The AM Forum

Ok new problem.. Radio motorboats.. Any Gonset Experts out there??

When squelch is off turned fully CW and I turn up volume at about 1/2 way its motorboats.. Then if I keep turning the volume up it gets loader and faster.. Have checked tubes and replaced as necessary.. Checked the resistors around the squelch circuit.. replaced 2 .. But still same .. Checked the 6AL5 voltages

Pin 2 Plate 99.3 volts no squelch 49.6 volts squelched
Pin 5 Cathode 108.9 volts no squelch 98.4 volts squelched

Other side ( other diode )

Pin 7 0 volts both ways
Pin 1 0 volts both ways

My B+ voltage is 259 Volts

Help

Skip...

have you recapped?

Yes.. There were only about 2 or 3 besides the electrolytic and the bypass caps..

Skip

this is not specific to a G50 ...motorboating audio stages usualy happen when one of the power suppy decoupling electrolytics or other large value dry out ... you might try bridging an additional 10 uF or so to see if there is any improvement ...

other things are resistor value changes and corroded ground connections ...

Ok will give it a try..

Thanks
Skip...

I also have a G50. Some years ago, I redid the RF and modulator, and a power supply recap. The receiver worked OK so I kept that stock. Some years later I noticed the same described motorboating around the squelch threshold. later on I could never break the squelch. The last few times I used the G50, I used it with an external receiver.

I have not spent any time researching the circuitry. Anyone looking for my G50? If so, email me.

Jim
Wd5jKO

Ok, Well I did bridge an additional 10 uF or so to see if there is any improvement.. There was no improvement..

Still motorboats.. I can get volume about 1/4 way up ( very low volume) and no motorboating.. But as I go farther it starts..

When I inject a signal via Signal Generator.. I can get it up to around 1/2 way then can hear motorboating thru the signal..

I have replaced the tubes 6AV6,6AL5,6U8,6BE6,12AX7 Etc..  with no change..

Skip

Can you post the schematic diagram?

I found the schematic diagram, so no problem. It seems to me that the second cap in the PS (voltage CC) is not sufficient and critical, Can't read the value nor the number. This because the problem is frequent, so a design flaw.  I should add an other RC to the anode resistor R34 of the 6AV6 (V2) Coming from the HV cc  10 - 47 kOhm than  10 uF to ground and  to R34, the 470 k. anode resistor
You can try add an other cap to the VCC across the second electrolytic cap, but that should be a big one, not 10 uF
I use a lot the small caps out of the trow-away camaras con flash, they are in general approx 100 uF, 330VDC and are very small.

My HQ170 has a similar issue. It motorboats when the band width is in the 2 KHZ position but is OK otherwise.

Any ideas?

Thanks, Rich

Try unsodering the 10 uF cap and tacking something bigger in its place....

KLC

Here is my schematic.. PA0NVD The parts you list are not on my schematic.. Attached is one close to mine.. Only difference is mine uses a 6BE6 tube instead of a 6BH6 tube for V9.. This schematic is much clearer..

Skip 

Thanks for the diagram
I did look into it, and to my opinion the cap C! of 80 uF in the HV filter is quite marginal in relation to the load. Especially because there is no filtering additional at all for the HV to the AF pre-amp and the IF amps
I stick to that advice to add an extra AF filter in the power line to the V2 6AV6 anode, so in series with R34 of 470 K
Add e.g 47 k coming from EE and add 10 uF to ground between this resistor and R34 of 470 k
If this does not help, the feedback is entering to the HV variation of the IF power, BB or AA
As said before, you can also try to make C1 bigger, but substantial bigger, e.g. a few hundred uF But I consider that not a sound solution, adding extra filtering should have been done in the design phase, like a R in series with the HV feeding AA, BB and EE and after this resistor an Electrolytic to ground.
succes

Ok, I feel stupid.. Motorboating gone.. This is what I did.. When I was recapping I got the 80uf and the 30 UF Electrolytic swapped.. Meaning the output of T2 ( B+) was a 30 UF instead of the 80 it should have been..

When I changed it to way it should be motorboating stopped..

Next I will see if I can see a signal..

Skip..

That's a good message Skip
Glad to hear it's gone. But that means that indeed the value of the cap is critical. Should be better to add an other 100uF or so to prevent feedback in more extreme situations..
Nico

Skip

No need to feel stupid,  at least you didn't wire them in backwards.  In fact, folks who have wired their electrolytics in backwards have been banned from the Forum.

So. you're still in good standing.

Fred

 In fact, folks who have wired their electrolytics in backwards have been banned from the Forum.

 :'( :'(

Well, I haven't done that yet...

But I took your advise and increased the 80uf cal.. I had 100 uf and put that in..

Now on to RX issues.. Started alignment the 455 Khz went fine as did the 2.3 mhz IF's.

injected a 51 mhz signal from generator was able to hear it. Then after about 5 minutes signal disappeared..

