The AM Forum

As noted in my "Christmas Twins" thread, the A4 is becoming a real had scratcher.
The biggest problem I am having is instability. Example: While on any band, I tune into a decent signal
and it just dances around like a drunk chick at a Grateful Dead show. This happened before and after my re-cap job, so I reasoned I am not the reason it occurs. To give you a better idea of what it sounds like, imagine the musical saw sounds of a bad B-rated horror movie while using the Calibration signal.
I have targeted a possible problem that I mentioned earlier, of the 175 vdc on the mechanical filter, occurring on the side of the shield after V5 second mixer. There is supposed to be a blocking cap (1000) C-144, but it was never there. I tried 1000pf and .001 and all it did was attenuate the hell out of the signal.
At this point I am starting to realize my limitations of receiver understanding. I know this radio has the potential to be a star performer, but right now it sounds awful, and not very sensitive either.
I have not started the Mica replacement quest either, as it will take a few hours to accomplish. I am awaiting the sage advice and direction from those who know better than I.
I will supply some photos to support my descriptions here. All of the original paper caps were replaced, along with the electrolytics.

Here are some pics of the area of V-5 & V-6.

If anyone has any pictures they can share, this would help greatly.
Thank you.

Sounds like the PTO.  Sometimes the lubricating grease hardens and causes the cam on the slug thingy to ride flakily over the linearity guide.  Requires breaking the seal and working on the inside of the unit. See the articles in ER or in the compendium amongst available sources.

To rule out the crystal oscillator, try 160m, since the rcvr is single conversion on that band and no xtal is used.

Don
Thank you. On 160, the calibrator still jumps all over the place. Tuning into distant heterodyne's  confirms that the instability exists there too.

crystal filter has nothing to do with frequency stability.
Sounds like you have a bad cap in the pto. you might get lucky and it is just a bad tube in the pto.

Sorry, I misused the term. I know you meant Oscillator.
I replaced the PTO tubes and swapped out the mixer and IF tubes, same result.
I tune into a loud station on 40M for example, and it sounds like I am jogging the PTO back and forth randomly.
The task of dealing with the PTO disassembly doesn't bother me. Ughhhh.

Don't forget the passband tuning. It's mechanically linked to the PTO. Mine flakes out from time to time, needs a good going over.

Thanks Todd.
I just now replaced the mica bypass caps on the SSB/CW detector (V-11) and the transformer grid feed cap into pin 2 of that tube. It improved the audio somewhat, but no real improvement on the "wiggle".
Before I commence taking the PTO apart, can anyone suggest something else?

Before you disassemble the PTO,  make sure the ground strap (wiper) where the shaft enters the body is clean and has tension on the shaft.

Yes, exactly. Before opening up the PTO...The PTO shaft needs to be clean and the wiper needs to press against the shaft at all times. See if that fixes the problem. Also check the DC voltages at the PTO tube socket to make sure they are stable.

Check!Check!

Are there any molded mics'a within the transformer cans?

Why not find the problem and fix it rather thqn shot gunning guesses.

Not sure what the problem is.

One way to tell for sure whether or not it is the PTO is to monitor the oscillator on another receiver (one with a BFO or CW/SSB mode).  The pto operates at 455 kc/s higher in frequency than the 75A4 dial readout on 160m, and has high enough output that you should be able to hear it with another receiver on the bench,  with a short wire connected to the antenna terminal and the other end poking inside the A4 cabinet.  Try poking the wire between the oscillator tube and shield if you can't find the signal with the wire just in the vicinity of the oscillator. Once you find the oscillation on the other receiver, turn on the BFO in the second receiver and get a heterodyne. If you notice the same wobble in frequency on the received oscillator signal, then the PTO is changing frequency.  

I assume you notice this shift in the CW/SSB mode, and you are hearing the beat heterodyne shifting about. If the main oscillator seems clean, look at the BFO. That could be the culprit, and the BFO is a lot easier to work on than the PTO. If you are hearing this shift in the AM mode, the PTO jump would have to be extremely severe to affect the signal as received through the pass-band of the i.f. filter.

I'm also assuming you  have eliminated a flaky problem in the  crystal controlled 1st oscillator circuit.

