The AM Forum

I did the audio mods from the AM Window site to my 32V-1, the audio is very nice but I do have one problem , but I'm sure someone here will be able to help me. The audio is fine, but when I hookup the negative feedback there is a hum on the carrier.  I switched out the 6SL7 tube with no change. I doubled the resistor value and it reduced the hum, but it also lowered the gain from the mod. Any suggestions of what to check next would be greatly appreciated.

  If you refer to the 1 meg in the WA1HLR 32V2 mods, then doubling the feedback resistor should INCREASE the gain. I think Brian pegged it and your feedback phasing is backwards. So pick the other side of the interstage transformer (E308 instead of E307).

   The hum you are hearing is there, just amplified. The schematic does not show the low B+ supply, but maybe the ripple is too high at that point. Are the filtering caps old, or low in value by design? There may be an opportunity for improvement there. Also, looking at the MOD by Tim, he changed C211 from .05 uf to 10 Uf, a big and good change. That said, I'd go farther, and put in at least 50 uf there. I say this because the 6SN7 parallel audio driver will literally "modulate" it's B+ supply unless the B+ source has a VERY low source impedance which it currently does not have.

Good luck!
Jim
WD5JKO

Thanks Guys I'll check all of that out tonight when I get home. I dont think it's the wrong phase, I did try the other side and got a squeal out of that one. I'll add more cap to C211 and give it a try.  It does not say to use a shielded cable, but I did try that too. Thanks for all of the suggestions, I'll post later and let everyone know whats going on. ;D

Maybe you get the hum from the mic amp filiment, and adding the feedback extends the response down to where you pass the hum through the audio chain, a GOOD sign.

When the feedback is backwards, I always got a high frequency (sometimes supersonic) problem.

What I did to eliminate the hum was run the mic amp filiments off dc.
I diode in series with the filiment, and an electrolytic cap filters it enough to eliminate the hum.
The size of the cap (UF) sets the voltage, as a bigger one will charge up higher, so I just experimented with the value of cap that gave 6 volts dc.
Yes its likely poorly filtered with half wave and little filtering, but its not AC and removes all traces of the hum.

I use a D104, and a 5 meg input and AC floating around don't mix well with extended frequency responses...

I also put in KT88 hifi tubes and removed the driver transformer, regulated the screens with VR tubes on that little deck where the mod tubes sit....

Brett

Brett
 

Is thsi the mod???   http://amwindow.org/tech/htm/32v.htm


I did those mods to my 32V-2 years ago and remember having the same problem.  I was unable to figure it out.  Gary, W2INR contacted Tim, WA1HLR who noticed the schematic was wrong-perhaps in the transition to the web?  The value of one of the components in the feedback circuit was incorrect on the schematic on the AM Window page.  I have since sold the transmitter and can't recall the correct item and it's value.  If I was to guess, I'd say the value of the the .01mf cap in series with the 1 meg 2watt resistor should be .1mf.  Maybe you can send Tim a PM or an e-mail to verify.   Sorry I couldn't be more helpful. 

I did all the mods with slightly less vauled coupling caps than Tim's and used no feedback circuit.  Left the 1st AF amp's input resistor at 1 meg. but fed it with an MPF102 follower with 16 megs input.

Dont' forget to switch phasing (switch 807 plate leads, etc.) as necessary to get best peaks on a scope.

Added 20uf/450v around all the bathtub B+ bypasses.

Sounds great. BTW the driver xfor in the 32V is as large as a lot of 75 watt modulation xfor's and easily takes increased bandwidth.

My greatest reduction in hum was had by simply running a short ground lead directly from the d104 stalk to the front of panel's mike input jack.  Eliminated coiled ground wires in the mike cable and all the other screw terminals, etc.

Hi Guys,
  Thanks for all of the great suggestions, I tried a lot of things tonight. I changed C211 to 22uf and that made the biggest difference, but the hum is still there. I also added a diode to the filament voltage for the 6SL7 and the 6SN7 and put 1000uf cap to get the voltage back up to a steady 6 volts, but still have the hum. I took the caps out of the mic lead and things got worse so i put them back in. Just for the heck of it I put a 1N4508 diode in the feedback line and no more hum, so i turned it around and still no hum....so my question is am I just rendering the circuit inoperable with the diode, or will the circuit still function?  The phase is correct, if I hook up to the other phase it squeals ...at least I'm still at the point where I still have more time than money invested here ...any other ideas, or do I just forget the feedback circuit?    ???

