Sideband harmonics 7mhz away...wheres your test tone?

[TheRealPorkchop]

It’s probably that D-Rail or however he spells his name, radio. Some hopped up Cobra 29 that’s supposed to put out 1.21 jiggawatts or whatever.

Or a Trucks CB Sales radio.

I guarantoldya it’s a crappy, dirty, super splatterbox radio and not a run-of-the-mill average black box splatter box.

 

[kopcicle]

they did not have any leakage that I could see.

on what?

into a 100W amp and checked again and still no discernible 7.8 or 35Mhz leakage.

measured how?

I can't say I have ever seen VCO or carrier osc leakage out of one of these radios

I have. Granted they needed a bit of a touch up.

substantial enough to notice

Questionable as to "how much" , but I have seen it. It took a bit of verification to be sure it wasn't noise. It was approximately 23db down and at the limits of my antique gear.

generally good enough to knock this down to almost nothing

"generally"

so it

not so simply

gets lost in the rest of the chain.

I have had more issue with image and inter-modulation products in the output of the PC-412 than VCO and carrier leakage. To Uniden et all a very good job at carrier suppression . To golden screw drivers in general , not so good a job at VFO suppression.

Martian, I'm not busting your balls over this at all. I'm genuinely curious what you have seen in a worst case scenario. I'm about to do things to this mixer arrangement that it was never designed for.

 

[kopcicle]

Cobra 29 that’s supposed to put out 1.21 jiggawatts or whatever.

More than likely a 2166/1969 with a bit of AC divider gymkhana between driver and final.
As for the variable output ask Nomad about the "lick and stick" variable.

 

[TheRealPorkchop]

Every time I build that variable it ends up being a firecracker. I gave up on that

 

[Martian]

on what?
measured how?

Radio transmitting 4 watts, channel 20, no modulation.

I did the test on a Siglent SSA3032X with BBIs span with and without an amp. Never saw anything. Using a Bird variable RF sampler between radio and dummy load.
At that span with an RF sampler knocking 50dB off its in the noise +20dBm on the instrument.

Now, if I repeat this and go looking for VCO leakage I see it.
On my guinea pig radio the VCO leaks out at levels around -90dBm. (ch 20, so VCO @ 35.005Mhz) SA connected directly to the antenna. (NOT transmitting, never do that!)
With no attenuation in the instrument the VCO is around -82dBm with the 50Mhz span. Reducing the span to 100Khz I have a solid -82dBm VCO output with noise floor down at ~115dBm.

Found 7.8Mhz down at -120dBm
(with sample port connected, radio in TX obviously, ref levels set accordingly, 10Khz span.)

Seems inline with your observation of "not so good VCO suppression"
7.8 just isn't leaking out in any level to be concerned with and even the VCO isn't terrible. Could be better.

With a 100 watt crap built amp the VCO got brought up around 5dBm.
7.8 was like a couple dBm. it's just so far down there already and I have not swept this amp to see it's frequency response. It may not be doing much at 7.8Mhz despite being capable.

I would still say it's not enough to be concerned with. But whatever BBI has going on I would say is a serious problem. It's not right.

 

[Handy Andy]

I'm all for and with you folks on the Cobra / Uniden way of thinking. But there is still the issue of the sheer existence of the 7MHz - and the best I can come up with is from the tweaking of the B.P.F and output tank.

Heck, most Base radios of the day also had a 52MHz tuneable trap in them to fine tine (SIC) the shunt past the Pi network - it's sad to see that they have done away with this simple filter - for even a simpler cap change and retune can remove the spurs of the 7.8MHz (35MHz and the 19~20MHz) this is pretty close to many Stereo receiver IF downmixers from FM and the 19kHz stereo "detect" embedded in the signal. So was the days of when you'd get phone calls or pounding on the doors from angry neighbors wanting a piece of you or your equipment - it's simply the "rattle - trap" of the mixing artifacts from the DA that doesn't set the Carrier Balance right...

The evidence of the down and dirty is there, no point in hashing over burnt ashes from the discovery - we still have to repair the dang barn, the door came off of...

 

[TheRealPorkchop]

Radio transmitting 4 watts, channel 20, no modulation.