Don
Thank you. I will try that. When you refer to the "BFO" on the 74A-4, I assume you are referring to the PBT control. I'll hook up my HQ-170 and loosely couple an antenna close by. The thing is, no matter what mode or band, the problem is occurring, and with the crystal calibrator on, I can zero beat and it shifts right along with the RX signal.
Should I be concerned with the 175 vdc on the mechanical filter? If someone had a photo of a stock 75A-4 showing the area of V5 & V6, it would clear up some things for me because C-144 is missing and apparently never was installed, or was removed at some point for a mod. When I tried putting a .001 in place I lost a ton of gain.

I don't have the schematic right here in front of me, but one of the first modifications you want to make  to an early run 75A-4 is to take the +DC plate voltage off the mechanical filters.  A simple tube or component failure could place the full plate voltage across the input side of the filter and likely destroy it.  Collins later made the modification on their production runs.  Feed the DC to the plate of the tube through an rf choke, and couple the plate to the input of the filter through a disc ceramic blocking capacitor.  I use a 1000 volt rated capacitor on mine - in fact, two in series, just in case one ever shorts out. The bottom end of the filter is removed from the +HV and as I recall, wired directly to ground.

That mod is in the ER compendium, and no doubt can be found on line. Instead of a regular rf choke, in mine I used the coil from the remains of an old Heathkit Q-multiplier, which as I recall is self resonant in the vicinity of 455 kc/s.  It may have a slug to adjust the inductance, but I don't recall using a fixed cap across it.  I'll have to look next time I open up the receiver; don't think I documented and recorded the details of the mod. A single coil and cap from the primary of an old 455 kc i.f. transformer could be made to work.  The idea is to use a coil that is broadly resonant at the i.f., in order to get more gain than what you would get from a simple rf choke. I believe Collins just used the choke. The choice of choke may make a difference in any loss of gain you might experience.

That DC voltage on the mechanical filter could destroy the filter if something broke down and shorted to ground, but I don't believe it would have anything to do with the receiver jumping about in frequency.

Here is the schematic of the area that is troubling me as it pertains to the filter.
Mine isn't wired in this manner, and the LC network feeding the plate of V5 and the wiper of the switch is missing entirely. Mine must be an early version before the schematic that I have. Sorry for the low res picture.

I ordered the entire compendium today. This ought to help clear up some mysteries contained in my receiver, which is serial numbered 1621. Here is a photo I found of one that has been modified to match the schematic drawing that I have. I just may have the molded inductor too. :)

Yep, that's what you have. C144 (1000pf) blocks the DC off the filter, and the bottom end of the filters go directly to ground. C145 and L32 form a resonant circuit at 455 kc/s, to eliminate the loss in gain that a simple rf choke would cause. In yours the plate goes directly to the filter and the bottom end goes to the B+ line. This is a disaster waiting to happen.  If the tube plate shorts to ground or somehow the wiring to the switch happens to get grounded out, the full B+ is directly applied right across each filter as it is switched in.  You could  wipe out all 3 filters in one shot!  Even if the tube doesn't short out, who knows how well the insulation inside the 50 y.o. filters will continue to hold up to the B+ voltage.

You need to immediately re-wire the circuit according to your schematic.  Use the best quality capacitor you can find for C144, and use something rated for several times the actual B+ voltage, since if that capacitor shorts out, the full B+ will appear across the input side of the filters.  Making this mod and using a poor quality capacitor is even worse than leaving the original circuit intact.  In mine, I use a couple of 2000 pf capacitors in series, each rated at 1 kv, for C144.  That way, if one of the capacitors were to short out, the other would serve as a back-up and save the filters from destruction.

C145 and  L2 form the tuned circuit.  Due to variations in components, just any 2 mh and 62 pf capacitor may not hit exact resonance. This would not only cause signal loss; it would detune the input coil in the filter, resulting in excessive pass-band ripple. If you can find a physically small mica trimmer, about 75 to 100 pf maximum, use that for C145.  That will allow you to adjust the input circuit for exact resonance. Tune for the best  compromise that results in the least overall ripple in all three filters.  Getting one filter perfect may result in more ripple in others, so a compromise may be necessary.  Collins decided to save on costs by using a fixed 100pf mica (C61, C130, C132 and C64, C131, C133), instead of a variable trimmer, across each end of each filter for resonance.  This makes it impossible to adjust the input and output of each filter for optimum performance. The filters still perform pretty well anyway, but this might be upset when you add C145 and L2 if they don't happen to resonate precisely at the i.f. frequency.

Don, is your receiver still using the 6BA7 as V5?

