开云体育

Having some issues with weak/no response crystals in the Drake 2B.? The 11 mhz 40m xtal works just fine.? I've tried two 18 mhz 14 mhz that looked OK on the scope/sig gen check but won't oscillate in circuit.? I've also tried two 25 mhz 15m xtals and one will oscillate but very weakly.

I've swapped out the 6U8 for a 6EA8 with the same results.? Has anyone tried increasing the gain of the xtal osc stage to get weaker xtals to work?? xtal rolling would be inelegant but OK if there were more of them available.

Any other ideas for measuring the xtal's output?

I've seen folks build external oscillator's that drive the circruit without xtals which is interesting.

I'm interested to here others experience with this.

Thanks,

de John, K8NN

I've had no trouble with HC-49 crystals in my 2B. At the time I found them on AF4K's website. See for details.

73, Niel - W0VLZ

John,

Redesign is unnecessary, from my perspective. ?The 40 m band oscillator is a Pierce oscillator (parallel resonance) that needs no tuning.
The 80 meter band is a phase shift oscillator, no crystal involved, and no additional tuning involved.

The higher frequency bands, 20 m through 10sub3 m, use the appropriately assigned crystal in a series resonance configuration which requires proper tuning of a coil: ?from the manual, “... the band switch inserts an appropriate coil resonating the circuit on the crystal overtone frequency and the crystal is switched between cathode and a low impedance point of the tuned circuit.” ?Since you did not mention it and not assuming you have completed the Crystal Oscillator Alignment on page 22-23, I recommend you follow this procedure using a VTVM with 10 Megohm input impedance or greater, or a DVM with the same DC voltage input impedance. ?Follow the complete procedure. ?There is a coil for 20 m, E band, and one coil for 15 m and all 10 m bands. ?You will adjust L9 first on 15 m, then check/adjust the same coil for 10 m. ?Oscillating on the third overtone for these “higher” bands, the radio’s oscillator will not perform sufficiently, unless the associated coil is tuned.

The band switch is an important part of this circuit for proper operation: you can check and clean the switch contacts, if not already do so.

If you actually need to replace a defective crystal, Neil’s (W0VLZ) recommendation is a good one, especially if you are not sure of causality.

73,
Michael Smith, N4KZO

Thanks Niel.? According to his website he's out of business.

Thanks, Michael.

Yes, I've done this.? Problem is that there is no neg. voltage on the 20M position so pointless to tweak the 20M coil.? Seems to indicate that the xtal is not oscillating or a bandswitch issue perhaps.? Will check that.? Strange thing happened yesterday.? When I plugged in either of my two 20M xtals, the 15M xtal quit oscillating.? Take the 20M out and 15 starts right up again.? Now that sounds like a switch issue.? This morning while tuned up on 15M and peaked on the noise.? Good neg voltage and lots of noise that peaked on the preselector.? Plugged the 20M xtal in and 15 kept oscillating.? Go figure.? Time for a few resistance measurement on the bandswitch and a bit more DeOxit.

Still not getting either of the 20M xtals to oscillate.

Going to be out of town for a couple days but will let you know what happens in a few days.

Thanks for the update Mike

John,

It’s not really pointless to adjust the 20 M coil, with L10 used exclusively for 20 M band resonance. ?There is a threshold inductance value that will cause oscillation: correct inductance is essential in this design. ?It is possible, if the coil is so far out of adjustment (or DC open-circuit in the winding), oscillation will not occur, confirmed with no grid leak bias. ?Consider noting the present position of the coil’s adjustment screw, then experimentally adjust this coil, without going beyond the mechanical stops. ?You can always return to the present coil position if an adjustment doesn’t help. ?But, yes, it sounds like point-to-point continuity checks and short-circuit isolation is in order, first. ?(I expect you have already done a close visual inspection.) ?Check the DC continuity through L10, from the junction (node) of C73 and C74 to ground, with the Band switch in the 20 M position; the capacitors should provide a good DC block to the remainder of the oscillator circuit, although you may see some capacitor charging. Then, switch to the 15 M band with the ohmmeter probes in these locations and confirm open circuit.

Do you know that the 18000.00 kHz crystals you have are series resonant, third overtone crystals? ?I have a Tedford Labs HC-6/U crystal in my radio — verifying, there are no markings on the container indicating parallel or series resonance, so I know of no way for you to confirm series resonance at this frequency, unless you have a dedicated test circuit.

If you haven’t already, try lightly tapping on these dysfunctional crystals with a finger nail during in-circuit operational check: this may provide indication of an circuit connection problem within a crystal.

73,
Michael Smith, N4KZO

N4KZO de K8NN

Thanks for the comprehensive analysis (as always).? I've tweaked L10 a bit but will do a continuity check and them another more extensive tweaking to see if I can get either of the 18Mhz crystals to oscillate.

Roger on the third overtone from all except the 11Mhz 40m xtal.? I can get a response putting a 5V P-P 6Mhz signal in series with the xtal then to the O'scope.? Freq is correct but the voltage output is quite low compared to the large voltage gain seen at resonance on a fundamental freq xtal.? Notice the same on both crystals so that leads me to believe it might be a characteristic of the overtone type of crystal?