So now I will have to see whats going on.. Something heating up??

Skip

Do you do hear noise? The same as before? The IF at 2.3 MHz still works ok?
Try first to change the 6U8 osc mixer. I saw various oscillators that did cease after a number of years.

Yup Both IF frequencies are OK.. RX noise even when signal quits.. Has developed a new problem..  When signal generator is modulated I cannot hear tone.. The Signal is there but no tone.. Both the IF 2.3 and 455 I can hear the tone..

I will try the 6u8

thanks
Skip

In the diagram I only see 1 IF at 2.3 MHz, no 455 kHz
If you hear a tone when you drive at the IF, the IF strip is ok. I think that you are not at the correct frequency when you hear no modulation. The signal generator does give a signal that does diminish the IF or RF gain, reason why you hear something, but the mixed signal is not at 2.3 MHz. Sweep your RF generator over a wider range, the oscillator of the receiver may be set wrong. I suppose that the receiver oscillator is below the RF frequency, don't have info about that. If so, try to measure the oscillator frequency, add 2.3 MHz and try again. If the oscillator is above the RF frequency subtract 2.3 MHz and try again.
A counter may be not sufficiently sensitive to pickup the oscillator frequency without excessive coupling which may pull the oscillator frequency, so checking the oscillator with a receiver may be better (if you have a receiver covering that frequency range)
Nico

Ok there are 2 schematics floating around  One uses 2.3 mhz as both and 2nd uses 2.3 mhz and 455 Khz.. Thats the one my radio is.. The other difference is mine uses a 6BE6 as the RX 455kc IF.. The other schematic uses a 6BH6 for the 2.3 mhz IF..

I will post mine  its a little hard to read values..

Skip

Ok, I see. Except a second mixer and IF, the rest is the same. I still think hat you are at the wrong frequency. Especially when the RF generator signal is a little strong and with the RF amp, it is quite easy to get signals that do limit the RF amp or the first mixer or first IF without passing the second IF. Than the noise goes down like you receive a signal, but it is out of band for the second IF.
Can you measure the first oscillator frequency?
If not, a shortcut may be to tune the receiver at e.g. 51 MHz, put the signal generator at 51 MHz, approx 100 uV and turn the core of the first oscillator coil until you hear the modulated signal.
Nico

I think the Osc is OK well maybe.. Here is what I found when a wiggle either or both of the tubes 6BE6 and 6u8 I can get good signal and tone if in right position..so I think either both sockets or maybe solder joints.. tubes are not making good contact.. once I find the spots on the tubes it works for a little bit then I have to try to find spots again..

PS I did clean sockets..  Also tried different 6u8 and 6be6.. with same results  wiggle right and it works..

Maybe I will try to tighten the pins a little?

Skip.. 

So you found it! congrats.
Indeed check if the sockets are tight and not bend open.
When I have that problem, I normally scratch the tube pins a little, put some contact cleaner and insert and retract the tube a number of times to clean the sockets. Scratch pins are a little rough and do the job.
I like that set Skip, when aligned good, should be a fine set for 6 meters AM Unfortunately I don't have an AM set for 6 meters other than a Alinco DX70. But that doesn't glow in the dark like yours does :-\

Well, scratching pins and lubricating didn't do anything.. Probably did a little worse.. I cannot find a spot that will work perfectly now  ..I can get somewhat of a signal but very scratchy.. 

Skip

That's weird, so resolder the sockets again may help. When moving the tube does change things, it must be a bad contact in that area. I don't know what kind of resistors the set uses, I had once a carbon composite resistor that had a bad contact inside the body. Took me quite a while to find.

Ok re soldered connections..That seemed to help alot.. Went thru alignment procedure with no problems.. Can receive signals from SG with good tone..

So that is all good.. However Sensitivity is extremely poor.. Says a 50 UV signal should read S9 .. I can barely hear a 50 uv signal. If at all .. S meter doesnt seem  to work right either.. Will not zero.. reading at about S2.5  No signal..

Skip

PS so making progress .. one step at a time..

Hi Skip
If you hear noise, I suppose the problem is in the RF amp or first mixer. Noise means sensitivity (in general)
Due to the experience with the solder problems of the other tube, start with re-flowing the solder at the RF tube socket.
If that doesn't solve the problem, try to measure the following:
Try fist to inject 2.3 MHz via a small cap (approx 10 pF) to the grid of the first mixer, but you have to de-solder C47, the 47 pF
Align the IF for max signal.
Than change the frequency to e.g. 51 MHz, tune the receiver and define the sensitivity again. Shouldn't be very much different.
Than re- solder C47 and change the signal injector cap to  1 pF. The sensitivity should not change much,  (re-allign L7, the anode coil of the RF amp for max signal).
Than change the signal via the 1pF to the grid of the RF amp, re-allign L7 and T9 and define the sensitivity again. Should be at least   3 - 4 times more sensitive.
I am curious about these figures.
They are NON-impedance matched figures, so after impedance matching (going to a normal 50 Ohm input) the last measurement should improve at least 10 dB
Regards
Nico

Well here we go.. Followed your instructions.. When all was said and done  Had approx 40 db difference..

step  1       Very  little difference
Step 2         little bit better may 1 DB difference

Step 3.       Approx 40 db difference  Now could hear 50 Uv signal at around S5..  Could actually hear a 1uv signal but very weak..

