I'm currently, slowly, trying out all my radios, one at a time, to see what still works and what doesn't. This involves tuning to 2UE and let it run for 6 hours or more.

I've noticed that every Kriesler 11-99 is subject to noticeable drift, given time they drift right off the station. Is this something others have encountered? Is there a solution?

This is not unheard of but can be exacerbated by resistances getting hot & volts on the Pentagrid changing: 6A8, 6A7, & 2A7, (same tube, different heater & basses) being the worst. Radiotron manual will tell you that most of the Pentagrids suffer from changes is voltage, but will normally drift upwards in frequency until they stabilise & reach their operating parameters.

Being voltage sensitive in a big city, they are going to have issues at meal times. The only way to get improved stability is to regulate the power to the Pentagrids screen Plate & Oscillator, if its a problem. Some precision stuff also regulates the heaters.

Need to check current draw but an LR8 may handle the one valve on its own, or shunt it with a TIP 50

Marc

All of my valve radios will drift to one extent or another. I haven't noticed that the 11-99 is any worse than the others, however, I never run them for anything like 6 hours, especially nowadays since 2CH went sport.

I guess it's due to component value drift, probably the result of heat.

The thing about the 11-99 is that the station dial is directly connected to the tuning capacitor's shaft, making it fiddly to get the station bang on compared to my other sets where some form of gearing is involved. The 11-99 is more like a pocket transistor set in that regard.

2UE is fine as a 2CH replacement, and it has less adverts.

Maybe I should connect a meter to certain points and see what happens over time. The frequency changer is a 6AN7 which is usually quite stable. The schematic shows the anode voltage is only 80 volts so I suppose it's lucky that it works at all.

Ditto with GTC .11-99 sits beside my bed tuned to ABC. Fires up every Sunday morning and bang on station after 30 seconds warm up, runs
for 20 to 30 thirty minutes while I listen to Macka then off for another week. Always on frequency from power on to power off, week after week.

Robbbert, for the purpose of the exercise, (and curiosity) next Sunday I will let it run overtime and post my report.

In any sort of analysis one try's to eliminate the variables. In accordance with Murphy's law, these variances will all be cumulative.

I will start a repaired radio, with an analogue meter tied to the "B" supply; This to note if it is doing as it should. It is never unusual for a supposedly new dodgy, or underrated filter cap, to get to a voltage & then let go. That is why I put the ones that have been sitting awhile, onto the reformer before using them. IC clips are good for both of these jobs. Record B+ voltage over time.

Ventilation plays a part as well, transformers and resistors all generate heat, not just the valves. Do Not overlook the AGC it can destabilise some pentagrids and if there is a fault , which should be obvious, it fluctuating will not help.

As noted fluctuating line voltage is easily dealt with. There are valve regulators. I have one in the bench power supply and the mains power supply for a Philco BC-221-N also has one in the HT (150V) That is originally battery.

Obscure is the dial drive. The tuning cap is often rubber mounted & if it can creep, due to a tension issue it will move.

Thunderstorm over head. Chain lightning. Shutting down. UPS died again so there is only my system to stop the surges.

Marc

Hi Robbert

Not a pentagrid issue, the 11-99 uses a 6AN7. They are known to be stable.

Not paper caps, only the very early 11-99s still used them.

I'd be suspecting C5 since it's such a simple circuit and there's nothing else there to cause frequency drift. Not the padder, there isn't one.

Is it a sudden or a slow drift? If C5 is a mica, you'll have a case of silver mica disease. Causes sudden jumps and crackles, usually.

If it's a ceramic, it may have the wrong temperature co-efficient. That will drift slowly with heat.

Hi Ian, it's a slow drift. So far, I've tested all bar one of the 11-99's in my collection - I'll test the last one next week. All the tested ones drift.

Apart from the drift, they work fine.

Just as a matter of interest I located my two 11-99’s , repaired/restored some years ago.
Both have been partially recapped, but retain the original mica C5, 50pF. And have not been turned on for years.
Both were turned on about 4 hours ago and are still rock solid tuned in to local radio station.
Will keep running to get some more hours up, perhaps even throw a cloth over them.
So far so good.
I think if one of mine would drift, I would as Ian says replace the C5 mica to start with, or maybe a can of freeze.
JJ

Just checked again, both radios tuned in rock solid after approximately 9 hours running.
JJ

Thanks for checking. It's not exactly a complicated design, so it shouldn't take too long to track down the culprit.

The most likely issue will be in the oscillator circuit caps like C5, I check for leakage (rare in Mica) at 500V with an insulation tester.

As noted B+ fluctuations & resistors heating cannot be ruled out. fluctuating current draw can be gleaned from the back bias resistor R12.

Is C5 a mica? Kriesler did quite a few parts subs in that model.

As I have said previously, the unusual use of the 6GV8 was made partly to use up stocks of them that failed test in TVs and which Philips would not replace under warranty. End of production for the 11-99 followed not long after the end of valve B&W TV production, i.e, the introduction of the 49-7 all solid state TV chassis.

All 3 of the radios that I could find (I thought I had more, but they must be hiding) drifted noticeably. Turned out the cause was the same in all 3 - a manufacturing oversight. The 15k resistor that supplies the screen voltages to the 6N8 and 6AN7 was physically lying on the mica capacitor, and so as it slowly heated up, it caused the cap to drift. Moving it away seems to have fixed the problem, although I need to test a bit more.

The first 2 radios have a 6V4 rectifier, and real mica caps. One of them has a whole bunch of wax paper ducons too, so it will need a recap, although it works ok. On this radio the chassis is held to the front plastic panel by 5 tiny self-tapping screws. The plastic breaks up allowing the case to shift about, which adjusts the tuning knob! Something to watch out for.

The last radio is much newer, May 1974, and comes with philips caps and no mica ones. The replacement for the mica is a tiny tubular thing, but it was still placed right next to the resistor.

I think you might be right about the rejected 6GV8s - 2 of the radios have noticeable distortion. I have a bunch of spare 6GV8s, so I might try doing some replacements. I believe the 6GV8 was not a very reliable valve anyway (in its intended TV role).

There may be cause if there is an oscilloscope and a function generator, to use them and after checking with the oscilloscope that both the Signal Generator and function generator have clean signal, apply them.

One of the reasons I calibrate using an oscilloscope is that if there is distortion, in the audio, it will be seen. A Generator in conjunction with an oscilloscope is a formidable combination for finding lost signal and the origin of distortion. The biggest issue with an oscilloscope is interpretation of what it is showing you.

I note that that particular tube is driven fairly hard. With 75mA cathode current quoted as a deflection amp. 6N8 also had its moments.

The shared screen bias was always interesting. In HMV nippers the RF/IF screen resistor was two 22K resistors paralleled. The bean counters must have seen it & later it became a single 10K. Current squared by resistance prevails, as the current draw is over one watt.

Do check that one, & make sure that the Kriesler has not had a similar change, or the wrong wattage resistor is in place?

Stability can only be considered from the point where all parts have reached their operating temp.