I did the test on a Siglent SSA3032X with BBIs span with and without an amp. Never saw anything. Using a Bird variable RF sampler between radio and dummy load.
At that span with an RF sampler knocking 50dB off its in the noise +20dBm on the instrument.

Now, if I repeat this and go looking for VCO leakage I see it.
On my guinea pig radio the VCO leaks out at levels around -90dBm. (ch 20, so VCO @ 35.005Mhz) SA connected directly to the antenna. (NOT transmitting, never do that!)
With no attenuation in the instrument the VCO is around -82dBm with the 50Mhz span. Reducing the span to 100Khz I have a solid -82dBm VCO output with noise floor down at ~115dBm.

Found 7.8Mhz down at -120dBm
(with sample port connected, radio in TX obviously, ref levels set accordingly, 10Khz span.)

Seems inline with your observation of "not so good VCO suppression"
7.8 just isn't leaking out in any level to be concerned with and even the VCO isn't terrible. Could be better.

With a 100 watt crap built amp the VCO got brought up around 5dBm.
7.8 was like a couple dBm. it's just so far down there already and I have not swept this amp to see it's frequency response. It may not be doing much at 7.8Mhz despite being capable.

I would still say it's not enough to be concerned with. But whatever BBI has going on I would say is a serious problem. It's not right.

You gots to get you one of them D-Rails, come on.

 

[Shockwave]

Further investigation of the "8719" schematic makes it somewhat doubtful the radio is responsible for what we are seeing here. Directly after the SO-42P mixer chip there are two 27 MHz. tuned circuits. The inductors do not have enough range to tune 7 MHz. off of resonance (in both directions) and if they did, they would no longer be passing much of a signal on 27 MHz. As I recall, being off a little on the alignment here makes a big drop in RF output. It's looking more likely that the analyzer was being over driven. Ten more db of attenuation on the input might make the whole problem go away.

 

[Tallman]

I have no leakage on my radios what so ever. Radio desk is always dry as a bone!

 

[Handy Andy]

You don't really need to. The 7.8 is the signal applied to the base on the mixer amplifier in the circuit - it's a ringing effect from the mixer "pulse"



The 7.8MHz is used to remove or slow-down a 35MHz IF - you see that artifact not from stopband or skirt - but from "ringing" WITHIN the passband that is from the pulse the SO42P uses to stop the 35MHz signal - it's the very one that comes off of the Xtal filter - if the filter is off or the guy modified it, you'll see such stuff more pronounced.

Hence my reference to the "Carrier Balance" not set correctly.

PDF of the SO42P attached.

The newer replacement is even simpler and in some ways easier to fool to produce artifacts that will cause these spurs...


Notice there is no other means to dampen - attenuate, filter or otherwise cutoff any artifacts. This is what (Those Gosh Darn People That Think They Can Get Away With Anything) being used as workarounds for type-accepted designs folks. They don't re-issue a certificate if they can make a chip to replace the one that was there. Yes, that wonderful loophole - not exact... but Hey! Use what works...

 

[Martian]

This is what China is using as workarounds for type-accepted designs folks. They don't re-issue a certificate if they can make a chip to replace the one that was there. Yes, that wonderful loophole - not exact... but Hey! Use what works...

A bit O/T but that's not a "China" thing but an industry thing.
Can't blame the Chinese for that one.
I have done it myself to avoid the regulatory capture that is the FCC/IC, etc.

 

[Handy Andy]

Yeah, you're right - they use what we use and do what we do - so nothing new - updated it to be more specific...

 

[TheRealPorkchop]

Martian is Tom and BBi is the little one. Hahahahaha

 

[Shockwave]

Neither mixer has any internal components to filter out the IF frequency spurs from the output of the chip. Since this is dependent on the specific frequencies feeding the input of the chip, filtering takes place externally in the tuned circuits formed by L45 and L46. These are the first parts the output of the TX mixer passes through on the 148GTL.

 

[loosecannon]

you can pretty much guarantee that any radio BBI uses to drive his amps has a 10.695mhz IF.

the old uniden chassis don't do that 16:1 modulation to carrier ratio nearly as well as the 'exports'.
LC