 So big improvement..

So is the alignment procedure from factory wrong??

Ok s meter is still not zeroing.. in fact when tuning from one end of band to other the s meter moves probably moves 2 s units.  Zero pot is at the stop..

Skip 

Hi Skip
Good to hear the improvement. I don't think that the factory adjustments are wrong, but you have to take into account that value's of caps and resistors may have changed. In addition, there is interaction, especially with band filters, so many procedures have to be repeated. A problem are the magnetically or capacitive coupled band filter. When the Q x k, is less than one, you can peak at the signal. But when critically coupled (ideal) this becomes difficult and when over coupled, there are two peaks. So bandfilters should ideally be aligned with a spectrum analyser and tracking generator or with a wobbling system. But to keep it simple, a repeated peaking at a weak signal in general will do.
So, in order to avoid drift during tuning, enter a small, just audible signal at 2.3 MHz at the grid of the first mixer pin 2 of V8 via a few pF. You don't have to un-solder anything. Now peak both IF strips repeatedly. Keep the signal just audible or measure the detected signal with a voltmeter or scope and just adjust for max signal.
Next is the adjustment of the oscillator coil T10 and the trimmer C52
--Tune to the lowest frequency of the scale and adjust the core in T10 for reception.
--Tune to the highest point and see if the scale is correct. If the receiver receives higher frequencies that the scale suggest, tune C52 for a larger cap until you receive the signal that the scale indicates and OVERSHOOT that alignment just a little
When you receive a lower frequency than the scale indicates, tune C52 for a smaller value and overshoot the adjustment just a little
--Return to the lowest frequency of the scale and readjust L10 again for reception.
This procedure has to be repeated several times until the scale indicates the correct reception frequency. During the last adjustments you should not overshoot the adjustment anymore.
If all ok, the alignment should be acceptable over the whole scale.
Now you can adjust the two RF coils T9 and L7 for max signal in the middle of the band and you should be ready.
When you tune across the band with a dummy attached, the noise  and sensitivity should not vary more than half an S point or less.
When not, tell me and I try to make a procedure to correct the problem.
Success Skip
Nico

Hi Rich
Did the above conversation solve your problem with the HQ170 ?

Ok followed your procedure.. All went well.. Looks like set is more sensitive at low end of band since AM activity is at 50.4 should be good to go..I would like to get S-meter to zero. and get it adjusted for 50uv=S9..

What I have read was the voltage divider circuit is R40/R41  Pot (5K) and R57  4.7K 1W..

After zero if reading does not equal S9 then either increase or decrease R44 ( 33K) For S9 reading..

Skip

That's indeed so Skip, the pot R20 (5k) is for zeroing the meter and R44 (33k) is the sensitivity
The meter depends on the anode current of the 6BH6 affected by the AVC. So the complete alignment, the amplification and the quality of the 6BH6 all affects the meter reading.
So when you change a tube, the meter reading may be off again
Please let me know when you had the first QSO with the set!!  I did sent an Email upwards to ask for good conditions at 6  ;D
Nico
nicomaria1945@gmail.com

Well, On to the Transmitter.. I guess I am still not done yet.. The spot function works well. However when trying to load I get NO Plate current at all..So of course NO dip..

Skip...

l

Seems that the set keeps you out of the bar Skip So this set will keep your wife happy  ;)
I suppose that the 6146 glows normal The tests below suppose a working 6146
When the spot function works, that means that the oscillator runs fine. If the driver Q2 works, you should get grid current. except if there is no cathode connection or if the 6146 gridleak R10 is open. Drive to a grid with an open grid leak will result in a tube cut-off and hardly any anode current.
Measure first the voltage at the anode during transmission, the choke may be bad.
If there is no grid current, measure the value of R10, 15 k, 1 Watt grid leak for the 6146
If there is grid current, than:
In order to exclude a meter circuit fault, please measure the voltage at the cathode of the 6146 during transmission. Work fast and keep an eye on the 6146 that it does not get red hot or so. The voltage should be something like half a volt.
If no voltage, measure the voltage of G2, pin 3, should be HV. If no voltage, measure at the other side of R15, he G2 resistor of 18k, 2 Watt. That resistor may be toasted. If so, please change C23 first, that may be the cause of an R15 failure.
Hear the results
Regards
Nico

Ok, Prelimary readings.. Resistance

R15  Should be 18K  Reads 19K
R10  Shoulod be 15K Reads 17K

R11 Should be 470 Ohms  reads 500 Ohms
R12 Should be 1.5K Reads 1.665K

But here is the kicker.  From Pin 1 to Pin 3 there is a 6.8 Meg resistor  Not on schematic.. Reads 7.8 Meg

Nothing shorted to Ground.. Did not measure all the .01 or .02 Uf caps yet..

