Forums :: NEW! Web Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Dec Tue 07, 2021 4:14 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 14 posts ] 
Author Message
 Post subject: Strange problem with Philco 38-7
PostPosted: Sep Thu 23, 2021 11:38 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
I've had this Philco 38-7 for about 7 years. When I got it, I did a chassis restoration, and every since then I've had a problem with the audio will be low/station weak and it will be noisy like intermittent static, and the sound will increase/station will come in stronger. Yesterday I removed the chassis to try to fix it. I had the oscope connected to the input of the 1st audio tube, and almost every time I touched my multimeter to B+ or anything with DC voltage, the sound out of the speaker would increase and the oscope signal would increase; B+, screen voltage to the 6K7, AGC, anything. My multimeter is a Fluke 87 III. I tried tapping using a plastic screwdriver, hitting the chassis, tapping on tubes, I changed the 6A8, 6J5, 6K5, with no effect. The 6A8 gave me a lower signal and sounded good for a while. I left the siggen connected and set to 1.4MHz. The radio had a nice signal with 50mv to the speaker. I left it for a half hour and when I got back it had 100mv on the speaker, not that bad. Then I connected the chassis to ground and got 600mv. That's great, I removed the ground and still got 600mv. Half an hour later it was back to 50mv. If I touched my multimeter to B+ the signal could have gone to 100 up to 600mv again.
I can't seem to make any changes myself, unless my multimeter touched B+ or ground touches chassis.
With my multimeter connected to AGC, it varies from -2.3 on no station to -6.5 on strong station. AGC sometimes changes with the sound out of the speaker, sometimes not.

http://www.nostalgiaair.org/Resources/256/M0013256.htm


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 12:31 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1627
Wow, the Philco model 38-7 that is my favorite tube radio of all.

It had a number of innovations in its circuit design, including delayed AGC, an interesting feature on one of the IF transformers with a single turn of wire around it placed in series with the cathode of the IF tube, the subject of a few interesting debates, a brilliant cone-centric tuning mechanism, great anti-backlash drive to the V/C, well designed RF & osc coils and more. The cabinet was brilliant too.

Since you have an intermittent problem affecting a circuit enclosed in a negative feedback loop (the AGC) you will have to "break the loop" to find where the problem is originating, or at least perform a test where the AGC is completely inactive. Because any change imposed on a system with feedback, the loop will attempt to compensate for it and all circuit points will change voltages. So it may be necessary to disconnect and disable the AGC before performing more tests.

It is good you have eliminated the tubes, because a bad soldered connection to the grid cap on a tube, to the wire existing the glass bulb with poor soldering to the cap, can cause this sort of erratic behavior. So you are left with an intermittent component, bad soldered connection somewhere, band switch contacts etc.

One possibility; the opportunity for a bad connection is always present when there is any conductor or material involved where it is near impossible to solder to it, because the connection relies on mechanical crimping instead. As I recall with this radio, the tapped resistor where the fixed bias and delay agc voltage is generated is a nichrome wire resistor, it cannot be going open or the radio would stop, but possibly one of its tags has a bad connection to the nichrome winding.

Finally you can have what appears to be an intermittent fault, but all the wiring & components are good, if there are oscillations, coming and going in the RF or IF circuitry as this activates the AGC killing the gain. Usually if that happens in this radio it is due to mis-alignment and appears at the high end of the SW band and whistling around stations is also heard. To check for this, place a 100 Ohm carbon resistor in series with the grid cap connection of the 6A8, normally stops it.

Other possible options could include intermittently shorting turns in inductors, that is rarer. Silver mica problems. Obviously it would be wise to replace the original molded silver mica caps, but you probably did that with the restoration ?


