Hi Conrad,
Yep, we have the same problems in that case. That's why I might investigate the control connector. You will find that those "sealed" single turn trimmers are also noisy. I will have to replace the entire lot. I had calibrated mine using a 3458A, but I no longer have access to anything like that. Not at reasonable prices anyway.
Does anyone have the manual for this thing? I could really use one to tackle some of these problems.
I'll try WWVB as well. That was my backup plan - sort of. Should be easy to convert the amp. The antenna is another story. It keeps getting longer and longer as the frequency drops.
-Chris
Yep, we have the same problems in that case. That's why I might investigate the control connector. You will find that those "sealed" single turn trimmers are also noisy. I will have to replace the entire lot. I had calibrated mine using a 3458A, but I no longer have access to anything like that. Not at reasonable prices anyway.
Does anyone have the manual for this thing? I could really use one to tackle some of these problems.
I'd be there if I lived in the 'States close enough. Someone should grab that! We don't get things like that in Canada very often (like never!).
I'll try WWVB as well. That was my backup plan - sort of. Should be easy to convert the amp. The antenna is another story. It keeps getting longer and longer as the frequency drops.
-Chris
I picked up a Valhalla 2701 for DC some years ago. I would describe it as the first true 24 bit delta sigma dac. The switches control a counter string and it divides the voltage by 1,000,000 counts. Its really pretty stable and accurate. In fact all of my primary precision DC sources and meters depend on the LTC1000Z Zener reference. I have a Fluke 8506A, a Fluke 732A, the Valhalla and a Prema 6001, all depending on the same reference and the same chopper amp. They all seem to be very stable (or at least they all drift together). And they are at least as good as a standard cell.
I also have the other form of 24 bit DAC, both an ESI and a JRL 7 digit KV divider. With the used market fuelled by eBay its not hard to build a museum of this stuff. The KV dividers are still about as good as it gets without cryogenics.
I also have the other form of 24 bit DAC, both an ESI and a JRL 7 digit KV divider. With the used market fuelled by eBay its not hard to build a museum of this stuff. The KV dividers are still about as good as it gets without cryogenics.
Those will have to wait for now, but thanks for the info.
I have found that the HP Z3801A will provide a 10 MHz reference from GPS signals. I'll have to save up my pennies as they are $250 USD, lowest price. Many people seem to think they are $500 or so. The thing that bugs me is that many sellers on Eeekkbay don't even understand what these things do.
They snag them for a profit only. Oh well.
-Chris
I have found that the HP Z3801A will provide a 10 MHz reference from GPS signals. I'll have to save up my pennies as they are $250 USD, lowest price. Many people seem to think they are $500 or so. The thing that bugs me is that many sellers on Eeekkbay don't even understand what these things do.
They snag them for a profit only. Oh well. -Chris
The Thunderbolt (recent versions) are as good as or better than the HP and can be had for $160 TBolt I have one and an Arbiter 1083A and they are essentially identical (except for the 5MHz and 1 MHz outputs). The Lady Heather software gives a lot of info and performance data and great charts. And they are small.
Hey guys,
I just replaced all PS electrolytic caps, keeping the same part ratings. No change here, even thw PS noise numbers are still the same.
I noticed something that Im not sure if it was there before (I dont remember doing this exact test). Setting the oscillator level to .1V and 1kHz gives me a very high distortion number. If I turn on the 30kHz LP filter the distortion decreases significantly. Turning on just the 80kHz filter also makes the distoprtion drecrease, but not as much.
What I sould I do now? It looks like there is some HF noise somewhere here.
I just replaced all PS electrolytic caps, keeping the same part ratings. No change here, even thw PS noise numbers are still the same.
I noticed something that Im not sure if it was there before (I dont remember doing this exact test). Setting the oscillator level to .1V and 1kHz gives me a very high distortion number. If I turn on the 30kHz LP filter the distortion decreases significantly. Turning on just the 80kHz filter also makes the distoprtion drecrease, but not as much.
What I sould I do now? It looks like there is some HF noise somewhere here.
Hi Demian,
Thank you very, very much for that info! I sure as heck do not have the cash to squander, so I'm very grateful!
I see that's everything you need to get going too! Now that is really a good deal, there is no room to mess up. I really owe you one there Demian!
I also see that there is maybe a group buy coming up on the time nuts group. I will ask them what they think of this deal. I just wish the exchange was better to Canuck bucks.
Hi bsgd,
Can you try taking the readings on the same ranges without an input? From another input?
I'm trying to see if your problems are in the THD section, or the oscillator section. From what you posted, there is certainly something going on at high frequencies.
-Chris
Edit: I see they have a kit with an improved antenna. Is going this way a good idea?
Thank you very, very much for that info! I sure as heck do not have the cash to squander, so I'm very grateful!
I see that's everything you need to get going too! Now that is really a good deal, there is no room to mess up. I really owe you one there Demian!
I also see that there is maybe a group buy coming up on the time nuts group. I will ask them what they think of this deal. I just wish the exchange was better to Canuck bucks.
Hi bsgd,
Can you try taking the readings on the same ranges without an input? From another input?
I'm trying to see if your problems are in the THD section, or the oscillator section. From what you posted, there is certainly something going on at high frequencies.
-Chris
Edit: I see they have a kit with an improved antenna. Is going this way a good idea?
Unfortunately I cant test with another oscillator, as the only one I have is the one built into the HP.
Here' are some interesting observations:
1- Internal oscillator directly connected to the analyzer input. INPUT RANGE set to .1V and OSCILLATOR LEVEL set to .1V (1kHz freq). ALL filters OFF. Measured distortion is .0215%.