Will power up and measure Voltages..

Oh it has a 6146 tube in  Not a A or B version.. If that makes a difference..

Skip..

There is little difference between 6146, A or B. The B version is especially made for VHF and may be preferred. But I doubt if you notice a difference
I don't think that you have to replace a resistor, all sufficiently close. That 6.8 M is not that strange, goes from G2 to the cathode. It seems to me that the open cathode at reception may have caused problems and they pull-up the cathode softly to keep the tube completely cut-off during reception. Not uncommon practice, a floating electrode is not a good defined thing. I should keep it like it is.
So seems that there may be no HV, bad anode choke, or one of the decoupling caps is bad. Other possibility is that the relay contact in the cathode S1 doesn't make contact. If so, you will measure quite a high positive voltage at the cathode at transmission, around a +50 V or more

Ok.. Here is what I measure..

6146  B+  RX  230 Volts
6146  B+  TX  328 Volts

Cathode  RX  197 Volts
Cathode TX   .4 Volts

Also measured pin 5 of 6146 with scope  Had signal in TX..

But still no grid or plate current indicated ..

Skip

meter switching and current shunts

That's why people should always use non polarized caps! Be very careful less you be Shanghaied to QRZ.COM..........  Oh NO the shame :P :P :P

Indeed what W4BFS said, meter circuit. When you measure 0,4 V at the cathode, the 6146 pulls 40 mA
So I suppose that the ground connection of the switching circuit is open at S5 because you don't measure Ia nor Ig or mod. If only R16 of 2k2 was bad, you should measure Ig

Well, Checked out meter circuits all is good. ground connection is good ..I guess I will try a alignment.. Maybe its so out of wack that there is not enough drive to do anything..

Cannot think of anything else to do at this point..

Skip

Must be in the meter circuit Skip. When there is 0.4 Volts over the cathode resistor R14, there is running 40 mA. With the meter series resistance of 2.2 k, the meter should show 0,4/2.2=0.18 mA (approx. ) that is almost 20% FS
Does the relay  S1 below the meter make good contact in Xmt? Does the meter show when in MOD and you modulate?

You can put 2.2Volts from a low voltage power supply at the left side of the 2k2 series resistor R16. Than there is running 220 mA through R14, the cathode resistor, transmitter off, not powered at all. In Ia the meter should show FS. If not, go fault isolating from there.
For Ig, you can put 1,5 VDC NEGATIVE over the 470 Ohm resistor R11 at the connection point of R10, R11 and R12
In the switch position IG the meter should show FS

I have not tried to  modulate.. I will have to rig up a mic to try that.. But will try to see if meter circuit is good by doing the voltage trick..

Thanks

Skip

The Xmt relay should be energized in order to switch over the contacts below the meter. Or you can push it over by hand

OK.. here are measurements with power supply..

Grid meter  -.66 volts  Reading 3.5 on scale
               -1.772 volts  Reading 9

Plate meter..   .646 Volts 2.5 on scale
                    2.5 volts    9

Ok so meter circuits working..

There is no relay.. All connections are made thru a  12 pole 2 position 3 section switch.. you manually go from RX to TX via the switch  So to take readings switch had to be in TX position..

I  think I need to work on the meter Zero.  When in RX the meter reads about a 3.. I cannot zero it using the Zero pot..  Probably has nothing to do with TX  .. I will see what readings I get on Grid position and plate when powered up..



Skip...   

Hi Skip
Can you replace the meter temporarily with an external meter of 1 mA to exclude meter faults? It seem that the meter reads quite low, perhaps the zero is completely off..

OK,, All is working fine now.. This is what I did..

UPDATE::::I was going to follow the alignment procedure I have for TX  So firts thing it said was to tack solder a 680 Ohm resistor across the coil L4.. So I did that..  Then it said  to hook up a VTVM from Junction of R10/R11  and adjust L5 for MAX.. So when I went into TX I all of sudden had Plate current about and 8 on meter..But no indication on VTVM.. So I decided to disconnect resistor.. Then went to TX and still had Plate current  Was able to dip and peak controls to get power out put on my Wattmeter..  I didn't have the right slug for bird meter so I am not sure what power is.. But its there..

So maybe a bad solder connection???

I will hook up the right meter and try again in a few to see if its still working..

Skip...

