Replacement for a leveled sine wave generator.

[Zenwizard]

I have been working on doing some calibrations on older analog Tek scopes as I have been building up the lab. I would like to do a full calibration on a Tektronix 7904A which is a 500 Mhz main frame. In preperation for this project I would getting my SG504 up to spec and had a catastrophic failure of some of the mechanical components in the unit. 60 year old plastics. unfortunately this is a custom part and as I do not have a lathe to fabricate another one at the moment I am stuck with out an appropriate signal source. I would like to get a signal generator used or new that can do sine waves accurately between .5 and 5.5 Vpk-pk and cover a frequency range from 9Khz to > =1050 MHz if possible. I don't mind new equipment or old however I would like it to have a 10 Mhz ref in if at all possible. Space is not an issue so it can be on the large size if needed. I have an HP 8657A but it will not go high enough in voltage for my current needs.

Zen

[David Hess]

Despite its simplicity, it is a pretty specialized instrument so there are not a lot of alternatives.

What I have done is use an unleveled RF generator, and manually level it using a sampling oscilloscope or sampling RF voltmeter.

[bob91343]

You can add a power amplifier at reasonable cost.  They are available in many styles.  For a while I was shopping for one but decided I didn't need that much amplitude.

[edavid]

You can add a power amplifier at reasonable cost.  They are available in many styles.  For a while I was shopping for one but decided I didn't need that much amplitude.

That doesn't help, since the amp is outside the leveling loop.

[bob91343]

Yes but you can run a calibration.  Its gain vs freq curve isn't likely to change.

[srb1954]

I have been working on doing some calibrations on older analog Tek scopes as I have been building up the lab. I would like to do a full calibration on a Tektronix 7904A which is a 500 Mhz main frame. In preperation for this project I would getting my SG504 up to spec and had a catastrophic failure of some of the mechanical components in the unit. 60 year old plastics. unfortunately this is a custom part and as I do not have a lathe to fabricate another one at the moment I am stuck with out an appropriate signal source. I would like to get a signal generator used or new that can do sine waves accurately between .5 and 5.5 Vpk-pk and cover a frequency range from 9Khz to > =1050 MHz if possible. I don't mind new equipment or old however I would like it to have a 10 Mhz ref in if at all possible. Space is not an issue so it can be on the large size if needed. I have an HP 8657A but it will not go high enough in voltage for my current needs.

Zen

Use a sweep generator with provision for external power leveling e.g. HP 8350B/83522A. Connect to the input of the scope under test via one output of a power divider with the other output of the power divider connected to a diode detector (HP 432B) back to the leveling input of the sweep generator plug-in. The use of the external detector close to the scope input ensures that the frequency dependent loss in the connecting cable is compensated for. This is equivalent to the function of the leveling head for the SG504.

For lower frequency applications the external power leveling arrangement can be dispensed with and, with knowledge of the frequency dependent loss in the cable, you can have the sweeper internally compensate its output level to cancel the cable loss variation.

While the amplitude flatness performance probably won't be quite as good as the SG504 specs the sweep generator arrangement is more versatile as you can easily sweep over a wider frequency range. With a different sweep generator plug-in you can extend your measurements to >20GHz which might be useful for testing sampling scope performance.

[bdunham7]

I have an HP 8657A but it will not go high enough in voltage for my current needs.

Is that 5.5V_p-p into high impedance or 50R?

[Zenwizard]

The SG504 is 5.5Vpk-pk in to 50 ohms at the moment my main needs are 500 Mhz or less sence I am doing a calibraton on a 7904A 500 Mhz frame. So really all my scopes make use of the sg504 low range.

[Zenwizard]

However in reviewing the spec sheet for the SG504 the leveling is +/- 4% of the set point at the carrier 50 Khz or 6Mhz. This seems like it is not too exacting as you roll up the frequency? I would imagine modern gear should be able to meet or beat that spec. but I am not sure how that translates in to the db world.

[David Hess]

4% is 0.35dB.  When I looked into it a few years ago, common RF signal sources had at least twice that leveled amplitude error.  It is not something modern instruments do better because there is no need for it in common applications.  A signal generator is either a leveled signal generator or it is not.

When I measured my SG503 with my sampling oscilloscope, it was level to within 1% or 0.09dB, and much of that error could be attributed to the cable calibration even at the lower frequency.