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 1:21 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
I wish it had a bad or loose connection, I could find that. I thought of an intermittent oscillation, but the waveform at the 6K7 anode is good, if does vary sometimes with the changes out of the speaker, but sometimes not.
One thing I'm thinking of, is there are no film bypass caps, there's a 4uf to the 6A8, but I'm used to seeing a resistor like 2k, and a 47nf or so cap filtering tubes, like the 6K5 and 6K7, but there are none. I'm sure my German radios have these RC filters.
Makes me think it's some kind of filtering needed, when my multimeter, with the black to the chassis, just barely touches either side of the 99kohm resistor going to the 6K5 and the audio goes 2 to 5X louder. Remove the meter, and the audio stays. Sounds like the meter adds a couple of nanofarads and stabilizes the radio.

I didn't change the mica caps, might have to check those.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 1:30 am 
Member

Joined: Jan Mon 16, 2012 4:15 pm
Posts: 8323
Location: Near Brandon, Iowa
My favorite cause of inconsistent volume is the "sheared solder joint on the grid cap". Insidious in that it won't show up as bad on a tube tester (since it isn't in the tester long enough to heat up and then lose contact). Only affects "G" type tubes (the "coke bottle" glass envelope style); metal-envelope tubes and "GT" tubes with caps are not subject to this problem, at least as I have observed.

To fix: break the water glass bond between the cap and the glass envelope (usually the cap is loose anyway). Heat up the cap and lift it off with pliers. With the cap still hot, knock out all of the old solder. Dress the grid lead wire with fine sandpaper, tin it, replace the cap and add fresh solder+flux. You will need to do this for all three tubes with grid caps.

Plan B: you can cut off the tops of the tube shield bases that are riveted to the chassis and install GT or metal (RCA style) tubes if you don't want to fool around with fixing the G tubes: the G tubes have skinny bases which just fit into the shield bases- the other two styles are too big to fit without modding the shield bases. Also, you may need to extend the grid wires if you intend to use the original tube shields. If you ensure that the tube sockets have pin 1 grounded, you can simply not use the external tube shields (the GT and metal-envelope tubes have integral shields which are connected to the chassis via pin 1).


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 1:50 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
I thought I had dirty tube caps, so I cleaned them, didn't fix it. Everytime I touched the grid cap wire to the 6K5, the sound would change, I cleaned the tube cap and the connector, then I replaced the wire. Still didn't fix it, so I subbed in a metal 6F5, still had the same problem.
Last night I opened the second IF can, loosened the adjusting screws, and sprayed them with alcohol. That really messed up the tuning, so I left the screws loose over night to dry. I aligned the IF and for now the radio is working great. I can pick up all my stations with no antenna. Just have a whistling tuning on the high side of the stations.
I'm going to wait a while to see if it breaks, if it don't, I'm going to install a couple of 0.1uf caps on the B+ and call it good until it breaks again.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 3:29 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
I kept fiddling with it until I broke it. I didn't like the whistle on the high side of the stations, so I redid the IF alignment on 470KHz and it made it worse. I put the oscope probe on the 6K7's grid and the signal looked ugly. Then I laid the probe just over the grid wire and there's a very high frequency about 30-40MHz at 50mv that develops with little or no signal. I tried switching the 6A8, but it did nothing. I'll try the resistor on the grid.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 6:19 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1627
Tin Omen wrote:
I kept fiddling with it until I broke it. I didn't like the whistle on the high side of the stations, so I redid the IF alignment on 470KHz and it made it worse. I put the oscope probe on the 6K7's grid and the signal looked ugly. Then I laid the probe just over the grid wire and there's a very high frequency about 30-40MHz at 50mv that develops with little or no signal. I tried switching the 6A8, but it did nothing. I'll try the resistor on the grid.


You need to renew the mica caps.

The 38-7 has a habit of being unstable if the IF alignment and the two capacitors on the V/C for the osc and antenna are not spot on. As I mentioned, it has that interesting small coil on one of the IF transformers. If the IF remained unstable, after a formal alignment, and if it looks like anybody has tampered with the wiring there, try reversing the polarity of the two thin black wires.

The whistling indicates instability, try the 100R stopper resistor in the grid of either the 6A8 or the 6k7.