Well, I messed it up again.. I was trying to get more power out by adjusting L2 and L3 for Grid current.. got about 15 watts output  Then I turned it off put it right side up and powered it up again.. Now I get NOTHING.. NO RX and NO TX.. Have no clue what I did to make it stop Receiving..  Only thing I touched was L2 and L3 ..

Not happy with myself I should have left well enough alone..

Skip

   

Congratulations Skip, you for sure found the problem, bad solderings.
So the set has quite a few bad solderings. Perhaps it is good to just re-flow the solder joints. may be a metallurgical problem with a cleaning process or so. You may run in a few more bad solder joints. I think your new problem is just related to that. Take some time and reflow all solder joints till you see they solder is flowing well, perhaps add some flux.
I remember when I did work in a microwave lab, that the joints and flanges of 10 GHz waveguide were soldered. One day we had failures and we saw that the flanges just came off the waveguide, one by one, after a few years of use. The lab did investigate and did track back to an aggressive flux, that kept attacking the surface of the copper after soldering, somthing similar to S39. We had to clean and re-solder almost a hundred flanges.
When I don't have the means to measure power, I normally connect a small vacuum diode with the anode to the output and the cathode to 10.000- Pf to ground. The DC voltage over the cap is the peak-voltage of the RF. The RMS voltage is 0.707 x Vpeak. The power is RMS squared divided by 50 So easy to calculate the power (if the RF load is a good 50 Ohms). Is quite accurate.
73
Nico

Well, I hope its that.. We will see.. Will let you know if I get it  working again and what it was if I know..

Thanks for all your help.

Skip

I am sure you will fix the set Skip. Please take into account that these kind of failure are extremely difficult to find, they are not obvious. Re-flowing is the best option
Good luck and happy to help

I have been watching this thread evolve. Nice back and forth with good information, and good results forthcoming.

Concerning reflowing connections, I will add my 2 cents. If possible use solder wick or solder sucker to remove the old solder. Then remove residual flux with brush and solvent. I use 100% (IPA) alcohol at work, and the 80% stuff from a drug store at home. Then use new solder with some added flux to renew the connections.

The G50 modulator has serious design limitations that can be removed without too much trouble. The main issue is the Heising choke has too much current flowing through it, and the core is close to saturation. As a result audio below 500hz gets distorted, and full modulation is not possible. Gonset turned this into a Marketing "feature" describing this as a way to prevent over modulation!
See Reply 2 and 4 at this thread. I talk about my G50 modulation changes, and define design limitations:
http://amfone.net/Amforum/index.php?topic=34714.msg267903#msg267903

I look forward to more progress.

Jim
Wd5JKO

A good option for the choke saturation is finding a balanced AF output transformer. e.g. for two KT88 or EL 34 tubes or so.  The center tap to the HV, one side to the modulator tubes and the other side to the 6146. Than the DC currents will almost cancel and you can modulate with a substantial smaller iron without distortion. Because the iron can be smaller, should be easy to find a transformer that fits.

Ok, Re soldered a lot of connections.. Radio now receives and TX .. The only problem now is power output is lower than spec.. I get about 15 watts .. Spec says about 27 or 28 watts?  Could be my 6146B  tube  . I have a lot more I could sub along with some drivers 6AQ5..

So far so good..

Skip

That's a good message Skip, makes me Happy :)
If the grid current is normal and it tunes with a nice dip of approx -25%, indeed can be the 6146. If you are looking for a replacement, try to find the A version, is slightly more suited for VHF. It is the equivalent of the QE05/40

Yes, It tunes great..I do get grid current but didn't look at it for a value.. I put in a 6146W and get about 20 Watts out now.. I will look to see if I have a 6146A.. Not really concerned over 7 to 8 watts.. But would be nice to see what the differences are between tubes..

Thanks again for all your help.. Without it It would have taken me a lot more time to get this running..

Skip 

You  are more than welcome Skip. Whenever you plan a trip to Costa Rica, give me a PM

Ok , Found some more bad solder joints.. Mainly Ground connections..  so now all seems to be good..

Wired up a D-104 Microphone.. I now have modulation..

But its downward..Is that normal for this radio? or does it require special tuning?

I remember back in my CB days in early 60's..  had to do some special tuning to get upward Modulation..

Skip



 

No, should not be downward. Measure if the screen voltage is ok. See if the grid current is sufficient, low drive level may be a cause.  A too heavy load may be a reason as well, tune for a deeper dip until the power goes approx 10% down.
An other cause may be low emission of the tube. If you raise the filament to 7 - 7.5 Volts and the modulation improve (becomes positive), the tube doesn't have sufficient emission. Raising the filament should have just a minor improvement of both power and modulation. A substantial improvement indicates low emission, both for indirect and thoriated filaments.

  A scope looking at the modulated envelope would be most helpful. My last comment about the G50 modulator earlier shows that in my case, the Heising choke was saturating, and audio below 500 hz was progressively more and more distorted as the frequency goes down. Distortion worsened, and upward modulation was limited.