2- Internal oscillator directly connected to the analyzer input. INPUT RANGE set to .1V and OSCILLATOR LEVEL set to .1V (1kHz freq). LP 30kHz filter ON. Measured distortion is .0044%.
3- Analyzer input shorted. INPUT RANGE set to .1V and frequency set to 1kHz. All filters off. Measured distortion is .006%.
4- Analyzer input shorted. INPUT RANGE set to .1V and frequency set to 1kHz. LP 30 kHz filter ON. Measured distortion is .0015%.
There is also something 'weird' going on: When the oscillator level is set to 30mV or below (and its directly connected to the analyzer input), the LEDs that indicate where to turn the INPUT RANGE control to select the correct one seems to misbehave. The LED indicating I should turn the range to a lower value stays lit all the time (it only goes out when I reach the lowest scale: .1mV).
At 100mV or more it behaves diferently: When I reach the proper scale, both LEDs go off, but if I keep down-scaling they dont light up. For example: If I have a 100mV signal being fed to the analyzer, and INPUT RANGE is set to 3V, the LED indicating I should lower the scale lights up. When I reach the 100mV scale it goes out, but if I keep turning the scale down to 30mV or less, both LEDs stay off and dont inidicate I should turn to a higher scale.
Here' are some interesting observations:
1- Internal oscillator directly connected to the analyzer input. INPUT RANGE set to .1V and OSCILLATOR LEVEL set to .1V (1kHz freq). ALL filters OFF. Measured distortion is .0215%.
2- Internal oscillator directly connected to the analyzer input. INPUT RANGE set to .1V and OSCILLATOR LEVEL set to .1V (1kHz freq). LP 30kHz filter ON. Measured distortion is .0044%.
3- Analyzer input shorted. INPUT RANGE set to .1V and frequency set to 1kHz. All filters off. Measured distortion is .006%.
4- Analyzer input shorted. INPUT RANGE set to .1V and frequency set to 1kHz. LP 30 kHz filter ON. Measured distortion is .0015%.
There is also something 'weird' going on: When the oscillator level is set to 30mV or below (and its directly connected to the analyzer input), the LEDs that indicate where to turn the INPUT RANGE control to select the correct one seems to misbehave. The LED indicating I should turn the range to a lower value stays lit all the time (it only goes out when I reach the lowest scale: .1mV).
At 100mV or more it behaves diferently: When I reach the proper scale, both LEDs go off, but if I keep down-scaling they dont light up. For example: If I have a 100mV signal being fed to the analyzer, and INPUT RANGE is set to 3V, the LED indicating I should lower the scale lights up. When I reach the 100mV scale it goes out, but if I keep turning the scale down to 30mV or less, both LEDs stay off and dont inidicate I should turn to a higher scale.
bsgd, I'm sorry I don't have the same analyzer or I'd set it up and try to match readings. FWIW, my other analyzers, as expected, only produce their specified readings near their maximum signal levels. At low signal levels noise dominates and the readings will be high. As the lopass filters are switches in, the readings will be reduced. Check the manual for the correct levels.
anatech, I think I can scare up a manual for the 8200- give me a few days. The antenna for WWVB is generally a medium sized ferrite rod with a coil of wire around it. Get it high up and run a coax to several stages of tuned amplifiers. It used to be easy to lock to the 60kHz signal, but they changed the modulation such that it drops by a huge amount for part of the time. I haven't tried this, but it seems like a locked oscillator would now need a track and hold circuit, such that when the signal dropped there was no tendency to drift rapidly up or down.
1audio, yes, I pretty much have a museum of old standards and dividers. I bought serial number 1 of my divider direct from JRL, and have another one, plus a couple ESI units. I maintain the volt with three Fluke standards that I can inter-compare, plus the 8200. My Fluke standards are of different vintages and I hand adjusted one of them for near zero TC (the master that I take out for calibration). My logic is that ones voltage standards shouldn't be identical, lest they tend to drift the same way. Also have the Fluke resistance standard, plus a bunch of L&N cans. IMO, there's less and less that needs anywhere near this level of absolute accuracy.
anatech, I think I can scare up a manual for the 8200- give me a few days. The antenna for WWVB is generally a medium sized ferrite rod with a coil of wire around it. Get it high up and run a coax to several stages of tuned amplifiers. It used to be easy to lock to the 60kHz signal, but they changed the modulation such that it drops by a huge amount for part of the time. I haven't tried this, but it seems like a locked oscillator would now need a track and hold circuit, such that when the signal dropped there was no tendency to drift rapidly up or down.
1audio, yes, I pretty much have a museum of old standards and dividers. I bought serial number 1 of my divider direct from JRL, and have another one, plus a couple ESI units. I maintain the volt with three Fluke standards that I can inter-compare, plus the 8200. My Fluke standards are of different vintages and I hand adjusted one of them for near zero TC (the master that I take out for calibration). My logic is that ones voltage standards shouldn't be identical, lest they tend to drift the same way. Also have the Fluke resistance standard, plus a bunch of L&N cans. IMO, there's less and less that needs anywhere near this level of absolute accuracy.
I have a question to you test equipment buffs:
I was offered an HP 3721A Correlator:
http://www.helmut-singer.de/pdf/hp3720a-3721a.pdf
What's that good for??
Jan Didden
I was offered an HP 3721A Correlator:
http://www.helmut-singer.de/pdf/hp3720a-3721a.pdf
What's that good for??
Jan Didden
janneman said:
In short, a boat anchor!
An early (all hardware!?) FFT converter of time to frequency domain (and vice-versa?)
Output to punched paper tape sets the vintage!
Can be used to de-noise (ie recover) a repetitiive signal as well as other functions.
Stuff we take for granted now on a cheapo PC with free software, or any real time embedded FFT engine.
Cliff
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.