Ray, I suspect the signal across the diode is not making the threshold, so in effect the diode is not there. You really don't want a non linear component in the feedback path.

Since bridging C211 with a 22uf made a big difference, recall my suggestion to go to eat least 50uf? Better yet try 100uf or more. With that dual triode 6SN7 driver, source impedance of the power supply is important. There is more going on there than just ripple on the buss from power line filtering. The schematic shows two sources of B+. I'm curious how these hook up. Is the low B+ source not shown in schematic sick, or just in need of an update (bigger cap on output side of pi C-L-C filter??

Do you have a audio generator and a scope? If so using these tools can be very beneficial in determining what is going on. For example, changing C211 from 0.05 to 10 uf to 100 uf while modulating the rig near max (whatever that is) should show step changes in modulating capability as seen on RF modulated envelope with scope, or maximum undistorted 807 G1 audio waveform.

If you have a scope, also AC couple the input, and look at the two B+ supplies for preamp and driver. What kind of bumpy ripple do you see? What is the peak to peak magnitude? A 10X probe with 600v peak capability is needed for these measurements. Keep the probe OFF the high B+ though, or ZORCH! :'(

You might also be fighting a ground loop that is exasperated by closing the feedback path. Situations like this take a lot thought. The interstage transformer if NOT magnetically shielded (iron box) can pick up stray magnetic fields too.

One thing you said early in this thread still bugs me:

"I doubled the resistor value and it reduced the hum, but it also lowered the gain from the mod."

If this is the 1m feedback resistor, then that should NOT have happened.  Do this test again with scope and audio generator.

All the Best Ray,
Jim
WD5JKO

I doubt the grid resistor is the problem. I've modified mine to 10 Megs and have no hum.

Yes the hum is still present without the mic attached.

Well, it's getting in somewhere.
Audio trace from last stage backwards, etc.

-you know, all the ol' service man's tricks starting with the dumbest thought and easiest fix first.
You did the tubes.
Maybe it's just a bad ground from some stage's bypass capacitor.

Borderline for what?

It is not a grid leak resistor. The tube bias is set by the cathode resistor.

Grid leak implies that grid current flows as in a Class C circuit. No such current flows in a Class A amp.

On most tubes? None that I've ever used. Using 10 Megs caused zero problems.

LOL.

10 Megs does not equal infinity.


You have said enough.  ::)

Ray and Brian,
 
   Yes RF in the audio might be possible causing weird symptoms. Using a dummy load might help if the RF is getting in from "radiated RF". Your thought about the RF bypass of the plate choke causing "conducted RF" is also possible. Keep in mind though that too much capacitance there will attenuate the higher frequency audio from the modulator.

   So far we've heard from Ray saying the hum goes away when the negative feedback is removed, and the hum is decreased when he bridges C211 with a 22 uf capacitor. Also when the feedback resistor is doubled, the hum is reduced but the gain drops. Then removing the "caps from mike lead" the hum got worse...

Some ideas:

*  Have we tried removing the preamp 6SL7 to see if the Hum goes away?
*  Have we tried bridging C211 with a bigger cap? I mean 50-100 uf or more
*  Try test with dummy load, and try various bands say 80m and then 20m
*  Have we tried bridging the plate choke RF bypass cap as Brian suggested?
*  Is the hum seen on a scope, or just heard on a local receiver? If a local receiver, is the hum real, and not an artifact due to a local setup grounding issue between your receiver and transmitter? I ask because monitoring yourself with a receiver is complex to get the levels correct, and to eliminate any ground loops. In my shack I have been fooled many times hearing hum that is really NOT being transmitted.
*   If the hum is real, what is the best guess magnitude? For example if your voice modulates upward to say 100%, then what mod % is being transmitted when you don't speak?
*   Is the hum 60 hz or 120 hz?
*   Could the hum be coming from the Mic PTT line? Is the DC for the PTT half wave or full wave rectified?


That's all I got for now.

Good luck Ray,
Jim
WD5JKO
 

Thanks for all of the great suggestions, I really didnt get a chance to look at it any more last night. I am using a dummy load, and I am listening on my 75A4. I'm going to try Brian's suggection onthe plate cap tonight and look at the others too. I'll try pulling the 6SL7 tonight and see what it does to the hum, and I'll try the 6SN7 too. I'll post my results later tonight ...hopefully they will be good.
       Thanks Again!

