Help with 10:1 oscilloscope probe

[neazoi]

Hello, I have build the Fig 2. load and probe from this page https://www.ab4oj.com/test/pwrmeas.html in order to measure a 1w @50R transmitter with my scope.

First I just connected a 50R load without the divider. I think (without the divider) the writer assumes a 1M scope input for the measurements, am I right? Because if the 50R scope input is used it will be paralleled with the 50R load resistors ti give 25R. Am I right?

Second, when I build the 10:1 probe (voltage divider) together with the load as shown in figure 2, gives about 2.5vpp (measured at 1M scope input). But my signal is 6.3vpp, measured at the 50R scope input alone.

What is the problem?

 

[FvM]

You should talk about signal frequency, too.

 

[neazoi]

You should talk about signal frequency, too.

Yes sorry. HF 1-30MHz

 

[FvM]

Consider probe capacitance and see why the circuit doesn't work for 30 MHz. 1 MHz should be O.K.

 

[tggzzz]

As FvM says, you have to consider how your measurement device affects the circuit being measured. In addition at 30MHz, your construction technique and components might be changing the performance.

A "normal" way of achieving your measurement would be to keep the signal 50ohms all the way to the scope, but to insert a 50ohm 10dB or 20dB attenuator in the signal path. Such things widely available, but make sure it is rated dissipate the transmitter's power.

Alternatively you could consider making your own "low impedance Z0" probe, which should be OK to GHz and avoids many of the problems of conventional 10:1 probes. See https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/ for more information.

 

[neazoi]

As FvM says, you have to consider how your measurement device affects the circuit being measured. In addition at 30MHz, your construction technique and components might be changing the performance.

A "normal" way of achieving your measurement would be to keep the signal 50ohms all the way to the scope, but to insert a 50ohm 10dB or 20dB attenuator in the signal path. Such things widely available, but make sure it is rated dissipate the transmitter's power.

Alternatively you could consider making your own "low impedance Z0" probe, which should be OK to GHz and avoids many of the problems of conventional 10:1 probes. See https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/ for more information.

I have tested the circuit at 7mhz. The scope input is 1M. The 50R load is connected across the TX output and parallel to it there is a voltage divider, as shown in the schematic. The leads of all resistors are kept as close as possible, the whole construction is made on a proto board and it is not larger than a coin. So I think at 7MHz there should be no problems with it.

When I measure the TX directly at the 50R input of the scope (without an external TX load), the TX outputs 6.3vpp (about 100mW).
So with the 10:1 probe, connected to the other input of the scope (1M), I should read 630mvpp. But I read at least half that or so.
What could be wrong?

 

[tggzzz]

The scope input is 1M.

No, it is not.

By stating that you have demonstrated that you (1) ignored FvM's reply where he wrote "Consider probe capacitance..." and (2) haven't read any of the references I provided.

The 50R load is connected across the TX output and parallel to it there is a voltage divider, as shown in the schematic. The leads of all resistors are kept as close as possible, the whole construction is made on a proto board and it is not larger than a coin. So I think at 7MHz there should be no problems with it.

When I measure the TX directly at the 50R input of the scope (without an external TX load), the TX outputs 6.3vpp (about 100mW).
So with the 10:1 probe, connected to the other input of the scope (1M), I should read 630mvpp. But I read at least half that or so.
What could be wrong?

Hint: a normal scope's input capacitance is ~20pF, or a 10:1 probe's capacitance is ~15pF. I'm sure you can calculate its impedance at 7MHz, and therefore the effect that will have on the measured voltage.

 

[neazoi]

No, it is not.

By stating that you have demonstrated that you (1) ignored FvM's reply where he wrote "Consider probe capacitance..." and (2) haven't read any of the references I provided.



Hint: a normal scope's input capacitance is ~20pF, or a 10:1 probe's capacitance is ~15pF. I'm sure you can calculate its impedance at 7MHz, and therefore the effect that will have on the measured voltage.

Thanks for the info.
I spotted the error. It was the voltage divider. This was connected parallel to the load. The TX is only a few watts so the 1M scope input connected in parallel to the 50R external load can measure directly without any 10:1 resistive divider.. I have tested it and the measurements were very close to the ones when the TX was connected directly to the 50R input of the scope, without an external 50R load. The scope has 250V max input at 1M and 8pf capacitance @ 1M, so it should be adequate for measuring a few watts.

I believe the small error is due to the capacitance and the things you mentioned.