Oscilloscope recommendation for measuring SX-770 10.7 MHz IF signal?

[paulcalif]

I'm troubleshooting a weak FM tuner on a Pioneer SX-770 and have another working SX-770 available for comparison.
I'm trying to measure the 10.7 MHz IF output from the FM front end, at the point where it enters the IF board (W31-002, terminal 3, marked FM IN). My goal is simply to compare the signal level from the working receiver with the problem receiver and determine whether the loss is occurring in the front end.
I tried a FNIRSI 2C53P 50 MHz scope, but even on the working receiver I haven't been able to get a reliable, repeatable display of the IF signal. The lowest vertical setting I can get is about 10 mV/div using a 1X probe. I've confirmed the scope will display a known 2 MHz signal, so the scope itself appears to be functioning.
What type of oscilloscope/probe would you recommend for reliably viewing the 10.7 MHz IF signal at this point in the circuit? Is the issue likely to be scope sensitivity, probe loading, or am I approaching the measurement incorrectly?
I'm not trying to do a complete FM alignment at this point—just compare the front-end output of a known-good SX-770 with the weak one.
Thanks for any advice.

 

[Olle]

Have you tried the 10x setting on the probe?

 

[paulcalif]

Thanks for the reply.

Yes, I tried the probe at 10X with the scope also set to 10X. At 10X, though, the lowest vertical setting the scope allows is 100 mV/div, and I still couldn't get a usable signal.

 

[ConradH]

An X1 probe will have high capacitance and is unsuited to high frequency work. It will severely limit the bandwidth of the scope. The 10X probe is the correct probe for the job, unless you have some fancy powered FET probe, which most of us don't. The problem is now the signal is 10X smaller, though in practice the 1X probe would have loaded and attenuated it anyway. That can also be an issue with the 10X probe. We all talk about 10X probes, but they're really /10 probes. Anyway, I'd expect the signal at the output of an IF strip to be large enough to see and trigger easily, since that's where most of the gain usually is. Note that the scope ground is critical and you can't get away with some old long clip-lead. You need a short ground connected at the probe and a ground near the output of the IF. I think you need a timebase setting of about 10 nS/div but somebody double check me on that. Digital scopes will lie if the timebase is way off. Also, I don't know the specs but turn off the unused channel. Most digitals divide the bandwidth as you enable channels, so it might be a 25 MHz scope if two channels are used. Can you post a screenshot of the 10.7 signals?

 

[paulcalif]

Thanks, that's very helpful. I did try the 10X probe, but switched to 1X because the scope goes down to 10 mV/div at 1X, while the lowest setting at 10X is 100 mV/div. I hadn't considered how much the capacitance of the 1X probe could load the circuit at 10.7 MHz.
My ground lead is about 6 inches long and I had it connected to the chassis, not to a ground point close to the FM IN connection. I had assumed a poor ground would mainly show up as interference, but it sounds like at 10.7 MHz the ground lead itself could also affect or attenuate the signal.
The second channel was off.
I don't have a screenshot of a confirmed 10.7 MHz signal. I briefly thought I was seeing one, but I couldn't reproduce it consistently on either the good or bad receiver, so I now suspect it was interference or an artifact.
I'll try again using the 10X probe with the shortest possible ground connection near FM IN and see what I get. Thanks again.

 

[mrehorst]

The BW you get from a digital scope is a function of the sweep rate, sample rate, and memory depth available. The number advertised is the maximum possible BW at the maximum possible sample rate, not what you get for every measurement. The scope will fill the available memory by adjusting the sample rate (and therefore BW) to do so. If there's 1 megabit of memory available, and you set the sweep to 10us/div=100us for the whole screen, the scope will fill the 1 megabit in 100 us. So it will sample at 100 Msps. BW is 1/10 of the sample rate. So the BW for that measurement will be 10 MHz.

If the scope has multiple channels, the A to D converter is usually shared so memory, and therefore BW gets divided between the channels.

There are a lot of cheap digital scopes out there that look great until you really dig into what they can do. Then they don't look so good. The marketing stuff almost always overstates the BW, and people don't realize that the BW changes each time you change the sweep setting or turn on a second channel. Better digital scopes will show the sample rate on the screen so you know what BW applies to the measurement you're making. Another thing to look at is the bit depth. Good scopes will use 12 (or more) bit A to D converters.