Will get back to you after fussing with L10.

de John, K8NN

Mike:

Thanks for that.? They're not fully up and running on the site but I did e-mail requesting a quote on the 18Mz 3rd overtone xtal.

de John K8NN

It’s not really pointless to adjust the 20 M coil, with L10 used exclusively for 20 M band resonance. ?There is a threshold inductance value that will cause oscillation: correct inductance is essential in this design. ?It is possible, if the coil is so far out of adjustment (or DC open-circuit in the winding), oscillation will not occur, confirmed with no grid leak bias. ?Consider noting the present position of the coil’s adjustment screw, then experimentally adjust this coil, without going beyond the mechanical stops. ?You can always return to the present coil position if an adjustment doesn’t help.

Michael:

Thanks for making me take a closer look at this.? I was able to get one of my two 18Mhz xtals to resonate by running L10 in and out until the neg voltage popped up on the meter and then tweaking to peak it.? I also peaked the 15M xtal which had been resonating but not a? very high level.? Now all three, 40M, 20M, and 15M peak at pretty close to the same value, around -260-270mv, relative to the 80M value being zeroed out, per the book instructions.

The only problem I have now is that none of the crystals fit tightly into their sockets.? The 40M is so loose you'd think it would fall out if you held the radio upside down.? You have to push the 20 and 15 xtals around until you get them to a point where they oscillate and if they get jarred, they will most often stop oscillating until I wiggle them again.? Doesn't look like it will be easy, if possible, to get at the bottom of the xtal sockets and try to squeeze them a bit.? The access is awful and they're really small.? I'm open to suggestions on this.? Off hand I can't think of an easy way to add a mm or two to the xtal pins diameter either.

de John, K8NN

John,

Your perception is on track. ?Considering a passive LCR equivalent circuit model of a Quartz crystal mounted in a holder with wire connection leads, the third overtone, the next highest possible resonance after the fundamental, has a lower impedance, purely resistive at resonance, and much, much lower than the nearby parallel resonance, w.r.t. frequency. ?The series resonance occurs first, then as frequency increases slightly beyond series resonance, the crystal becomes inductive with a very high impedance at parallel resonance (again purely resistive), occurring at the fundamental and each valid overtone. ?With increasing overtones, the series resonant impedance (resistance) drops somewhat linearly. ?With the voltage across the crystal being proportional to the crystal impedance, the third overtone will (normally) result in a lower voltage.

Apparently when this radio was designed, these “higher” crystal frequencies were generally only available for third overtone. ?Hence, the Colpitts oscillator, using parallel inductance, was used for greater output. ?The third overtone series resonance crystal creates a low impedance in the cathode circuit, only at third overtone resonance resonance to lower negative feedback and increase the gain. ?With the tuned tank circuit in the feedback path, the Colpitts oscillator configuration is realized. ?The 2-B manual couldn’t accurately call this a Colpitts oscillator, because of the use of the low impedance path created by the crystal in the cathode circuit, and apparently didn’t what to take the manual-space to explain it thoroughly (alright with me). ?Even though fundamental frequency crystals are available at these higher frequencies these days, you what to use what the circuit was designed for, the third overtone series resonance (using AT-cut Quartz). ?I will be interested to know the difference in price between an HC-48/U holder and an HC-49/U holder, with 10 ppm crystal frequency accuracy, depending on how WD8DAS quotes your crystal.

Let’s see what the L10 adjustment will do, after checking the circuit wiring. ?Make sure someone before you did not modify the circuit.

73,
Michael Smith, N4KZO

John,

I am thankful that the oscillator is working now for all bands. ?You are welcome for any helpful suggestions. ?You have observed that sudden oscillation occurs when all conditions for oscillation are met, just as it should.

Concerning the crystal pins loose in their sockets, I assume the crystal holders are either HC-6/U or HC-48/U, same dimensions with both. ?In my 2-B, each crystal has both pins bent, very slightly toward each other to create a transition fit with the socket. ?The bend occurs at the holder insulated wall, as if someone gently pushed on the end of each pin toward the other, or used smooth-jaw pliers to make the bend. ?The offset from the perpendicular position measures 0.01 inches, using a Vernier caliper. ?The pin diameter measures 0.05 inches. ?Check your crystals. ?If you don’t see this slight offset from perpendicular, bend each pin toward each other for a transition fit. ?My crystals plug into their socket smoothly, and quite snugly, all the way. ?I expect that the socket barrels flex from perpendicular, conforming to the offset in the pin, when the crystal is inserted, creating the secure transition fit.

73,

Michael Smith, N4KZO

N4KZO de K8NN

Thanks Michael.? I had been thinking of that but in the other direction.? I managed to get the full alignment procedure done while all three were oscillating.