[alm]

The more modern replacement is an oscilloscope calibrator. Used scope calibrators from Wavetek or Fluke seem to cost less on eBay than DMM calibrators, probably because of lower demand, but probably still thousands of dollars.

Like with the SG504, make sure you get the levelling heads.

[2N3055]

Is this good enough:

Amplitude flatness -0.3 +0.3 dB 50Ω load, 0.5 Vpp, compare to 1MHz Sine

Limits from verification procedure:

Select Sine waveform of the generator and set the amplitude to 0.5 Vpp
and frequency to the values listed below in sequence.
Frequency CH1 (dBm) CH2 (dBm) Spec Range (dBm)

1 MHz -2.3412~-1.7412
5 MHz -2.3412~-1.7412
10 MHz -2.3412~-1.7412
50 MHz -2.3412~-1.7412
100 MHz -2.3412~-1.7412
150 MHz -2.3412~-1.7412
200 MHz -2.3412~-1.7412
250 MHz -2.3412~-1.7412
300 MHz -2.3412~-1.7412
350 MHz -2.3412~-1.7412
400 MHz -2.3412~-1.7412
450 MHz -2.3412~-1.7412
500 MHz -2.3412~-1.7412

That is Siglent SDG6052X AWG.

For more accuracy and BW than that, leveled RF gen is what you need.

[Zenwizard]

looks to be the closest I have seen to duplicating the functionality I am curious if it can drive 20 V pk-pk in to 50 ohms or if that is the unterminated the spec sheet did not say.

Zen

[2N3055]

50 Ohm is 10V p-p..
And there are some amplitude limits at higher frequencies.

[jonpaul]

Bonjour, we have worked for decades with TEK CAL, have 7104, 7904, many 2465B.

Rather than CAL of freq resp, we find the transient response adjustments most critical and that a linearity check and adjustment is also required.

The groups.io TekScopes group has many threads on the topic of CAL generators and scope CAL.

1/ You can easily solve the HF leveled with HP RF gens like the classic Cavity gen 8640B 500 kHx...512 or 1024 MHz. . You can  find at Ham fleas, heavy to ship on epay.

2/ We use SG 503 up to 250 MHz, we never got a good SG504 with level head.
The excellent TEK CAL set is TG501/A Time Mark gen
https://w140.com/tekwiki/wiki/TG501

and  PG506, Calibration Gen.
https://w140.com/tekwiki/wiki/PG506

Finally the rare and excellent (but flaky elastomeric switches!)  TEK 7000 CAL fixtures 067-0587-01 (7904 500 MHz) and -02 (1 GHz 7104)
https://w140.com/tekwiki/wiki/067-0587-01

3/ For transient CAL  an alt to the usual pulse gens and PG506 is the amazing Leo Bodnar 40 pS pulse, a tiny board with aBNC that fits direct on the 50 Ohm input of a DUT.
Just L50! http://www.leobodnar.com/shop/index.php?main_page=product_info&products_id=295

I hope this proves useful,

Bon courage!

Jon

[niconiconi]

At the RF/microwave section of the forum, G0HZU described how to create externally leveled sinewave with a standard signal generator (with AM support down to DC) and an RF power meter, by building an opamp error amplifier. I also talked about my idea on whether it's possible to create a self-contained external automatic level control module with a variable attenuator and a much cheaper RF detector (diodes or high-performance RF log amps). The main problems are the max input power (5 V is 24 dBm, for most power meters and variable attenuators, it's out of range), and the tolerance of the attenuators.

https://www.eevblog.com/forum/rf-microwave/homebrew-rf-levelling-amplifier/

[alm]

If you look at one of the earliest Tektronix leveled sinewave generator 067-0632-01, they use a very similar principle of a GR unit oscillator modulated by the feedback from the detector head.

[niconiconi]

If you look at one of the earliest Tektronix leveled sinewave generator 067-0632-01, they use a very similar principle of a GR unit oscillator modulated by the feedback from the detector head.

During my research for this project I also learned it's still a standard practice in calibration laboratories for general-purpose signal leveling, e.g. see https://download.flukecal.com/pub/literature/6005965a-en-RFCal-power-splitters-dividers-an-w.pdf

[perdrix]

If you have an SG504 with no levelling head goto: http://perdrix.co.uk/SG504Head/index.htm

I did that some years agao, but I still sell the kits.