 dBm. So the analyzer should read about that level (assuming it can operate at such low frequencies).
Make sure you understand how to convert from voltage to power in dBm so you can be sure not to overload the spectrum analyzer and damage it.
Terry
The generator output of the 339A really is 600 ohms from full output to the minimum.
Put a 600 ohm load on it and the level will drop to half. The manual for the 339A is available on the net from a few sources.The calibration procedures are in the manual. You don't have to use the same equipment named in the manual but look the equipment up to find out what is and what output/input impedance is specified. Be sure to use a standard or recently calibrated test fixture.
Put a 600 ohm load on it and the level will drop to half. The manual for the 339A is available on the net from a few sources.The calibration procedures are in the manual. You don't have to use the same equipment named in the manual but look the equipment up to find out what is and what output/input impedance is specified. Be sure to use a standard or recently calibrated test fixture.
Yes, the 339 output really is 600 Ohms and I should have mentioned that since this is a 339 thread. I thought Gwalters was asking about some other (non-specified) generator). It may also be a true 600 ohm output impedance, but I have seen some that are not, that's why I mentioned it. I don't know how common it is.
Yetty
Yetty
Making Cables
It might be helpful to other's as well. I know it would help me out too as I'll need some too. Yeah I have to buy the ends...but if I can make some that would help me out as then I'd just have to purchase the ends. I've got a couple of crimpers that I could use also.
@TerrySt or Others Terry do you know what font HP used? I have the same issues as other people as we all do with this vintage HP and polystryene knobs.
Heck, I just bought a different peice of gear just for the knobs they lookded great and like new. Imagine my surprise when I turned it and it split and shattered even before I removed it to put on my 339a.
Some of my clear pieces have disappearing number and are cracking too.
I need to do the same thing you are Terry. And I need to make them for my older eyes too with better contrast between clear and number etc.
I have an orginal manuel for mine as I hate being stuck with the computer. One thing I found handy was copying out the schematics on heavy card stock. I have a complete set of schematics on heavy 11 x 17 cards to make notes mark up changes etc. I'll post a pic of what I've done. I don't like working from the computer or small 8.5 x 11 inch pages having 30 pieces of paper or the computer sitting there.
That may change, but I like the card stock copies.
Cheers,
@dennismiller55,
It might be helpful to other's as well. I know it would help me out too as I'll need some too. Yeah I have to buy the ends...but if I can make some that would help me out as then I'd just have to purchase the ends. I've got a couple of crimpers that I could use also.
@TerrySt or Others Terry do you know what font HP used? I have the same issues as other people as we all do with this vintage HP and polystryene knobs.
Heck, I just bought a different peice of gear just for the knobs they lookded great and like new. Imagine my surprise when I turned it and it split and shattered even before I removed it to put on my 339a.
Some of my clear pieces have disappearing number and are cracking too.
I need to do the same thing you are Terry. And I need to make them for my older eyes too with better contrast between clear and number etc.
I have an orginal manuel for mine as I hate being stuck with the computer. One thing I found handy was copying out the schematics on heavy card stock. I have a complete set of schematics on heavy 11 x 17 cards to make notes mark up changes etc. I'll post a pic of what I've done. I don't like working from the computer or small 8.5 x 11 inch pages having 30 pieces of paper or the computer sitting there.
That may change, but I like the card stock copies.
Cheers,
Last edited:
Sorry, but I don't know what font HP used. As you can see from my pictures, I was not able to match the font very accurately. I played around for quite a while with different fonts, sizes, bold or not. And then of course playing around to get the right diameter and spacing to place the characters on the right part of the dial. What I ended up with is far from perfect, but it is functional.
I didn't have a good program to do the work on. I used an old version of Microsoft Word for Windows. It let me wrap the letters around a circle, but it was very hit or miss as far as how big the circle was and where the letters ended up. I'm sure there must be better programs to use for this. Something like Corel Draw, PhotoShop or maybe even AutoCAD or similar. But Word was all I had.
I probably have the file saved from the dial I made if that helps.
