Oscillation in STC 830 radio
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
Started with need to open IF cans because of loose internals and very stiff trimmer adjustment screws. At this point the radio MW band tracked perfectly and picked up distant stations well ie good sensitivity. SW1 and SW2 working OK. I previously had left the IF cans strictly alone because they were bound to be fragile, but because of the amount of work done on the radio including valve changes, rearranging the screen grid supply potential divider and removing and refitting the tuning gang, it was time.
Inside the IF transformers looked good except that 2nd IF had two frayed (not broken) litz wire leads. Wax cracked here and there with a little dust penetration around the trimmer adjustment holes. Frayed leads repaired endeavouring to get all of the litz wires soldered. Also removed trimmer adjustment screws and lubricated them with teflon spray, let dry then reassembled. Made adjustment easier. This after counting number of turns to fully closed - returned to setting after reassembly. The same lubrication of screws for 1st and 3rd IF trimmers.
Will post a photo of repaired IF2 transformer.
Trimmer micas looked OK, just needed some dusting, but didn't disassemble the trimmers. Trimmer metalwork appears to be nickel plated so no silver migration expected.
Loose internals fixed by tightening the nut that secures the ceramic of the trimmers to the top of central fibre tube of the transformer.
Replaced deteriorated top cap leads of IF1 and 2 with same length of new wire.
After this above, following symptoms appeared:
-oscillation (whistles) at threepoints between 640 - 680kHz. Whole set microphonic when whistles occur, but appears to be centred on 6A8G mixer.
-silent below ~640kHz, indicator tube showing no signal
-above ~680kHz tuning indicator shows low signal off station
-picking up strong local stations only above ~680 kHz, ie loss of sensitivity
-SW1 and SW2 appear OK
Alignment was now way out, but with adjustment of IF trimmers to maximise noise at low frequency in SW1 since I have no low frequency MW to work with, the strong stations above 680kHz on MW band were brought in a little stronger. 3rd IF adjustment screws need to be right out for best signal. Tracking not good and erratic, eg 1431 OK, 1314 about 1340 (26 high) and 1206 about 1220 (14 out).
Have checked the continuity of the IF transformer coils in the transformer and where they connect into the circuit- all OK. Also swapped out all IF valves (6U7s), the 6A8 mixer , 6H6 detector and 6U7 RF amplifier valve. No difference to symptoms. Also bypassed IF2 by connecting IF1 top cap lead to IF3 instead of IF2 6U7 IF valve - same symptoms. Opened up the IF cans again and carefully checked out the transformers and trimmers, and reassembled, checking earthing of the shielding cans undisturbed since 1989. Cleaning micas would require unsoldering the trimmers which I won't do unless necessary. After all, work on the IF cans & transformers the only work, apart from trimmer adjustments, done on the radio since it worked OK.
Whistles can be turned down with the volume control so does not appear to be an AF problem.
I suspect gross misalignment could account for these problems though I tried to avoid it. Or some other event that has happened coincidentally eg a shorted turn in the repaired IF transformer or a dry joint showing itself. 3rd IF trimmers are furthest from the original setting which is suspicious. Or maybe repair of frayed leads has improved the gain leading to oscillation.
Any other ideas gratefully received.
Look in https://vintage-radio.com.au/default.asp?f=12&th=120
for circuits.
|
|
|
|
Location: Toongabbie, NSW
Member since 19 November 2015
Member #: 1828
Postcount: 1313
|
Hi STC the IF section needs to be tested for active sweep response.
The IF strip is the heart of a Superhet and has to have response and gain over the specified IF frequency.
The whole tracking and performance of the set depends on the IF strip response having adequate gain and response width.
Adjusting cores or trimmers for a sound level is not the way to go.
The AGC and local oscillator need shutting down and then a sweep signal applied to the convertor grid.
The response can be read after the detector.
An appraisement of the response and gain can then be made.
Sorry to sound so abrupt.
This is a technical problem and needs a technical approach.
Guesswork has no place when faced with this sort of problem.
Once a deficiency is found then the cause can be looked for.
Cheers, Fred.
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
Don't disagree with you Fred but I don't have a sweep generator. I have always depended on doing one thing at a time, making things work OK and moving on. But am caught out this time.
|
|
|
|
Location: Hill Top, NSW
Member since 18 September 2015
Member #: 1801
Postcount: 2079
|
Looking at the schematic, that radio should run like a rocket, with its 2 IF stages and the RF stage. The 6A8 is not exactly the best mixer for short-wave performance, there's lots of better plug-in substitutes, but leave it until you've sorted out the issues.