   Besides the good stuff already mentioned by PA0NVD, a simple test would be to load the D-104 cartridge with something like a 100K load. This will attenuate most of the lows, and if the audio Heising reactor were saturating, getting rid of the lows should improve things a lot. It would sound like a stock Apache though!

   Lightly loading the RF stage should also help with modulation capability.

   Funny discovery in my G-50. The modulators were original Genelax KT-77's instead of 6L6's. What a find!

Regards,
Jim
Wd5JKO

Hi Jim
Indeed nice find. I love the KT77, in UL mode these tube perform excellent with very low distortion.
The choke saturation can be solved by finding a balanced output transformer e.g. for 2 x KT88 or EL 34. There are plenty. Center tap to HV, one side to the modulator and the other side to the PA. Than the DC magnetization is almost canceled and you can modulate the transmitter with less iron and less distortion. Simple to try temporarily with the iron externally wired with clip leads.

  A picture of the output/modulation transformer, and two scope plots. The tone modulation is nice up to 80%, but from there the negative peak shows a little "tilt" toward the center. There is little difference at 80hz versus 1 Khz. For voice modulation, with a properly phased D-104 element, 100% positive peaks are obtainable.

  The schematic I posted earlier shows some changes to the 6146 screen modulation. I added a negative peak clipper to the screen only, and compensated the screen dropping resistor for less phase shift. With a plate modulated 6146, "hard" limiting the screen voltage to something like +10v minimum (diode clamp) seems to provide a "soft" downward limit to the plate side envelope pattern. Full modulation with voice, and NO "white Lines" on the scope.

Jim
Wd5JKO

    Skip,  I am not stealing your thread, but while waiting for your next report, I am adding a little info pertaining to my G50 modulator experiences. I attach three schematics for you and Nico to mull around. Perhaps there is a better way, or a simpler way to get good modulation with low distortion. The D-104 trick I mentioned to roll off the lows is as simple as it gets. That is a big bang for less than a buck.  ;)

Jim
Wd5JKO

Hi Jim
Thanks for the diagrams. As you may remember, I once made a small AM transmitter using the trick with the balanced output transformer instead of the straight Heising coil. We did discuss  that before. I indeed had some white lines with negative modulation peaks, I like your trick with the negative peak limiter at G2  :D  When my house is ready and I have my shack again, I will try that for sure.
I think the balanced output transformer instead of the Heising choke will give most of the improvement and allows more lows in your modulation.

Well, Here we go again.. While running radio everything was working perfect  Good RX and TX.. But then I heard a small popping sound and the Receiver went dead.. The audio level dropped a lot . So tried to inject signal from SG  and got Nothing.. So back to square one.. This is getting to be a project from Hell..

Underside showed no burnt or discolored parts all looks like it did.. No smell as in overheated parts ETC..  So Not sure where to start again..But I will start to look to see  if I can find any parts that my have opened or shorted..

PS The TX still works fine..  But no spotting as RX is dead..

Skip

 

Hi Skip
A small popping sound does point a lot of times to a shorted cap. Measure at all receiving tubes the anode voltage and the G2 voltage. Go from there.
I don't know if you did recap, but that can happen. Please don't get de-animated, you came a long way successfully!!!

My thoughts also.. I tested the tubes again and found my RF tube a 6BZ6 now shows a short between the plate and  screen.. Thats where I will start looking..

Skip..

That's weird, never saw that before.

OK, Will start to get back on this thing tomorrow.. I did a little tonight . Injected signal  pin 1 of V9  6BE6 455 Khz.. I could hear signal but had to raise Generator to about +6 dbm.. Normally this would be around -50 to -60 dbm or less..

The tube tests good..

Then tried V10 6BH6 pin 1 455 Khz.. Same results  Tube tests good.. 

Skip

V10 has anode and g2 voltage? And if you touch pin 1 (grid) of V11, the 6AV6, do you hear loud hum?

Ok here is what I get..

V10 Pin 5   +176 volts
      Pin 6   +151 volts

Pin 1 OF V11  touch=  Load Hum.. ( yes)

Skip

That looks ok, May be a problem with he squelch circuit driven by pot R54. Please put a 0.01 cap from pin 2 to pin 5 of the 6AL5 double diode. That should disable the squelch without affecting DC levels.
Lets go from these results.

Measure the voltages across the pot R54. One side shoud be approx 200VDC, the other side approx 80VDC
C65 at the 200VDC side may have shorted. that should give a dead receiver because the squelch is completely out.

Ok Receiver not completely dead.. I get noise with the volume control.. Not as load as before .. The squelch works  I can mute receiver..

When it was working before pop.. The s meter was reading about a 2.5..  Now its reading a 1.. And volume is a lot lower.. and I could not Receive the Spot function anymore.. So I said Receiver was dead..