Well after trying all sorts of things I had given up hope of ever getting the negative feedback circuit to not induce a hum on the carrier, so I moved on to another problem. I had bought a pair of Weber Copper Caps to replace the HV rectifier tubes, but when installed they caused the fuse to blow as soon as I applied the HV. I had been told by someone who uses solid state in the HV supply that he was able to do it by adding caps on the rear terminal strip. So I added 50uf between terminals 4 and 6 on the rear strip, still blew the fuse when I applied HV with the Copper Caps though so I gave up on that too.  I left the caps on the back and figured what the heck I'll try the feedback one more time....well with the caps on the back the feedback circuit does not hum,  go figure ...I was about to give in and sell the 32V...now if I can just get it back in the case without something else happening .....Thanks for all of the help everyone, I really do appreciate it it gets frustrating, but I try to keep telling myself, it's my hobby and I'm having fun  ;D

Ray,

I am a little curious where you added that 50 uf capacitor. The following schematic for a 32V3 shows what I suspected.

See "list of illustrations", and schematics A, B, C

http://www.qsl.net/wa5bxo/32v3/

So does adding the negative feedback increase or decrease the gain?  ;)

A similar story happened on a RCA AR-88 receiver where I added negative feedback, increased coupling cap size, etc to make the 6K6 audio output high-fi and go down to 50 hz.  The reusults were horrible until I realized that the audio output B+ had about 100 volts P-P ripple on it when I cranked the audio with a male voice speaking. The draw on the power supply was causing the C-L-C-L-C filter to ring at the filter resonant frequency (about 80 hz if I recall). When we take an old boat anchor like your 32V and boost the bass, it is important to also upgrade the power supply to provide low source impedance down to the lowest audio frequency used. This includes both the HV and LV B+

Glad your out of the woods!

I bet one or both of them copper tops failed on you.

73,
Jim
WD5JKO

I guess the feedback actually cuts the gain a bit, suppose to cut distortion from from what I've been reading. I've received good audio reports and I'm scared to say everything seems good. Thanks for all of the input, I hope all of this will help someone else sometime. I was surprised at how many guys said they did the same mod and had the same problem. Well it's back into the cabinet tomorrow, wish me luck.  ;D

I think you want to remove as much load off the LV transformer as possible (and that includes filiments) as the LV transformer seems to be the only weak link in the 32v rigs.
I removed all the tube rectifiers and replaced with solid state diodes 30 years ago, and also added 25uf 1500 volt oil filled cap to the terminals on the back for energy storage.
I think the stock 8uf cap is a bit light for the rf and modulator......
If you find the right cap, you can sometimes get a 10 uf in there...

Brett

I was not suggesting the hum problem was from the LV supply, just that you want to remove as much load from the LV transformer as you can, since that is what seems to fail on the 32v series rigs.

Oil filled caps last forever and dont tend to explode like electrolytics in series can.

The little cap across the choke improves regulation of the power supply, and does not normaly cause any hum.

When I did my audio mods, I had a lot of hum on my audio, but the dc on the mic amp filiments eliminated it, I only added the extra capacitance for energy storage....

Brett

That oil cap across the LV filter choke blew to smithereens (twice) in KA1KAQ's 32V. 

that cap looks like some attempt to resonate the choke. the minute you change the other values, it''s probably way out of resonance hence the blowing up. That cap would only work in a bone stock rig.

In my 32 V I thought the fixed mica loading caps were a bit too fragile. Not a big fan of that transmitter. Too hard to work on and too crammed together. I'll take a worked over Viking 2 any day.

I wonder how the modification posted at the following link can be updated to illustrate that the audio modifications are only part of the package, i.e. beefing up the power supplies are required before satisfactory results are obtained.

http://amwindow.org/tech/htm/32v.htm

Jim
WD5JKO

I just looked at another modification at amwindow.org, and this one is for the Viking II. Take a look at the circuit:

http://amwindow.org/tech/htm/vikinga.htm

So what is wrong here?

1.) negative feedback is eliminated with 6C4 cathode bypass capacitor (might work fine without NFB)
2.) If the 807 screens are tied to the low B+ as in note 1, won't the 807's screen grids burn up when the high B+ is turned off??
3.) ??

