Boonton 102F on Ocilloscope

[DannyKlenz]

Forum,
      I recently acquired a Boonton 102F. I hooked up to my oscilloscope and I am getting an unusual signal back from it. Im not sure if its my cabling causing reflection or what. My goal is to fix this and learn as much as I can while doing so. Electronics have always been an interest of mine but I never found the time to really dig in to a project like this.

Things I have tried:

1. I have tried every combination of probe/terminator combo I have on hand and still get the same result.

2. I recapped the power supply using Nichicon caps. Original values were 2x 2400uf 40v DC for the +15 and -15 rails and 1x 5300uf 20v-30v Surge DC for the +5 These were replaced with 2x 2700uf 50v DC and 1x 5600 35v DC

3. Adjusted the power supply so all voltages are within spec

4. Bought a copy of the service manual.....

Below is a link to a youtube video I just posted showing the results I am getting on my scope.

https://youtu.be/hgCUPs4UhLA

[Mr Evil]

It looks like its clipping. Does the problem go away if you reduce the output amplitude? How much DC offset is there?

[DannyKlenz]

The link below is a picture of the lowest output I can measure with the scope. Sign wave is still distorted. I used the top measurement which I think is the DC offset? There are a dozen or so tantalum caps throughout the device. Even on the power supply board. I've read that they can be an issue but only in certain situations. There are also a couple electrolitic caps as well. Maybe these are part of the problem?

http://imgur.com/a/J023PMg

[Mr Evil]

I don't know what "top" is, but it looks like you've set it to DC-coupled, and the signal is only a few mV from perfectly centred, so the DC offset is ok, on the output at least. It does still look distorted, but it isn't toally flattened on one side like in the video. If you run it like in the video so the peaks are flattened, then reduce the amplitude a little at a time, does it become less distorted, like the peak amplitude on one side is limited?

I'm thinking that somewhere along the signal chain there's a faulty coupling capacitor, so the signal to the following stage is no longer centred. Presumably the service manual will have detailed enough schematics to let you locate the coupling capacitors. Tantalum capacitors can fail short, so checking those might reveal something.

[rf+tech]

Hi Danny,

I've serviced a number of the Boonton 102 A through D. The wideband output amplifier and ALC detector are both ahead of the step attenuator, so the distortion seen will be present at all steps regardless of range. ALC loop alters drive level via a PIN diode attenuator, located ahead of the wideband amp input.

With the output step attenuator set to the highest level, vary the vernier level control, while observing the distortion. Let's see if reduced drive to the wideband amp resolves the distortion. From experience, it would not be unusual to find one or more carbon composition resistors in the wideband amp with increased resistance, altering bias on the final as the culprit.

RF+ Tech

[DannyKlenz]

I will definitely look into this. I had some time to play around with it last night and noticed when I raised the voltage on the power supply above spec the clipping of the waveform was reduced. I then went around measuring inputs on all the boards and im almost -0.5v down on the 5v and about -0.4 on -+15v so something is going on there as well. I suspected carbon resistor having dealt with them in a Heathkit SB200 I currently have in pieces under by bench…… So far this has been an awesome learning experience and I appreciate the help! Strangely I now find myself glued to the usual places looking for a good price on various pieces of test equipment to help with this project and others.

Dan

[rf+tech]

Dan,

Glad to be of help. The 102 series power supplies have single-pin push-on terminals for the harness connections, a great source of !*R drop. Easy enough to measure by comparing readings at the solder side of the PS pcb and individual modules, so as to include the drops across the terminals with the harness drops.

0.4 Volts low on the 15 Volt supplies is within ±5%. 0.5 Volt low on the 5 Volt supply is outside of the ±0.25 Volt spec. The 102D uses an LM323 fixed regulator, presumably the 102F uses the same. Measuring the LM323 output directly between its output and ground pins will give a clearer view of the situation.

RF+ Tech

[DannyKlenz]

I bypassed the attenuator and have the probe on the output of the wideband amplifier. The vernier adjustment is not having an effect on power output.

[rf+tech]

No ALC may very well account for the distortion. Fortunately, this is dirt simple to troubleshoot. Enable 1 kHz AM, set the scope timebase to 200 microseconds/div and see if two cycles of 1 kHz sine wave are present, riding on the RF envelope. Varying the % Modulation control should show a corresponding change in the 1 kHz amplitude. At zero modulation, just the RF envelope should be observed.

This verifies the PIN diode attenuator, which also serves as the amplitude modulator. Since the AM amplifier does double duty as the ALC feedback amplifier, it needs the correct DC bias on its non-inverting input from the ALC detector, to produce linear AM. My guess is the 1 kHz AM will also show some distortion, since there is no vernier RF level control

The 102D ALC detector uses some hand-soldered SMD components, a chip coupling capacitor and two chip resistors that connect to the schottky detector diode cathode. Any one of these might be open circuit. Or the detector diode might be blown.

Please post the 102F Output PCB schematic page, as I'm working from the corresponding 102D schematic.

RF+ Tech

[DannyKlenz]

Alright, below is a video of my scope while following your instructions. And a link to the amp pcb for your viewing pleasure....

http://imgur.com/a/tsEoVct
https://youtu.be/h6UyjUtcQd8

[DannyKlenz]

CR 14 is reading .350 in the bad direction. I have a bunch of 4000 series diodes 01 04 07 etc... Going to check if I can drop one in

[rf+tech]

Dan,

To see the modulation envelope, it is necessary to slow the sweep speed d-o-w-n to audio. Reread my previous reply. The actual RF cycles will be indistinguishable. To see two cycles of 1 kHz requires a sweep speed of 200 microseconds.

I see that the initial RF signal without modulation looks pretty good. What changed between the last video showing distortion? Disregard the distortion at high modulation depth. Focus on the fact that before turning up the AM control, the signal looks clean. What made the difference?

Thanks for posting the schematic. I see Boonton changed a few things here, between 102D and 102F. The above mentioned coupling capacitor at the Schottky detector diode has been removed. And the vernier control signal gets applied off-board to the AM input (bottom edge of the Output PCB schematic). Let's see, this may be located on the Output Level Meter PCB. We need to locate the page with the RF Output vernier control, to see where the control voltage gets summed into the AM input.

Hold on a moment, CR14 cannot be replaced with a 1N400x. The only suitable replacement for CR14 is another Schottky diode such as 1N5711. The forward drop of a Schottky diode is about 0.35 Volts. So your measurement is good, not sure what you mean by "bad direction".

RF+ Tech

[DannyKlenz]

Sorry I got ahead of myself. Went and read the datasheet and realized my mistake. I'll see about locating the vernier adjustment.

[rf+tech]

Umm, the vernier is right there on the front panel...  the control just above the N connector, marked OUTPUT.

Vernier, as in fine adjustment and as opposed to fixed 10 dB steps.