Skip

Ok, that does change things. Lets go over the receive stage gains to find a low gain stage.. I did estimate some levels. Put a 10 pF in series with the signal generator and measure the following points. They should give approx the same signal strength +/- 1 - 2 S points Set the generator at 50% modulation.
V10, pin 1, 10 mV 455 kHz
V9, pin 7, 500 uV, 2.3 MHz
V8, pin 2, 25 uV, 2.3 MHz  and also 25 uV, 51 MHz
V7, pin 1, 3 uV, 51 MHz

Ok here is what I get..

V10, pin 1, 10 mV 455 kHz    4  Good tone
V9, pin 7, 500 uV, 2.3 MHz    0
V8, pin 2, 25 uV, 2.3 MHz  and also 25 uV, 51 MHz   0,0
V7, pin 1, 3 uV, 51 MHz   0

V9, pin 7, 50 mv, 455 KHz    4  Good tone
V9, pin 7, 158 mv, 2.3 MHz   4  NO TONE

Skip..

Seems that the 6BE6, V9, does not mix anymore, or very low level. measure with a scope at the cathode pin 2. There should at least be a few volts RF of the correct frequency, 2.3 MHz + or - 455 kHz (I do  not know if the oscillator is above or below the 2.3 MHz.)
First try an other 6BE6 V9
If no solucion
Pin 5 and 6 should have the same HV
When there is RF at the cathode, the oscillator is functioning, than the frequency of the oscillator is wrong. Most probable a bad soldering in the oscillator circuit.
Re-flow all connections in relation with the oscillator. If still the wrong frequency and very much off, the two tuning caps C30 and C64 may be off and should be measured, The value of these caps is important, otherwise there is no good tracking between oscillator and RF circuits when the frequency is adjusted.
If the frequency is just slighly off, you can try to change the core in L12, the oscillator.
BUT REMEMBER THE FORMER SETTING!!
When the oscillator is dead, measure from pin 1 to ground, should be 20 - 25 kOhm.
If ok, measure the 3 caps in the circuit, there may be one open or shorted
Keep my fingers crossed

Ok,, The Osc is working..It runs 2.3 Mhz below the Carrier  so at 51 MHZ Osc runs at 48.7 Mhz  I can see on spectrum ANA the osc signal It varies along with frequency..

But hear is the catcher.. I used to measure the OSC frequency by putting a Antenna on my Freq counter.. And probing the chassis by the 6u8  IT would read the Frequency.. But now I get no reading on counter.. So I hooked up spectrum ANA with antenna and probed.. I can see signal not real low.. But maybe not as strong as it was.. Not sure because I didn't do it that way before so I have no reference to go by..

Skip..

Hi Skip
That is the first oscillator. You need to measure the second oscillator of V9. That should run at 2,3-0.455=1.845 MHz or 2.3+0.455=2.755 MHz. You should measure at least a few volts at the cathode of V9, pin 2 using a scope. and a counter

Ok  Understand  I put a scope on the cathode..  Got nothing..  I will try again tomorrow to make sure I am doing it right..

Thanks
Skip

That makes sense, there is your problem. First try an other 6BE6

Ok, Here is what I did..

Put scope on pin 2 of 6BE6 Got nothing .. Put in a new 6BE6  same .. So started to check a few components.  Noticed on pin 7 schematic says 68 Ohm.. Mine was 100 Ohm  But measured 130 Ohm.. So replaced it with a 68 Ohm

Next The cap on Pin 1 schematic said 47 UUF  I thought..  It had a 5 uuf..  so maybe schematic should have said 4.7 uuf ??  Took it out measured and put it back in..resoldered all connections on tube socket and coil L12..

Fired it up  It works again.. Checked pin 2 again and got a nice sign wave about 1 volt PK to PK.. Re aligned  and now getting approx 7 on scale for 50 uv input.. Spot function works again..

For how long it will work?? Who Knows..  Will try again in morning to see if still working..Still curious to what the pop was..

Skip...

Hi Skip
I really don't think that C89 should be 4.7 pF, it should be 47 pF, 4.7 pF may not give you a reliable oscillation or no at all.at this low frequency. That may the a part of the problem. Soldering quality is for sure an other problem. Please try to change to 47 pF and measure again the RF voltage at the cathode. I expect more than just 1 volt.

Ok.. I will try that.. Have to see If I have a 47 PF Cap..

The one in there was tubular.. with colored bans.. looked almost like a resistor..

Colors Left to Right

Grn,Wht,Blk,Grn,Blk

Skip...

The value is not critical Skip, anything between 33 and 56 PF is ok.

Ok will give it a try right now..

Skip

When it works ok, you may have to adjust the oscillator coil L12. Put 2.3 MHz at point 2 of V8 via a few pF, just audible and adjust L12 for max signal.

Yup did that.. Seems to be working so far.. Noticed that L12 peaked a little broader than before.. But signal strength is the same.
  I used a 33 Pf in parallel with the 5 pf  to give me 38 PF..