I know a lot about the 38-7, because I got my first one 49 years ago and have worked on it for most of my life.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 2:32 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
The only mica cap I found is component #8 a 3500pf micamold. The cap itself has a 3 dot code of purple-blue-purple or 760,000,000pf :shock: Well I don't have a 3500pf mica cap, have to find one. The schematic shows component #7B as a 250pf fixed cap, but it's 250pf variable.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 3:50 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
3500pf is apparently difficult to come by, Mouser has 1 at $88, ebay has 1 at 2500 volts, at bit too large to fit, but at least is cheaper.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 9:26 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
When you turn the radio on, and once warm, the 30MHz signal shows up, it could have been there since I recapped it. So I grounded a set of test leads and put a 33nf film cap on the other end. When I touched it to component 11, the 4uf cap, or component 11a, the 8uf cap, the 30MHz disappeared, so I installed a 33nf capacitor across both 4 and 8uf capacitors. It probably worked fine when it was new, but the new capacitors I installed aren't that good at filtering above 10MHz. I have a book Audio Valve Amplifiers where a EE is designing amplifiers. When he designs a power supply, he'll use a 100uf to filter hum, and a 100nf to filter higher frequency noise. That seemed to work for me.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Fri 24, 2021 9:59 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1627
Make sure the 3500pF is replaced with the same value, as I recall it is the SW padder capacitor (I'll check the schematic). You could use a 3000pF and a 500pF in parallel, 500V 1% dipped silver mica are suitable.

https://www.ebay.com/itm/191771736238?h ... SwGotWs-s7

https://www.ebay.com/itm/202804992741?_ ... %3A2334524

In both cases try to get the 1% parts, or buy a few 5% and check /select the final value.

The adjustable capacitor is usually pretty reliable, it is a custom Philco type as I recall.

The way they organised the band switching, a faulty 3500pF capacitor could affect both bands, because in MW mode it ends up in series with the adjustable MW padder. But you could consider keeping it, if it tests ok, as it is a pretty looking capacitor.

From what you have said your electrolytics must be relatively high ESR, at least at high frequencies, it never hurts to put 0.1uF caps across them for RF decoupling purposes.Most 38-7's have had their electrolytics replaced at this stage, but it might have been a while ago or they could be on their 3rd or 4th set of replacements by now, the radio was made in the 1939 era.

(the extra coil I was talking about in the IF can is in series with the suppressor grid of the 6k7)


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Sat 25, 2021 1:44 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
The electrolytics I think I used are Nichicon UCA/UCY/UCS, they are high ripple current, high reliability, but tiny compared to the old ones. I did change them from the originals. I still had a whistle at the high end of the stations, I have a bunch of 1/8 watt 1K carbon comps, so I put one on the 6K5 grid and the whistle went away. Right now it's totally fixed, sounds great; volume is nice and loud, multimeter doesn't affect the sound, it is so much better. I'ma test it a while longer and put it back in.


Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Sat 25, 2021 3:09 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3082
Location: Cortez, Colorado
Thank you Acornvalve and everyone else. The radio is playing great. I refinished it last week, I'm waiting on the Philco decal to complete it.
About 4 feet away I have a Philco 91X, it works great too. I bought it back in March near Charlotte, NC.


Attachments:
Philco 38-7.jpg
Philco 38-7.jpg [ 311.76 KiB | Viewed 358 times ]
Philco 38-7-2.jpg
Philco 38-7-2.jpg [ 237.13 KiB | Viewed 358 times ]
Top
 Profile  
 
 Post subject: Re: Strange problem with Philco 38-7
PostPosted: Sep Sat 25, 2021 12:42 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 10867
Location: Beaver Falls, PA. USA
One thing to remember on prewar Philco radios is that the company had their own internal color code for mica capacitors.The code on the capacitor has no relationship to the actual value; one must go to the service literature for the correct value.

_________________
Tim KA3JRT


Top
 Profile  
 
Post New Topic Post Reply  [ 14 posts ]  Moderators: Norm Leal, Marcc

All times are UTC [ DST ]


Who is online

Users browsing this forum: dirkbuddy, Google [Bot], Marcc and 16 guests



Search for:
Jump to:  




































Privacy Policy :: Powered by phpBB