I completed the changeover, not permanent but such that component swapping is easy. The gain is way over the top though and as you indicated Don, it has nothing to do with the warble and shift of the signal. The front end is very hot now and I have to use judicious amounts of RF gain reduction to make listenable. I imagine there are ways to combat this but at least I have successfully removed the high voltage from the filter.
There must be a pad ahead of this stage.
Onward.

The A4 PTO is notorious for defective ceramic caps and this has been documented several places on the web. Changing will affect calibration and you may have to remove the PTO a few times to get it on. It can be a PITA job and one I dont relish doing.

The ground wiper as mentioned is an important first step as is being sure its not a tube problem.

Have you tried a new PTO tube, V-15? You have 3 available right in the radio.

Carl

Carl
Thanks for the reply. Unfortunately, I am up for the task but don't look forward to it if I have to.
All tubes have been swapped with no change in the problem. I haven't ruled out a bad cap or resistor, but as soon as the radio comes on from a cold start, this behavior starts as the radio settles in. This doesn't sound
like a bad resistor problem, at least from a thermal standpoint.
After I get back from my NYE gig in PHX, I'll have another go at it. What I plan on doing then, is to go stage by stage making careful measurements of the components and voltages. While I have a very nice Sencore LC-75, I do not have a real honest in circuit able ohm meter. My Fluke 87 III isn't the ticket for that sort of thing. Thanks for everyone's replies and guidance. I'll see you guys on the other side.
Happy New Year.

I have three 75A4's here in PHX and you are welcome to come by and take pictures or inspect them if that helps you.  I have a low Serial, a mid and a very high last run unit. 

Yes.


I think there is an i.f. gain adjustment pot somewhere IIRC.  I use a 20 dB pad between the dipole and vertical and receiver.  Don't use it with the beverage or indoor loop.  The rx seems to have adequate rf sensitivity, maybe a little too much.

 

I had a real low serial for awhile, like 330 or something.
The one in regular use is middle as in the 2400's and another is very late. So late that it has a factory noise blanker and a Collins historian presently has it trying to trace its history. The last time I actually used it was in the 80's and its never been recapped and the blanker wasnt very effective unless someone with pure copper ignition wires drove by. I shelved it when I sold the pair of Drake C Lines and got the first TS-930.

My ER 75A-4 Compendium showed up today! :)
Sure enough, there is a section devoted to repairing the frequency jump problem.
I guess I had better get organized and prepare the bench for extracting the PTO.
Repairing it looks to be easy and straightforward.

Just take your time especially with the front panel and wiring.

I will, trust me. I don't like re-working my own work.
The sticky tubing for the PTO wires will finally get 86'd too.

BTW...
When loosening the shaft coupler, can I remove the PTO without messing with those pesky
cam disks?

The cam disks are inside the PTO body.  If you are refering to the Oldham Coupler on the PTO shaft, it is not a big deal to separate it.  It will need to be cleaned anyway.

Be careful you do not over rotate the the shaft of the PTO.  The turns limiter is external to the PTO (on the front panel).  If you start cranking on the shaft when the PTO is out of the A-4, it can be turned too far and the result is not pleasant at all.

Agreed! If you turn the shaft on the PTO far enough beyond its limits to feel physical resistance to further rotation, damage may already be done.

Before  removing the PTO, best to set the frequency at 500 or somewhere else at midrange, carefully note the setting, and keep track of any rotation before the unit is disassembled. Once apart, note the exact position of the slug assembly as it rides oven the cams, even to the point of counting the leaves.  Try to re-assemble it exactly as found.

While it is apart, that would be a good time to re-adjust the leaf-cams for maximum linearity.  Most older PTOs have developed some error over the years.  It might not be a bad idea to fabricate an extension cable to allow the PTO to operate outside the receiver, and a frequency counter would be very useful for checking linearity. But be careful! This requires running the PTO all the way across its tuning range, so be wary of the range limits.  Temporarily attaching some kind of 360° precision calibrated dial to the shaft would make adjustment a lot easier.

Those PTOs are basically all the same.  I have worked on ones from the T-368, R-390 series and T-195, but never have (yet) gone into the ones from my 75A-4s.

That little slug that expands or compresses the tuning range to exactly 1 mc/s per 10 turns of rotation should be checked and adjusted as well.  If the slug is against the  stop and all the adjustment is used up, that can be somewhat corrected by removing turns from the coil (or adding turns if necessary, but you usually need to remove turns).  I believe this is all explained in the compendium.