So far so good.. I said that before.. we will see when I try to put it in the case..

The Transmitter still puts out about 20 watts..

Skip

Putting it into the case is when problems start, Murphy is still alive!!!. But no problem, as soon as you get it out of the case, it should work fine again to prevent that you find what was wrong... :)

Ok, Stopped working again.. Still have a intermittent connection somewhere in the 2.3 Mhz IF's  If I tap the IF case it makes all kind of noise and squeals ETC..   I think I will take them out and inspect the insides.. Maybe the Caps are bad or intermittent inside cans..

Oh goody did not  even have a chance to get it into case before it stopped working this time..

Skip

you are lucky  ??? If you should have put it into the case, you should have to take it out again.....
Don't know what they used as a flux during soldering, but it was really bad Skip. Did't hear so many problems wit bad soldering s. May be an other one?

Yup, Could be.. Well I am pretty sure its around the 2.3 Mhz IF cans.. T4 and T5..

So I will remove the cans.. I did that before and could not find anything that looked bad.. I will have to do a better inspection this time..

Skip..

Ok, Removed the coils  All looks good could not find any problems.. However I was tuning one of the slugs and it broke. I cannot seem to get it out .. Anyone know of a way to get this out without destroying the coil?

Skip.. 

That is the nightmare of the restaurateur...I went through that several times.
If it is mounted with wax (many are), heating will sometime help. When I can't get one out, I break pieces off and put a second slug on top to compensate for the loss of material.

OK Finally was able to get back to this.. Reinstalled the 2.3 Mhz IF Coils.. Could not find anything wrong with them..

Realigned radio.. Radio now works as before.. I did fix the S-Meter circuit that I had not gotten to before ..

Now a 50 uv signal =9 on meter..  Reinstalled the front panel.. Radio is still working..

Now to put it in case.. going out to garage to do that now..Wish me luck..

Skip

Ok back in case.. Still working  Did a drop test a couple of times  No problems.. TX still at 20 watts or so.. Did notice the in VFO mode it puts out about 2-3 watts less power than in Xtal mode..

No big deal..

So on to the next..

Thanks to all for helping me get this up and running.. Especially Nico..

Skip

   Now that's interesting; mine puts out a few watts more in VFO mode than in crystal.  Go figure.  What's more, when in crystal mode my power output will do a slow drift, maybe three or four watts or so, slowly falling off and then returning just as slowly.  I haven't tried changing out the 6AU6 yet, but I think I will.  Power output drifts very slightly in VFO mode.  I'm getting 30+ watts from mine most of the time.

Wonder why its different.. Mine seems to have more Grid current in Xtal mode  Maybe that's why..

Mine only loads to about 5 in plate position both Xtal and VFO..

Mine is rock solid no drift..

Skip

    I'm thinking the slow drift in power output must be heat related - the change takes place over a span of 30 seconds or so.  Something intermittent like cold solder or a cracked carbon-comp would be all herky-jerky spastic, while this is slow as molasses.

Hi Skip
Fine to hear that all your wrk had good results, congrats!!!
Bill, doe the grid current and anode current also drift? 

    Haven't checked grid and plate currents at the oscillator yet, but at the PA they do show a slight change, in sync with the drift in power output.  The plate voltage at the 6146 drops with increasing power output and rises with decreasing power output, so that appears to be effect rather than cause.

The anode voltage rises probably due to the decrease in anode current. See is the grid current also goes down when the power goes down. If it does, the problem should be the driver or oscillator stage. If not, check the G2 voltage. If stable, put a 10 000 pF ceramic from cathode to ground. May be the original cap is weak.

Concerning adding the push-pull transformer, what might be the impedance that is needed, or maybe it doesn't matter that much. I'm not sure how impedance should be calculated in that configuration.
Thanks...........Larry

 Hi, Larry,
   There are people here far more knowledgeable than my humble self, but my take is that since a Heising configuration - center tapped or otherwise - doesn't offer any impedance transformation between the modulator and the plate circuit, it's not something we'd have to worry about with this mod.  The Hammond 1645 that I'm going to be using for my G-50 has a 5K ohm primary, center tapped, so you'd have 2.5K on either side of the tap.  The secondary has taps for 4, 8, and 16 ohms, so it'll match your speaker.
    As a side note, at the recent Sussex hamfest I picked up a nice Clegg 99'er, and my friend picked up beautiful little Utica 650.  At the time I didn't know much about the workings of either rig, but a check of the schematics revealed that both rigs use the very same modulation scheme we're talking about using in our Gonset mod: a center tapped Heising choke.  I'm in the midst of a two-part vacation, so I haven't had time to play with my 99'er, but I'm looking forward to finding out how that modulation scheme sounds, and I really can't wait to hear my friend's Utica 650 on the air - that is one sweet looking little rig.