If you're looking to buy a digital scope, expect to pay at least a few hundred $ for a decent one.

 

[paulcalif]

Thank you. That explanation helps a lot. I hadn't understood that the effective bandwidth of a digital scope could change with the timebase/sample rate. I was thinking that since the scope is rated at 50 MHz, a 10.7 MHz signal should simply be visible regardless of the horizontal setting.
That may explain some of the trouble I've been having. I'm new to oscilloscopes and was using AUTO quite a bit, so I wasn't necessarily controlling the timebase myself. I now know how to set it manually.
For a 10.7 MHz IF signal on a 50 MHz digital scope like this, what timebase would you suggest I start with to make sure the scope is sampling fast enough? The scope is a FNIRSI 2C53P.
I actually boxed it up to return it after failing to get a repeatable signal from either receiver, but I'm beginning to think I may need to unpack it and give it another try with the proper settings.
Or are you saying this little scope isn't capable of measuring this.

 

[mrehorst]

Finisri says the memory is 13M, but that isn't the waveform memory. I believe those scopes have 64kbit waveform memory, though I can't find it in any of their marketing BS. I don't think it has sufficient BW to be looking at 10.7 MHz waveforms.

eeblog always has good, real info on digital scopes. They often do comparisons, and there are a lot of user comments.
See: https://www.eevblog.com/forum/testgear/new-handheld-touch-dso-fnirsi-2c53p-claimed-50mhz250msps/
The consensus is that the Finisri scope is a toy.

I have a Siglent SDS804X. 200 MHz BW. They sell for about $500 these days. They were $400 when I got mine about 2 years ago. You buy the 70 MHz version then apply a simple hack to unlock the full 200 MHz BW. See: New test equipment: Siglent DSO and AWG

 

[paulcalif]

Thanks, that's very helpful. I think you've probably saved me from wasting a lot more time trying to make this scope do something it isn't well suited for.
I bought the FNIRSI specifically because I wanted to compare the 10.7 MHz IF signal from a working Pioneer SX-770 with one that has very weak FM reception. On paper, 50 MHz/250 MSa/s sounded like it should be more than adequate, but in practice I haven't been able to get a reliable 10.7 MHz signal from even the known-good receiver.
I had already boxed the FNIRSI up to return it, then some of the earlier suggestions had me considering unpacking it and trying again. After reading your explanation and the EEVblog discussion, I think I'll leave it in the box and return it.
I'm not opposed to spending more for a proper bench scope if that's what it takes. I'll take a look at the Siglent SDS804X you mentioned. Most of what I work on is 1970s audio equipment, but I would like enough capability to troubleshoot FM tuner/IF sections when necessary.
Thanks again for taking the time to explain this. I'm learning as I go, and this has been very helpful.

 

[ConradH]

I use the Siglent SDS8xxX HD series and think they're a great scope for the money. Not perfect, but no scope is. I gave away all my other scopes, 4 big Teks, and just use that now.

It's been a while since I looked at an IF. Be like a lawyer and don't ask a question unless you know what the answer should be! What should you expect to see at the output of an If strip? It's a narrowband 10.7 MHz filter with a lot of gain. (I think) the only thing you should see is a 10.7 MHz signal that's the sum or difference of the oscillator and the FM RF, IOW probably not much unless you have a good FM signal coming in. Or, the response if you sweep the IF input with a signal generator around 10.7 MHz. I don't think most people put a scope on this but infer what's going on from other points down the line.

 

[paulcalif]

“Be like a lawyer and don't ask a question unless you know what the answer should be!”
I love that. I think you've identified one of the mistakes I've been making. I'm new to scopes and was basically looking for “a 10.7 MHz signal” without first establishing what I should actually expect to see at that particular point and under what test conditions.
I have the SX-770 service manual and also have another working SX-770 for comparison, so I'm going to go back through the manual and figure out exactly what the expected signal path and test conditions are before probing further. I also just ordered a Rigol DHO804, so when it arrives I'll have a proper scope rather than trying to determine whether the FNIRSI is showing me the truth.
Thanks — that gives me a much better way to think about this.