Unfortunately, the problem with the PTOs going out of calibration appears to be from ageing of the powdered iron slug composition material, somewhat analogous to the resistance shift in carbon composition resistors.  It may not be possible to restore the PTO to perfect calibration, but it can be adjusted to minimum error. In extreme cases, perhaps some kind of outboard digital frequency display would be in order. In years past I used to see ads for those, adaptable to most vintage receivers. Otherwise, something else could be adapted to the purpose, or for the ambitious, homebrewed.

Remember, the factory specs for frequency calibration are only plus/minus 300~, when the nearest 100 kc/s calibration point is zeroed to the xtal calibrator.  That is a worst-case scenario, actually rather poor, given the inherent precision of the PTO and dial mechanism, and most receivers can be adjusted to meet or exceed that specification before/without even going inside the PTO.

One more hint regarding calibration and linearity adjustment: manuals for both the 75A-4 and R-390 series say to reference the end-point adjustment settings to the extreme ends of the tuning range: at the 0.000 and 1.000 mc/s points, and repeat several times until they fall in place at both ends.  I have found that this often leaves a substantial error near the middle of the range, where most of the popular ham frequencies lie.  Instead, use 0.100 and 0.900 mc/s for the adjustment points.  That may take it off a little at 0.000 and 1.000, but the intermediate calibration points will more closely line up into place.  Linearity error tends to be substantially greater within the first and last 100 kc/s of the tuning range, than across the intermediate 800 kc/s range between the 100 and 900 kc/s points.

Thanks for that clarification. I am referring to the several disks that reside in front of the PTO body
that have the little stop tabs stacked in a row prior to that final coupler arrangement.
I guess the trick here is to carefully map everything out as I disassemble and reverse order assemble.
According to the compendium, it echoes what you all have said about presetting the PTO frequency and marking it to make recalibrating easier when done.
I appreciate all of your sage advice about this operation. I can see how it would be easy to royally screw
things up and render it useless.
Onward!

Well, my 75A4 did exactly this.  I replaced the ceramic and mica caps inside the little box at the front and bottom of the PTO.  You do not need to disassemble the PTO.  It's tight in there , but you can replace them.  It totally cured the PTO instability I had.  It would sometimes sound as if someone was rocking the pto back and forth across a signal.  Try that before taking the PTO apart.

THANK YOU!
Here is what it looks like inside the compartment. Those little cubes of treachery are history!
While I am at it, the two tubular coupling caps are gone too. Replaced with Mallory M-150 series caps.
More to follow.

Problem solved :)
After replacing C-204, 205, 206 & 208, it is very steady without the "Dances with Wolves" that plagued it before.
Now that it is staying relatively on frequency, I can attend to the AVC and S meter problems.
THANK YOU EVERYBODY!!!!!!!!!! :)
Onward and upward!

Mazel Tov!  I finally helped somebody!  As soon as I read your post, I knew what the problem was.  GL with the rest.  They can really be a pain, but totally worth the trouble when finished.

73,
Mike WB0SND

Mike
It will be on frequency tonight for the West Collins Collectors 75 Meter Net.
Hopefully I can attend to the AVC problem. Shouldn't be too hard.
Thanks again!

Congrats!   I got my DC relay in the mail.  I should have my Gold dust twins on the air next week if all goes as planned.  I cant wait to make a contact with you, Twins to twins.

C

Yahoo!!!!
I will have mine on tonight! I just hope the band holds up... it was very sloppy last night.
See you next week!
Here's the Twins online and in the hunt tonight!

The disks you are referring to is the PTO turns limiter.  I don't recommend taking it apart, just flush them out well with solvent and lightly lube.  I tried to re-purpose an old one and the tolerances are pretty tight, have up on it.

The Oldham coupler I mentioned is the gizmo with the spring between the two outer faces and the slotted brass disk in-between. It connects the outer shaft to the PTO and the purpose is to provide non backlash compensation for any minor misalignment of the PTO with the outer shaft.  Oh, the spring will typically fly off into an adjoining room. Not to worry, you will find it a few months after you "make do" with something close.   

I was wondeing if it was ok to stack the units like that.  That makes me feel better about getting the rig into my small ham shack. 

I will get flamed by this next question by any Collins snobs that happens to be on the forum, but has anyone thought about Transcieving with these units?  I am sure someone has done this at some point.  EIther external VFO for both or having a buffer board that Drivers the other?  Probably to much work and you would really upset the CCA members at the though of any modifications.  It sure would be cool to have external Solid state ROCK solid VFO.