Maybe add a second note to key the screen grid of the 807 with a relay??

I realize that folks went through a lot of trouble to come up with these modifications, and again to document them and post them. I applaud this effort. I do hope though that folks contemplating any of these modifications realize that these posts are not necessarily complete. Therefore, ask questions on this forum first, and hit a book or two before warming up the soldering iron.

Regards,
Jim
WD5JKO

The screens could be at risk, but since the LV and HV operate on the same switch, the operator could never cause the problem. If the HV supply craps out, you have a different story. These mods have been done by many  people and I've never heard of this problem. That said, there is a possibility. Like any mods, there is some risk involved.

If you would like to submit improved mods, please do so. It would be greatly appreciated. They will be added to the collection or be used to replace older, less complete ones.

Nice catch guys. The mod info has now been modified. The Note 1 section has been removed.

Thanks for your input. We don't need to be burning up any 807s needlessly.   ;D

Steve said:

Oh the humanity!!!!!  :'(

Interesting point, Derb. Though nothing in the set has been changed or modified, it got me to wondering if a few shorted turns in the choke may be the culprit. It's munched a couple of original .25 mfd 2kv oil units in short order. Found an exact (but taller) 5kv unit at Shelby and installed it, so far no issues. Applying the old principal 'If it doesn't work, force it. If it breaks, it needed replacing anyway'.  ;D

Back to the issue of hum. Hmmmmmmmmmmmmmmmmmmmmmmmmmm. Got me to thinking of an issue I had with an old Zenith set years ago. Since the issue developed after operator intervention for mods, I'm wondering if it's a simple case of inadvertent installation or repositioning of a component or line near a source that it wasn't near before. This happened to me with the old Zenith set, changing caps and repeated revisits produced no improvement. Then I accidentally moved a wire added by a previous owner for phono use, and noticed a difference. A little experimenting with repositioning and voila - hum gone.

Seems like the problem has been addressed in the subject 32V, but still a good thing to remember. Sometimes it's the obvious things we miss looking for something more worthy of being called difficult.

I kinda specialize in that these days.  :-\ I dont work on things logically anymore, but more on emotion.

Hi

I agree with W5HRO,since the feed back is taken from the HV power supply,secondary winding of T202/mod.trafo ,the hum can be introduced from this supply ,because if the supply has a little ripple/hum it is amplified by the 2ND (V201) and the modulator.

When we changed a tube rectifier with solid state rectifier ,

the tube rectifier takes time to rectified the ac voltage because it needs to get "warm" to conduct ,so the in rush current is not high,

with solid state rectifier ,when you switch on the transmitter ,it directly conduct ,so the in rush current is high ,that makes the fuse open.
with higher value of C the inrush current is higher.

Try to change or parallel C 303 with a higher value of C ,the more higher the less ripple of the HV supply.
Still used the tube rectifier .

When using Solid state rectifier ,we must used surge limiting devise in the primary winding of the HV power supply.To limit the inrush current

Gito

Well, in the case of the 32v, the high voltage is keyed on transmit, vacuum rectifiers should be conducting by that time.
The low voltage power supply is on all the time, but on power on, there is almost no load as none of the tubes have warmed up yet.
Both supplies are choke input, so surge current is fairly low, even with big caps.

The cap across the choke was a popular method to improve regulation, its in the Bill Orr handbooks.
I have never had a problem with mine, and even added a .5uf across the choke in my big rig supply.

I have run the 32V3 up to 900 volts on the B+, I have my 32V3 rack mounted, and was using it as an rf exciter for the big rigs, and used the variac as a 'dial a power' knob, its mounted on a rack panel in the bottom of the rack.

As a standalone transmitter, I usualy run it about 800 volts and 180ma, it seems to modulate better that way.

This 32V3 I got as one of two for $90.00, both looked like new, one blew the LV fuse, and the seller thought the LV transformer was bad, but it was a shorted trimmer cap in the multiplier section...

I also had a 32V1, and never had any problems with any of them except one had the plate tuning cap arc over all the time because there was force on it shoving the plates off center, an easy fix.
All had been solid stated without any dropping resistors and run on the 700 volt tap.

Of all the 'ham' gear I have had over the years, its the only thing that has stayed in the shack for 30 years.
A well built transmitter!

Brett