Advice on my rapidly deteriorating DX-100b audio?

[kf6pqt]

So last weekend I aquired my new old dx100-b, a local ebay pickup.

Rig has obviously been recapped, modded to use 6146b's instead of 1625's for the modulators. The audio is about as nice as cat poop on a rainy day, is inconsistently better or worse (usually worse) with each transmission.

The pictures you see are from before I ever fired it up, I was surprised it even modulated at all.

Final plate current, and and final rf output go DOWN on voice peaks... I thought the opposite was supposed to happen?

I've tried a different mic, and a few different 12ax7's and 12BY7s. Also notice that the mod 6146 filaments are much more dim than the final 6146 filaments.

Possible scenarios:

My mod tranny is gasping its death rattle?

I just have a bad cap or resistor somewhere?

In dropping the filament voltage for 6146's, a resistor has burned up?

And, what might be this disconnected string of 1n4007 diodes? Circled in red, hard to see, with the arrow?

Thoughts?

Thanks,
Jason kf6pqt

[kf6pqt]

That "blackened hole" leads to the mod tranny.

[kf6pqt]

and...

[The Slab Bacon]

If your outpoot is going down on voice peaks, you may have something sagging in the power supply. 5R4s are good for this problem. HV filter caps that have lost some of their capacitance can also cause this problem as well. Also some average reading wattmeters will show this phenomenon when you modulate as the average power drops, but the peak rizes. Be sure it is not your meter.

Looks like you have a transmittah that has been "hammed up" You will have to do what I have had to do to too many rigs. Lay your hands on it and do the "vulcan mind tap" and become one with it.
Seriously though, Get a copy of the skizmatic, and have it blown up fairly large. Compare the original skiz to what you find and try to figger out what the last owner was thinking with his "mods".

And if all else fails, rewire it back to original and start your own "mods" from there!!

                                        The Slab Bacon

[The Slab Bacon]

Jason,
        looking at your pix, this is so typical of older BA gear. In my last post I refered to it being "hammed up" You find this quite often in older gear. I have restored / repaired much of this stuff over the years. If you are looking for high reliability in a piece like this, sometimes a pair of diagonal cutters and a lot of desoldering braid can be your best friend.
 
I run nothing but old BA gear and have little or no reliability problems with it. The trick is to do a meticulous electrical restoration on it and cut out all of the JSs and jury rigs.

Take your time with it and do the best job that you can with it, dont take shortcuts or jury rigs, and you will be rewarded with a nice reliable piece of gear. Do it half-assed and you will be inventing new cuss words every time you fire it up.

                                                 The Slab Bacon

[AF9J]

Jason, it looks like you're going to be doing the same thing I might have to do - undo a mod.  I looked in my HP-23A power supply this morning, to determine what caps I'd need to buy from Mouser, and realized that the the nice, big, newer 125 microfarad cap that's in the bottom is an extra cap that doesn't even belong in the supply.  It looks like it's hooked up where C1 (another 125 microfarad cap) is, but if it's replacement for C1, it's a quick & dirty, lame job.  Because, C1 doesn't look like it's been disconnected.  As it is, all of the other 125 microfarad caps (4 total), look like they're shot.  Two of these help with the HV B+ feed that is so low at the present time (550V no load, instead of the expected 820V no load).  Why do I get the feeling, that I"m going to end up rebuilding a power supply?  Good luck Jason, it looks like you're going to be doing a rebuild too, and yours is much worse than mine.

73,
Ellen - AF9J

[kf6pqt]

Yes, I know I'm in for it until I figure out exactly what this guy did. Hopefully I can get the modulator section schematic traced out.

What does everyone think about the choice to go with 6146's? I find it ironic, in that I have boxes and boxes of 1625's...

How well (assuming it too, is any good) would the mod tranny I pulled out of a BC-375 work in the dx100?

And, I thought  I read about this somewhere, but now can't remember... the differences between the 100 and the 100b are in the VFO, a couple front panel controls, (meter, grid drive) and the case, and thats mostly it, right?

BAMA only gives me the non-b manual...

Thanks,
Jason kf6pqt

[Todd, KA1KAQ]

The Slab has really nailed it, Jason. With some work and rewiring, that rig should be pretty much bulletproof. Much of the old gear is once it's maintained properly then used regularly.

I also have a case of milsurplus 1625s and was going to offer you a few, but it sounds like you're set. My vote would be with Frank's suggestion: rewire and rework, then enjoy.

[The Slab Bacon]

Jason,
        Someone may have subbed the 6146's in an attempt to get more audio power. After all they spec out better than 1625s. 6146s will work ok-fine for AB1 or AB2 modders. But if you have the 1625s why not use them.