Do you find you have to chase the reciever and transmitter around?
c

They are both fairly stable, with the KWS-1 being best. Stacking these is not an issue because of the venting of the cabinets. I cheat a little bit when I am in QSO. As a courtesy to the other OPS on frequency, I monitor my on air transmissions with headphones and my trusty 756 Pro. Therefore, I make absolutely sure that my TX is on freq. I am quite surprised at how well the rig stays on frequency but it will get better as I go through and eliminate those problematic molded mica's.
After being in QSO for over an hour tonight so far, the 75A-4 is rocking big time and staying relatively on frequency. I am also impressed with the floor noise quality compared to my other radios. The lack of AVC action is very apparent with strong signals and I must fix the problem. The beauty of the radio demands better AVC action.
I love this receiver. It will be a real outstanding unit when I am done with it because I am completely committed to getting it right. It has a lot of promise and is in very good shape considering it's age. It's 6 years older than I am and now stays on frequency better than I do! Hi! I am also observant of not going too far with the modifications, especially those that have been proven not reliable or warranted.
The compendium has been a savior, as well as several of the OPS here with the excellent contributions.
Having a real ball tonight.

Glad that the caps worked out. Better than getting into the PTO....Seems like PTO's get nasty after very very hard use / abuse
If the radio has been in a decent environment and not twisted violently around, the PTO may never need opening.
CAPS are our enemies. But hats off to them for lasting this long..

Good luck with your goodies

I recommend the AGC mods by George, W1LSB.  They are in Electric Radio about a year or 2 ago.  This mod is far superior to the more common and popular K7CMS mod.

If they are hooked up the way the manual says with teh BNC,  connector.  You can listen you yourself talk on the transmitter through the 75A4 without transmitting over the air. No need for the other radio on with headphones. You can flip the Driver stage on, Talk a little and fine tune the Kws1 vfo, Then flip the switch back and you are ready to go.

It sure would be nice to have them rock solid.  But I digress, I love old radios and this is one small anoyance.

The 75a4 I will be using is a very late serial number.  I have not tested it much more then about 5 minutes.  Its all working though.  The older 75a4 I have that came with the kws1 has a very fast AGC. TO fast for my tastes. About like my Kwm2a.  To smooth it out, I back down the Gain a bit and its very smooth.

C

P&H made a transceive adapter that only did 20/80 and Ive only seen one ages ago. They (or somone else) announced a multiband version but it didnt hit the market that Im aware of.

In the late 60's I tried designing a 160-10M one to go with my 100V, even using a lot of SS, at National but never licked the birdy problem. These days it might be a lot easier, and I even have the same rigs, but its not something I'll likely get around to.

Carl

I'm sure there is a way to do the transceiver operation like the S-Line does, but it sounds like a lot of work.
It would likely involve some internal modifications that aren't worth the hassle. They're a lot fun the way they are.
I had a ball on the 75 meter CCA net last night. There was a lot of the gang on board, including FOUR "Gold Dust Twins" stations! The band co-operated too and was very fun for a change.
This was the first time I had ever used a 75A-4 in QSO and was impressed with the lack of floor noise and the ability to navigate from some QRM. I think it can be better and certainly the AVC problem needs to be addressed. When KB6OQJ transmitted his first signal, the volume nearly destroyed my speaker! He was easily in
excess of 40 db over S9. The quality of the receiver audio is very good despite the crappy output circuit.
I performed the feedback mod in the audio amp and it made a big difference and has a smoother response than it was as stock. I will end up using a direct output into an external power amp once this has been properly installed in the station. Right now, it sits atop the KWS-1 while I go through the trial and test period.
Once both units are completely finished, they will be the center piece of the SSB station
Can't wait.

I tuned into the 3895 last night.  I heard a bunch of guys arguing and going on yelling at each other.  Strong signals.  BIG argument.  I got tired of listening and turned it off.

C

Wasn't our crew. We were off the air by 9:45 PST

I will be on the CCA 20 meter net today 14.263. Afterwards, I might be @29.000 and up on AM.
The Twins are holding up well. I think I have both a plate load and cathode resistor going nuts
in the speech amplifier circuit of the KWS-1, so you might hear a little crackle and pop if I don't get to it
today. I hope to work some of you on the air.
73, Billy