I'm having a bit of a problem with a Kriesler 11-25. When I tune in any station particularly at the low end , I get a whistle either side of the station. I've done a complete recap as well the RF bypass caps, changed any above spec resistors, replaced the 6AN7 and 6G8G, I even tried a 6B8G with no change, and a complete alignment . There was a 2 ohm reading from the shield to chassis which I fixed. There is no IF on the 6G8 anode. Double checked lead dressing, all components were fitted in the same orientation. It's got me beat.
I would appreciate it if some could explain how this reflex circuit works, maybe it's got something to do with it ?

Reflex sets like that do have two IF transformers and the 6B8 will have RF on the plates or the 2nd IF would not work. How did you determine that there was no RF on the plate? The one tube amplifies bot RF & audio.

That, like several in that 11-7 series had more changes than a nappy. A whistle means that you have feed back, or its picking up RF from another source. I would be tempted to probe around with a chopstick & see if you can change the frequency when you move wires etc. If you can its feedback.

New modern caps can cause instability as they have not got outside foil like the old ones and induction and radiation, particularly in the area around 6G8 and other 2nd detectors is common. Often I have had to add shielding and shielded wire.

Alignment data is not on many of those circuits however, if the there is a BC padder is is normally set at 600kHz first. If you calibrate the signal generator with a counter: The modulation / Tone should be off. Too much signal into the Pentagrid will cut in the AGC and you will not get the alignment right.

Should have mentioned, I removed the 6AN7.

"Should have mentioned, I removed the 6AN7." Why??? For what purpose?

If it was for alignment this is not an Autodyne with separate exciter.

You inject the calibration signal into the signal grid which is grid 1 of the Hexode that will be connected to the tuning gang. I normally use a blocking cap of around 200pF although I now have built an attenuator box that couples the generator, counter, and one of the oscilloscopes. That remains as virtually a permanent arrangement.

Hi Rudolf, this whistle: is it high pitched as you tune in and go down in pitch to nothing when the station is tuned in, then come up in pitch as you tune out?
If so its a beat effect between two "carrier" signals, that should not be near each other.
Fred.

A member of our local club suggested that by removing the 6AN7 I should be able to see if the IF is oscillating on the anode of the 6G8.

Marcc, I'll try your suggestion with the chop stick, shielding a couple of wires and replacing a couple of the RF filters with silver micas.

Fred, that is how it sounds.

The idea of the chopstick is the same idea, if wires are away from the chassis they can be antennas or radiators. Capacitors can induct: Some worse than others.

If nudging a wire, or component, changes the pitch then that is normally the problem area. Sometimes I have found the oscillator to be on the wrong side. Low not high.

I have had whistles in the mixer in a hacked Autodyne; discovered by an oscilloscope, being used to find out where the signal got lost? Factory lead dress fail: Rewired in a more appropriate manner

Rudolf, put simply the reflex works as follows.
The 6AN7 has the station carrier reduced to IF frequency with any audio modulation coming out of its plate and into the 1st IF.
This signal is applied to the control grid of the 6B8.
The 6B8 amplifies this AM 455KC RF signal to a higher level and its plate drives the 2nd IF.
The second IF applies the IF signal to the detector diode and AUDIO is left (by filtering the IF) at the volume control.
So far just like any normal set.
Next instead of heading to the output valve, the audio signal is applied to the bottom of IFT1 and flows up to the control grid of the 6B8.
The audio signal is then amplified by the 6B8 and that audio appears at the plate of the 6B8 and flows through the primary coil and appears across R177.
That point at R177 is connected via C27 to the control grid of the 6M5 and that amplifies the audio as usual to the speaker.
Thus the 6B8 has 2 goes at the signal, first as a modulated IF signal and then second as a audio signal.
The frequencies are sorted out by the IF coil trapping and directing the IF signal and the resistor trapping the audio, the RF being filtered out by capacitance "top cutting" the IF signals out.
Does it work well, no, sort of, the 6B8 has both signal streams inside it and does not do either job at the best efficiency. You will have RF where you should not and as Marcc points out a dead sitter for leakage coupling causing all sorts of unfortunate problems.
Not as stable or as efficient as using a separate valve for the audio pre amp.
Hope that simplified explanation helps.
Fred.

Basically it was to cut out one expensive valve. Price was important.

That was still used differently in sets like Thorn RG106 where reflexing tube was a 6N8 and that fed (stereo) 6GW8's (Triode Pentode) and you still actually had four physical tubes, but better audio quality.

It is probably the new parts and their physical position that has brought this one unstuck.

Marc

Thanks Fred, I'll read it and digest it, in the mean time I'm going to order a selection of silver mica caps from Just Radios, as I want to eliminate the new RF filters which were ceramics because I think Marcc could be spot on with them being part or all the problem.

Some disk Ceramics can pick up noise, but leakage was rare in many ceramic caps. There was a group I believe had a red spot on top of them & were common in transistor sets.... They were bad news.

Albeit that with the effluxion of time some parts deteriorate, one must be very careful to ape the original circuit. Changes can introduce unforeseen problems. Consensus dictates that a set worked when it left the factory and unless there is a known problem and factory mod. One stays with the original schematic. It is amazing how well most sets work when put back to as close as possible to original specs.

I still think it wise to do the chopstick test to see if you can ascertain the smoking gun, before spending money. It may be a simple shield or move something.

Just finished reading Chapter 28 of Radio Designers Handbook, learnt a lot.
I'll do the chopstick test in a few days as my mind is elsewhere at the moment, next week is my 6 monthly check following radical prostatectomy operation 3 years ago.

Good luck with that - I am on the terrorist rating at the moment - be alert but not alarmed pending next PSA following detection with a biopsy.

Re biopsy for those facing one, not as big a deal as I feared. Just a little bleeding which is normal.

My PSA reading for the first 2 years was less than .01 which was great, 6 months ago it was less than .03 not so great, hope it's not any higher on Monday.
At .1 treatment such hormone or radiation begins.

Rudolf, I had a radical in 2011, showed <0.01 for a while, that drifted up in 12 months to 0.02 by 2012.
Then I had 30 rads in 2012, result <0.01 ever since. I check PSA every 6 months now.

So that's 7 years or so at <0.01 so cautiously I think should stay that way.
There was no additional damage by the rads (done on latest machine at Westmead for 2012, very targeted) I would think 2019 machines would be better.
I had bladder damage from the surgery ( I instructed the surgeon to take wide margins) not the rads.
That damage was rectified by fitting a UCS, (urinary control system) so now I can pee on demand not dribble like an old steam engine.

I know my experience is very individual but it can be good.

Fred.