Would be good to go except the 20M xtal has gone intermittent again.? It seems to be something to do with the socket because if I wiggle it crackles and eventually I can get it to a place where it will stay and oscillate.? However, any vibration on the bench causes it to move minutely but enough to cause it to stop oscillating again.? Would like to reflow the solder joints on that portion of the crystal socket but it's hellish difficult to access and don't think its possible to get the soldering iron in there.? Moving the pins worked well on the 40 and 15M xtals and they seem firmly in place and are oscillating their little hearts out.

I've got a brand new 20M xtal inbound and will try that out and in the meantime see if I can poke around from the bottom and see if I can find anything intermittent there.

Other than that, it works great.? Going to stick a themistor in the AC line input just on general principles.? The receiver appears to be quite sensitive although it doesn't receive anything but the signal generator here in the basement.? Would like to get this one put to bed, put it on the market and move on to the other 2B I have that apparently has a hellishly loud parasitic oscillation that appears as soon as it's turned on.

Thanks again and I'll keep you advised.

Sounds like the socket needs cleaning. Some DeOxit and a GUM Proxabrush
dental cleaning brush will do it.

Mike, W1NR

N4KZO de K8NN

Michael:

Well, sometimes it actually is a weak and intermittent crystal.? Received the new 18Mhz crystal from WD8DAS, popped it in and boy did it take right off.? Checked the negative voltage on the test point per the manual's instructions and it was at least three times (approx. -6vdc vs -2vdc) what I had been seeing with the other crystals and solid as a rock.? No drop out when pushing it around with a finger tip.? If I was going to keep this radio, I'd invest the $$$ and put all new crystals in it.

Touched up the alignment and now only have to put on the polarized line cord and add a thermistor and it can go back in the case and onto the auction block.? Will then move on to my other 2B which was butchered by a previous owner and will be a challenge just to get back to the original circuit, never mind repair.

Thanks for all your help and will probably be calling on you with the next one.

de K8NN

John,

Sometimes more is not better. ?As you are seeing a significantly larger grid-leak bias magnitude, this indicates the feedback voltage to the oscillator grid is larger, which indicates the output voltage of the oscillator is significantly greater. ?Mixing, at the first mixer, is occurring by modulating the input RF through transformer modulation (see L2), ?with the output of that amplitude modulated signal (two fundamentals, sum, and difference frequencies) capacitively coupled through C72, to the first mixer tube grid, V2. ?If modulation is not balanced properly so the modulation index is unity (100%), harmonic distortion will be produced, and linearity no longer exists. ?Not knowing the value of L2, but knowing that L2 never changes value indicates that L2 and C72 are not series resonant, and they shouldn’t be with this signal varying in frequency. ?So the input is not resonant, but the output of V2 is resonant. ?This resonant tank output may attenuate undesired harmonics, but if harmonics are produced, signal power is lost in the harmonics, taking desired signal power from the desired difference sideband signal. ?This would reduce the sensitivity of the receiver. ?Intermodulation distortion could also increase. ?Without more information and in-depth analysis, I can’t predict circuit performance of the first mixer using the new crystal and -0.6 V grid-leak bias.

I suggest you perform an input sensitivity test. ?Compare?the new crystal in circuit with the intermittent one, that is much closer to Drakes specified value of -0.1 V (if you can keep it oscillating). ?With the much increased grid-leak bias associated with the new crystal, I expect this crystal is either operating at series fundamental, not third overtone, or the equivalent series resistance is much, much lower than an original crystal design. ?Beyond this test, how does the radio sound operating on 20 M, comparing use of the two oscillating crystals?

73, Michael Smith

John,

Additional thoughts concerning my reply yesterday:

Concerning the greater output of the oscillator with the new crystal, rather than a fundamental-optimized crystal, I expect that this new crystal is a series overtone, higher than the third; this would supply a lower impedance at resonance to the cathode circuit of the oscillator, and decrease negative feedback increasing the output of the oscillator. ?The manual does not specify what series overtone is used. ?I have received information from others that Drake used the third overtone for optimal design.

An oscillator input voltage higher than expected for optimal design could, actually most probably would, create an under-modulated AM signal: the difference frequency (sideband) voltage would be less than optimal. ?The (higher) oscillator frequency is being modulated by the RF signal (lower) frequency. ?With the Preselector circuit of the RF amplifier peaked and the RF signal power constant to the inductive-coupled coils (L2), an increase of the oscillator injection signal would create under-modulation and decrease the difference frequency (lower sideband). ?Whether over-modulated or under-modulated, the result would be the same: decreased receiver sensitivity. ?Unity (100%) modulation at the first mixer will provide the best receive-sensitivity performance. ?You could use a high-impedance oscilloscope to inspect the AM signal modulation on the input to V2 (Pin 2) if the scope has enough sensitivity, and the output of the first mixer, convenient at V3 (Pin 7).

Comparing the receive sensitivity using the new and old 18 MHz series-resonant crystals will quantify the effect of different oscillator injection levels, at a difference of |-0.6| - |-0.2| VDC. ?How the receiver sounds during normal operation will qualify performance.

73, Michael Smith

John,

Correction concerning series resonant overtone specification: the 2-B manual clearly specifies third overtone for all crystal frequencies at and above 18.0 MHz on page 11.


73, Michael Smith