As I said before, I cant tell you how many times I have looked at a piece of equipment to be repaired and said: "what in the hell was this guy thinking!!" As I reached for the diagonal cutters.

when I restored my Apache, it was missing the mod iron. I used a dixie 100 mod transfoma. I didnt have the space between the existing tube socket holes for the shouldered 1625 / 807 stile tubes, so I used a pair of 6146s. They worked OK-FINE once I adjusted the screen and control grid biass accordingly.

                                                       The Slab bacon

[WBear2GCR]

The diodes may be negative peak limiter diodes.
Or even part of a local bias supply... two in series for higher PIV... replacing an old selenium stack...

The downward modulation may be because you are not loaded up properly?

The 6146s may have more power than a pair of 1625s, depending on a few details, which include the B+ applied, the match to the xfmr and load, and the bias/screen voltage. Probably that's why they are in there.

If there is a blackened area around the mod iron at the bottom, that tends to indicate that one or more of the leads has cracked and flashed over. I'd probably pull the mod iron out and inspect it carefully, maybe even gently open the metal base to see inside the little hole - a dremel tool with an abrasive cut off wheel and a few minutes will make a nice clean cut out... depending on what you find, you may be able to repair the leads.

The mod iron is potted. IF you remove the entire bottom of the can - it may be riveted or soldered, you can put the thing into a jig above a pan into ur oven (or an oven you grab off the street for this project, since it will stink up the place) and heat the thing until the tar melts (or wax, depending on what's in it)... the stuff drips out the bottom, you let it go until there ain't much left... and then repair the leads. I'd probably shrink tube/spagetti (teflon is best) the leads so that they don't get gooked up in the process, and you can still clearly see what is what... then replace the leads back up at the windings... when ur done, heat and replace the goop... (after testing).

Or buy a "new" DX-100 mod iron off ebay, if that one is bad.

A high voltage test is not a bad idea either.

Or a bench test with clip leads using a dummy load for the secondary, no RF on the rig, just B+... see how it looks on a scope after running it with a sig gen or other input source for a while...

The problem could just be leaky coupling caps elsewhere in the audio chain... if the audio is degenerating in quality over time...

Great how simple things can suck up time and effort, eh? 

              _-_-bear

[WU2D]

"Other than that she's cherry"

The modders and conversion artists are our best friends - how else would you get a rig nowadays for a decent price?

I'm going though the same deal with my Apache project. I stripped the modulator section clean - I may even replace some sockets.

Mike WU2D

 

[WA1GFZ]

2 viking 2s in a row had the same problem. The coupling cap to the audio gain pot were leaky putting positive bias on the driver sending it into saturation so it had no AC gain. I could get a little audio if the pot was set low. It had lots of distortion. Best to check all the caps and replace the old ones. JS means jock strap. 6146s should work fine but the screen voltage should be lower than a 1625 screen.
BC375 trannie will work. I just picked up an ARC38 mod transformer to put in my rig that is nice and small 2 6146s modulating 3 (28 volt heaters in an A/C)

[W1GFH]

Don't assume you need to replace the mod tranny, at least yet.

Is that the butt end of a PAPER CAPACITOR I see?



http://audiofiles2.jerryseinfeld.nl/nononono.wav

Aak! You can NOT allow 50 year old paper 'lytic caps in the speech amp, yikes! (If there are old caps up behind that partition which the previous owner was too lazy to take out, a leaky one could be a culprit.)

And I see LOTS of extra components around the tube sockets. Extra huge power resistors to get the screen voltage down for the 6146's. Bizarre diode strings. Resistor kludges.

http://audiofiles2.jerryseinfeld.nl/nononono.wav

I'd rip it ALL out. Even if modded, that section should be much less confused, like k3msb's for example: http://www.k3msb.com/dx100b/dx100b_bot_audio.JPG

I agree with Slab and others: rewire and rework, then enjoy.

[KB2WIG]

   "Extra huge power resistors. "   That one pic shows a "litic" next to the resistor kluge...  not gud fer them 'litics.......  klc

[John K5PRO]

Re: Your comment on the 6146 in the modulators having less filament glow. If they are 6146B, this is normal, as the filament is different. You can verify by measuring the filament V at the socket, with HV off.
Also, if it is converted to 6146, you would have to rip out the socket and go back to that older style socket for 1625/807 style tubes. I would stick with 6146, if you get the screen voltage right and set the idling current via a cathode zener. If you look at the AMFone archives (search DX100), there are a series of reports I did on a conversion of mine from 1625 to 6146 then to 6CA7 where it is now. I found that the 6146 makes a better modulator, as is the 6CA7/EL34, as it has more cathode current capability than the older tube.

