Marconi 2024 Signal generator - high power 25dBm version vs 13dBm version

[G0HZU]

Hi, are there any users of the low and/or high power Marconi/IFR 2024 signal generator on here?

I was hoping to provide some info about the pros and cons of each type. I've used both the 25dBm version and the 13dBm version for many years. Although it's initially tempting to always choose/purchase the high power variant (25dBm) over the standard (13dBm) version the performance of the high power version has some issues that may make some prospective purchasers favour the low power variant.

The two main issues are to do with the quality of the source match at output levels in the mW region and also the noise floor. It would be interesting to do some comparisons between the two signal generator models to see if other people have found similar issues.

Thanks in advance for any replies... Jeremy

[G0HZU]

Not much interest so far...

However, see the results of the first comparison. This tests the source impedance of the 25dBm variant at low frequency. This test was done at just 1MHz. Once the power level of the high power (+25dBm) 2024 is increased above +7dBm the last section of the step attenuator gets switched out and this means the source impedance won't be as well defined. Most sig gen manufacturers do manage to define the source impedance via the ALC feedback and a series 47R or 50R resistor. However, this doesn't seem to be the case with the high power 2024 sig gen.

Here's a quick test comparing how well the sig gen transfers power to a suite of resistors:

The test resistors are: 18R, 22R, 24.5R, 27R, 39R, 51R and 68R. The Vrms across each resistor was measured using a Racal 9300 true rms meter at 1MHz.

Here's the results at +7dBm
R      mV rms   Power   
18   0.2595   0.00374   W
22   0.300   0.00409   W
24.5   0.324   0.00428   W
27   0.3446   0.00440   W
39   0.4325   0.00480   W
51   0.501   0.00492   W
68   0.5738   0.00484   W

Not bad... the 51R resistor has the highest power transfer and the power level is very close to the expected 5mW.
This is because the source impedance is improved closer to 50R by the step attenuator.


Here's the results at +8dBm

R   mV rms      Power   
18   0.336      0.00627   W
22   0.3765      0.00644   W
24.4   0.4013      0.00660   W
27   0.4183      0.00648   W
39   0.4942      0.00626   W
51   0.5465      0.00586   W
68   0.597      0.00524   W

As you can see the maximum power transfer is now with the 24.5R resistor and this suggests the source impedance  has changed to just 24.5R above +8dBm. This is not good. Not good at all if you consider this is happening at just 1MHz.

If I compare the open circuit EMF in each case the EMF is 1.03V when at +7dBm but when the sig gen is clicked up 1dB to +8dBm the EMF drops to just 0.834V.

This EMF issue at +8dBm is also bad but it makes sense because Marconi have adjusted down (fudged) the EMF such that the power transferred from the 24.5R source impedance is correct for a 50R load.

Can anyone do the same tests for their high power 2024 sig gen? The standard (+13dBm) 2024 should perform much better here. This is an issue that affects the high power (+25dBm) variant and not the low power variant.

At one time I had three of these high power Mi 2024 sig gens and they were all the same.

[G0HZU]

Note that it is possible to offset this issue by enabling the ATTEN LOCK feature on the sig gen but this only works up to +13dBm. Above this power level the poor source impedance issue returns.