Terry
I didn't have a good program to do the work on. I used an old version of Microsoft Word for Windows. It let me wrap the letters around a circle, but it was very hit or miss as far as how big the circle was and where the letters ended up. I'm sure there must be better programs to use for this. Something like Corel Draw, PhotoShop or maybe even AutoCAD or similar. But Word was all I had.
I probably have the file saved from the dial I made if that helps.
Terry
Terry,
Yes, please, link or attach file. I've found it is better to start somewhere than no where. I have two old versions of word. I also have an older drawing problem. PhotoShop is nice but if you don't use it all the time you forget its particulars.
Other DIYer would benefit from it too. There is a hobby lobby place not too far from me, I can get the decal sheets there I think. I have a Brother injet style printer. I'll figure out what's best to use.
Cheers,
Yes, please, link or attach file. I've found it is better to start somewhere than no where. I have two old versions of word. I also have an older drawing problem. PhotoShop is nice but if you don't use it all the time you forget its particulars.
Other DIYer would benefit from it too. There is a hobby lobby place not too far from me, I can get the decal sheets there I think. I have a Brother injet style printer. I'll figure out what's best to use.
Cheers,
Here is the file. In order to use the decals, you will have to 'erase' the old lettering. This involves sanding away the old lettering which is actually slightly embossed into the back of the clear disc. Then you have to use progressively finer sandpaper and polishing compounds until you get the disk back to clear again. A bit of work, but in the case of one of my knobs, the disk was so messed up that I had no choice in the matter. Another just had the lettering rubbed off in places, so I touched up the lettering with black paint applied very carefully with a toothpick. Then the edges were sharpened up a bit with a sharp exacto blade. A magnifier helps for this work.
Terry
Terry
Terry,
I understand. I guess anyone with some of that era HP gear does. Besides bursting knobs I've go scraped and screwd up clear skirts too (if that is what they are called.)
And they are weird too, they aren't just straight/flat they snap into the knob.
I"ll have to gather up stuff and see what I can come up with. I and others thank you. Let me see how it goes on this end and I'll keep posted from time to time when I get a chance.
Cheers,
Terry,
I think you probably had an easier time doing that than I am with the "new" version of word.
Excellent job by the way. It looks simple enough until you try to type or change something.
I'll try working with it when I have some quiet time. : )
I've got some good old software around here, but with Microsoft Win 7, Win 8 much wont run. I can only guess because Microsoft want people to buy or pay for a product license for it or something.
Next version Microsoft is going to be charging people $80 a year for it.
I understand. I guess anyone with some of that era HP gear does. Besides bursting knobs I've go scraped and screwd up clear skirts too (if that is what they are called.)
And they are weird too, they aren't just straight/flat they snap into the knob.
I"ll have to gather up stuff and see what I can come up with. I and others thank you. Let me see how it goes on this end and I'll keep posted from time to time when I get a chance.
Cheers,
Terry,
I think you probably had an easier time doing that than I am with the "new" version of word.
Excellent job by the way. It looks simple enough until you try to type or change something.
I'll try working with it when I have some quiet time. : )
I've got some good old software around here, but with Microsoft Win 7, Win 8 much wont run. I can only guess because Microsoft want people to buy or pay for a product license for it or something.
Next version Microsoft is going to be charging people $80 a year for it.
You guys are doing it the hard way... forget Word and try Front Panel Express which is designed to do such job.
Thanks Jazbo8,
I looked through the stuff they had and not sure how to make the skirts for the knobs?
It looks like and okay tool for designing panels though. Just wondering about costs. $40 - $100+ USD for some of the demo panels looks kind of pricy. Like the littel 2 inch by 2 inch panels. Unless they are huge in size.
I'll see if there is a smiple way to do knob skirts.
I looked through the stuff they had and not sure how to make the skirts for the knobs?
It looks like and okay tool for designing panels though. Just wondering about costs. $40 - $100+ USD for some of the demo panels looks kind of pricy. Like the littel 2 inch by 2 inch panels. Unless they are huge in size.
I'll see if there is a smiple way to do knob skirts.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.