Sounds like your rig may have been bastardized by a happy-go-lucky modder, without paying attention to details and observing the results of each change before adding more. Other option is that, like others have suggested, it has leaky capacitors, and other component failures which are symptomatic of your negative carrier shift. Sketch the circuit for the Modulator (and the RF output stage if that got changed, esp around the screen resistors for the 6146 output). Post it here and you can get free advice, from the galley here.

[N3DRB The Derb]

that rig has been seriously screwed with by a mental pygmy. That's why it dont work.

Strip every wire clean offa the term strips, and take it back to stock and get it working stock, then have the people here guide you into making some FB mods. Be careful with the low level end as thats how my dx100 blew up. a ultrasonic feedback built up and I was not around to shut the rig off. It melted everything including the LV transformer. That's how I met the Slab so many years ago.  limit your low level response to 10K.

You know those HV twist lock caps are bad right? I've never seen one that didn't need replaced. You'll find they have now undergone a gay sex change and turned into probably 200~700K resistors each. They always need replaced.basically ANY paper cap in the unit needs to be replaced.

I've forgotten most of what I knew about that rig. Too many years ago.

[kf6pqt]

Allrighty, we're off to a good start, even before I've posted a trace, we assume that the modder was on crack. That makes me feel better than say he knew his stuff but got stumped by a mystery error or glitch.

I was too distracted by the black marks, and too excited to fire it up to notice those paper caps. When I open it up again tomorrow, I'll try and do a much better high-resolution photo survey, and try to trace out the modulator section.

Thanks!

-Jason kf6pqt

[WBear2GCR]

The mods may or may not be fine... trace them and see what they are doing, or not doing. Imho.

Then of course they look ugly, but if they work....

Taking good extrmeme closeup pics with a digital camera? grab a standard magnifying glass, place flush up to the camera lens, use the LCD finder to focus and move the camera until the image is focused. I presume this will still work with active autofocus cameras, it has for me...

                    _-_-bear

[kf6pqt]

This new one is a 7 megapixel, trying to use it in macro mode on those diodes didnt work out so well, but that may have been user error.

Next attempt will be to take wider shots, and do detail-level zooming-in with the image editing software. (gimp.)

That, and I'll definitely be a lot more patient and look at the thing with a more critical eye.

Depending on how much work this is gonna be, I may first be back to slapping together a grounded-grid amp for one of my PW tx to remain operational while I take my time on the dx100!

Going from a DX-60 to the DX100, was like going from a 10-speed to a V-8, I don't think I can go back.

[W1GFH]

Be careful with the low level end as thats how my dx100 blew up. a ultrasonic feedback built up and I was not around to shut the rig off. It melted everything including the LV transformer. That's how I met the Slab so many years ago.  limit your low level response to 10K.

Tales of the Great Derb DX 100 Crapout are still circulated in hushed tones, and the horrifying details have been known to make strong men weep.

I must concur with Derb's assessment of the mental midgetness of your rig's modmaster.

[KB2WIG]

 "Tales of the Great Derb DX 100 Crapout are still circulated in hushed tones, and the horrifying details have been known to make strong men weep. "


or


scha·den·freu·de (shäd'n-froi'də)
n.
Pleasure derived from the misfortunes of others.



       klc

[N3DRB The Derb]

would NEVER trust the work of some ham I didn't know was not a crack repair guy. too many times, the reason why they did the mod is at cross purposes with what your goal is. They might have wanted to clip the peaks for contest work, you wanna sound like WLW. 

the only way is to find and root out ALL crap down to the term strips. Then get it working stock. Then apply ur mods in a no -JS manner.

[flintstone mop]

I'll have to agree with DERB and some of the others here. The previous owner was prolly a CBer and they are not known to have quality audio in their neck of the woods.You will have to get the Tx back to orginal and add the proven mods. 6146's as modulators really wreak and probably aren't a match to the mod transformer. The 6CA7's are really nice audio pubes.
Fred

[kf6pqt]

I'm not against going back to 1625's!

Also, 10/11m supposedly doesn't work on this tx, that plus the fact it doesn't reek of nicotine leads me to believe it was done by a non-chickenbander.

[W1GFH]

IMO, the 6146s are not worth having all that extra bullcrap in there. If you have 1625's and sockets available, it's easy enough to rebuild the mod section as it should be. The "stock" mod section uses a helluva lot less components. Strip and rip. Side cutters are your friend. While you're in there, you can upgrade the caps and resistors around the 12AX7 as Ed suggests to remove the bandpass restrictions, then step back and evaluate, and see if you want to pursue further mods.

Might as well get your capacitor order in to Mouser, or my fave vendors: http://www.justradios.com/capacitors.html