The schematic is missing the earth (or AGC) for the common of the RF coil secondaries.
If I'm reading your comments correctly, it seems it was working until you decided to have a fiddle - is that correct?
Without being hands-on, there's not much I can suggest. I'd probably return the slugs and adjustments back to where they were (or where they should be) and look for another cause. Obviously you've bumped or damaged something, probably a wire or solder joint somewhere.
|
|
|
|
Administrator
Location: Naremburn, NSW
Member since 15 November 2005
Member #: 1
Postcount: 7402
|
Photo uploaded.
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
A valve a day keeps the transistor away...
|
|
|
|
Location: Toongabbie, NSW
Member since 19 November 2015
Member #: 1828
Postcount: 1313
|
Hi STC, dont have a sweep gen?
Make up the SC "minimite" module, see special projects.
Handiest bit of kit since sliced bread.
Then you can sweep any AF RF or IF stage and with any old CRO read the result.
Fred.
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
Robbbert: It is (was) quite sensitive to the degree that I haven't built an external aerial for it but listened to MW on it (mostly cricket where it would get a wicket for Australia at opportune moments) with an untuned loop antenna. Short wave I never bothered with apart from making sure that it worked every now and again with 20 feet of wire strung around. Have plugged in other mixers but tracking is affected so didn't stick with them since I mostly listen to MW. Neither RF nor OSC secondaries are earthed - isolated by capacitors. Will investigate and see if this really is the case.
It has much fiddling over the years - this is the second restoration after the first in 1989. But have left the IF, aerial, RF and Osc coils strictly alone since they were working and would be fragile. And first time I open up the IFs, trouble. No slugs, all cap trimming, and all except the 3rd IF are about where they should be. I think it has to be something to do with IF transformers, especially the 2nd (repaired frayed leads internally), or due to removing and lubricating the trimmer screws with dry teflon spray. If it is not these causes then it has to be some coincidental fault developing and/or way out alignment.
Brad: Thanks for the photo
Fred: I was looking at the minimite last night. Looked at it some years ago and couldn't figure out Silicon Chip's ordering system and what exactly to order. Then medical matters intervened and didn't get back to it. My CRO really is any old CRO, with a distorted display, but no doubt would do the job.
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
Robbbert: Re your post #4, The RF coil secondaries do indeed have a resistance to earth of the order of several to a few ohms, so they are earthed even though as you point out it is not shown on the schematic. And it is there bold as brass, or rather tinned heavy copper wire from a transformer terminal straight to chassis. Don't know what this means, but this wire has been there since the year dot.
It might mean another STC draftsman error; there are a few others eg 6T5 called 6T7.
The 6A8 is not connected to AGC because it would even further mess up SW performance.
|
|
|
|
Location: Toongabbie, NSW
Member since 19 November 2015
Member #: 1828
Postcount: 1313
|
STC, in the meantime you could assess that the IF transformers are resonating.
You could shut the 6A8 oscillation down, ( short a winding) apply a IF frequency to the 6A8 grid.
That means the IF strip will be supplied with an IF single tone signal at the right impedance drive.
The level may need experiment to get the range of level right.
If we assume the AGC level is normally in the range of 5 to 10 volt set the input to get that sort of level.
Then use the AGC level DC as an output indicator.
Stay away from hooking any instrument near the IFT's, just read input level and output level.
Then, Each IFT core adjuster should be able to show a peak either side of resonance.
Of course, the AGC feedback should try to defeat any change, once you have common sense response you may ground the AGC line.
Then you may have to detect resonance in a different manner, like read the DC level of the detector output before the AGC point.
Each IFT winding should show a resonance, the last detector coil; may be flattened by the diode loading and not as sharp as the rest.
If an IFT refuses to resonate, show a peak, then find out why.
Ita a bit long winded but can point to any IFT coil or resonating capacitor problem.
another very sensitive indicator to a valve working on its AGC feedback is to watch the screen grid voltage.
As the valve is throttled back in gain the screen voltage should rise, say from 60 volt on full gain to 90 volt when biased back. That can give you a clue as to watch each stage is doing (so long as each stage has its own screen feed).
Each stage should amplify, resonate and gain control.
The oscillation may have a root cause if there is an IF fault.
Try clipping a 0.1 cap across each HT point, cathode point, screen grid point if you suspect any bypass.
clip a lead across each can or shield to ground while it is oscillating.
There is a lot of trouble shooting that can be done but a lot of it is instinctive guided by what each valve does in relation to what it should do.
Good luck.
Fred.
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
Thanks for those detailed instructions, Fred. Will attempt them when I have redone the relevant valve soldered joints.
Regarding screen supply, RF, mixer and IF valves have the same supply and currently hold around 90-100volt, but I need a full voltage survey in both normal operation and while oscillating. I have eliminated long leads from the voltage divider by reducing its size and location, but want to eliminate the divider as a cause by shielding it somehow, because this is a major change from the original circuit.
|
|
|
|
Location: Wangaratta, VIC
Member since 21 February 2009
Member #: 438
Postcount: 5401
|
Not sure as to how you went about this but there are traps. A photo of it in situ before messing with it is a good idea as is using the same coloured wire. It is easy to wire it backwards. The original position of the trimmers should have been photographed.
Just changing the wires should not have messed it up. wiring it backwards & messing with the trimmers will. ensure that the grid is not wired to B+ in the previous tubes plate circuit. The grid circuit will possibly have the coil of most resistance and will normally go to a grounding.
I normally line those cans internally with insulation paper. The trimmers can short especially if the wafer breaks, or has water in it. Spray joints with circuit board lacquer to reduce corrosion.
With the fly lead, generalising depending on the terminal, shrink tube reinforces and also prevents the TC terminal shorting on some cans. and stops fraying of the sleeve.
As that is a common screen arrangement did you actually put in the right decoupling cap? or is it perhaps missing. It is not uncommon especially in transceivers for decoupling caps on every screen to get rid of standing waves.
Worst offending area for post refit instability is detector 1st AF area lead dress. Probing with a chopstick can change the frequency in the offending area.
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
There are photos of the changes wrt the voltage divider in
https://vintage-radio.com.au/default.asp?f=12&th=120
Believe me I have gone over the IF transformers with a fine tooth comb; but the leads weren't broken when the can was removed, just frayed, and repaired one after the other.
The trimmer screws and their washers were removed, the screws lubricated, dried and replaced with the washer. Both trimmers for each transformer are on the one ceramic base. The wires from each winding go to their own trimmer. Will post some photos. Shorts of trimmer terminals to shield could occur, will investigate.
As Robbbert points out, the RF transformer secondary is not shown as earthed on the schematic, but it is in fact earthed. STC schematics show the shields as a dotted line and the winding earths are sometimes shown made to this dotted line. The dotted lines are very faint in the STC 830 schematic so the earth point doesn't show.
There is only one decoupling cap on the screen wiring, being C7 0.25μF on one of the terminal boards of the radio. This is some distance from the screens being decoupled so maybe this is something to look at by adding further screen caps.
I pulled out 1st AF 6J7 - oscillation still evident as indicated by deflection of the indicator valve - display fully closed. So I don't think it is 1st AF causing the problem. But that said, the 3rd IF trimmers need to be screwed right out for strongest signal. Originally they were about half position like all the others, so something might be going on near the detector; pulling out the 6H6G detector however also does not stop the oscillation as indicated by the indicator deflection.
|
|
|
|
Location: Wangaratta, VIC
Member since 21 February 2009
Member #: 438
Postcount: 5401
|
With top trimmers there are traps. Often the top will be staved in and that can lead to shorting., Cans not properly grounded, can cause havoc.
The penalty for messing with IF cans can be a wire break on a node.
|
|
|
|
Location: Toongabbie, NSW
Member since 19 November 2015
Member #: 1828
Postcount: 1313
|
Hi STC,
" But that said, the 3rd IF trimmers need to be screwed right out for strongest signal. Originally they were about half position like all the others"
That says it all!!!!!
The 3rd IF is NOT resonating.
Substitute with another IF of the same frequency or find out what is damping the dud IF.
Marc has outlined many common problems.
Fred.
|
|
|
|
Location: NSW
Member since 10 June 2010
Member #: 681
Postcount: 1303
|
OK Fred, thanks for that. I will see if I can pull that valve and bypass the whole section from IF2 and see what happens.
|
|
|
You need to be a member to post comments on this forum.
|