About 20 years ago the Kriesler 11-81 was one of the most common radios a valve enthusiast could find. Now, like the rest, they have largely disappeared.

The valve line-up is 6AN7, 6N8, 6BD7, 6AQ5, 6V4. It has a ferrite antenna. Despite this, they are only an average performer, with not a lot of volume.

There's 2 modifications that can improve the situation.

In my collection I noticed a pair of them with melted cases. One case was missing some bits of plastic, with holes and other disfigurement. One is brown and one is dark red.

The brown one worked but cut out after a while. The red one started up, then became very distorted and low.

So, the brown one was looked at first. I couldn't see anything obviously wrong - until I bumped the 6N8. Then it cut out and was very touchy. There wasn't any dry joints, so I supposed it was a cheap valve socket playing up. Given its generally poor performance and ruined plastic case, I decided to scrap it.

So, to the red one. The first thing I noticed was that the negative wire from the transformer to the 6AQ5 grid circuitry had fallen off. Further investigation showed that the 6.3v heater wires on the transformer were both dry jointed. Then, the 270 ohm resistor was also dry jointed both ends and open-circuit. This is why the bias voltage was over -40 volts and rising. The 680k bias resistor on the output valve measured 820k. So, all this was fixed and we had a working radio. It was quite a bit more sensitive than the brown radio, so I decided to keep it. There was only one knob, broken, so I transferred the knobs from the brown radio over.

In the end, one average radio and one scrapped radio.

The tone control on these radios barely does anything, and being in the feedback path causes loss of volume. A modification can be done to fix both problems at one go.

The other mod involves replacing the 6N8 with a 6BY7, giving much improved sensitivity and even more volume. You have to rewire the valve socket though. Once this is done you can experiment with other even-higher-gain valves such as 6BX6, 6EH7 and 6EJ7. However I found the 6BY7 gives the best improvement without becoming unstable.

Now, somehow, I've run out of solder. I thought I had enough to last me for life. Oh well, a trip to Jaycar is coming up then.

For whatever reason I’ve got three of these sets so I’m interested in the process of these modifications.
Two of the sets I’ve restored but as you say, average performance.

The first one I fixed had a huge wasp nest on the front and was in a shed. Cleaned up really well. Was just dirty really. Bit of car polish brought the shine back. Other than some scratches it’s not bad.

I've got at least another 4 of them - off white, stale custard, bluish-grey, and another brown one. The bluish one is modified and is in the kitchen, and is used most days to listen to 2UE.

I'll post the mods shortly.

As promised, here's the details from my notes. But you should check that I haven't made a typo.

1. Fix to tone control. On some radios, turning down the tone control results in instability (motorboating and squealing).

- The cathode of the 6BD7 needs to be earthed, so remove the 470 ohm resistor, the 0.047 paper cap, and the 12k resistor in that area, and directly earth pin 3.
- If you look at the back of the tone control, there's a green wire to the speaker, and a bare earth wire. Cut this earth wire.
- The green wire from the speaker goes to the left terminal of the tone control.
- The centre terminal of the tone control goes via a 0.047μF capacitor, to the grid terminal of the 6AQ5 (pin 1).

This gets rid of the instability, and the control is much more effective.

2. Replace 6N8 with 6BY7.

- The above mod must be done first.
- Remove 6N8
- Remove wire from 6N8 pin 7 to IFT2.
- Now, the 6N8 was being used for audio detection, so we will use the 6BD7 for this. The wire that went from IFT2 to 6N8 pin 7 now needs to go to 6BD7 pin 8 instead. So, remove the existing earth from 6BD7 pin 8 and join a wire to IFT2.
- Disconnect 6N8 pins 1 and 6.
- The wire that was going to 6N8 pin 6 now goes to pin 7.
- Move screen grid resistor and wire (was on pin 1) to pin 8.
- Earth 6N8 pin 6.
- Plug in your new 6BY7.
- Switch on, make sure it works.
- Retune IF.

Thanks Robert, I might get around to trying this out with one of them now I've finished exams

A 6EH7 could also be used in place of the 6BY7, it's pin for pin compatible but has even more gain.

Yes, I've tried 6EH7, 6EJ7, 6BX6, 6BW7 and 6BY7. However I found some instability sometimes with the very high gain of some of these. That's why in the end I settled on the 6BY7. But you might have more luck than me, and there's certainly no shortage of these others.

I think there's others with the same pinout. I seem to recall that EF89 might work too. There's a lot to play with, that's for sure.

Interesting as to why you interfered with the 6N8 in the IF. Despite RF floating around in the area of the detector I would have perhaps considered increasing the audio only. perhaps a triode pentode

I have aside from increased shielding in the detector area found the addition of an RF choke like many Plate detectors in Autodynes, tends to reduce the audio from amplifying stray RF & regenerating it, destabilising the whole show.

Watch the current draw sometimes higher gain comes at a price.

I did some other experiments at the time too. For example, if you remove the 6BD7 and simply jumper across (making it a 4-valve set), the volume is almost the same. The 6BD7 in the original configuration barely does anything.

What could be done, although I didn't try it, is to replace the 6BD7 with a pentode, such as 6N8, 6AD8 or 6BH5. That might help.

Hi Rob, just a thought, having mucked about in design and changing valve gain types in RF and IF stages
:
How does the AGC relate in these Kriesler chassis?
Reason, if you replace one valve with a higher gain, then the AGC feedback should just throttle the overall gain back to where it was and no change results.

That assumes the stage is hooked into the AGC loop of course.
In a good design the AGC loop determines the overall gain and I have watched an IF strip output remain constant even when manually forcing say the convertor valve up and down in gain.

Mind you it is a big jump to whack a 12EH7 into anything!
Those things have heaps of gain having used them in all sorts of radio stage designs.
Great valve, I just needed to watch the lead dress in a stage and keep the grid and plate ciruits well away from each other!

Cheers, Fred.

I'm sure you meant 6EH7...

The AGC argument seems sound, until you think about it. What if you used a valve with no gain at all? That argument says the AGC should adjust itself to as if a normal valve was there. No, more gain means more gain. It will be throttled back a little, but not back to what it was.

The real issue is to stop the stage from self-oscillation, which might need (as you said) correct lead dress, or even putting the valve in a metal can, like so many TV sets did.

OOPS, yep I did mean 6EH7! Getting old in the head.
I bought of box of 6EH7 and 6EJ7 from a guy at $1 a pop, 30 of each!, and used them for all sorts of things.

If you have no gain in a valve, say x1, then the AGC drops to a lower value and the other valves in the loop go to high gain and make up the AGC as far as their slope allows them to. Obviously there is a limit.
If you break the loop, say pull a valve out, yes then that is no gain ( zero) and no AGC and silence!
I am not theorising here, when playing with front ends I draw up the AGC curves for all sorts of conditions and try to get max gain, level output and no screaming at 100MHz!

Good to see someone trying things in the engine room and actually doing something!
Fred.

Yeah I meant x1 gain, not removal of the valve, sorry about that misunderstanding.

With the earlier repair jobs I sometimes experimented with AGC resistors and so on, the object was to tune to a very weak station and adjust the resistor(s) for best reception. Then tune to a strong station and make sure there was no distortion and the volume level was not so much louder. It was amazing how much more sensitive a radio could be with a new resistor or two.

For proper 100% AGC though, you really need some active devices to detect and control it properly. Consumer valve radios never had the cost base to allow for that though, you'd have to look at communication / ham radios to see how it's done.

As for the engine room, I think you're the expert in that. Your knowledge of matching up coils and so on in the front end surpass me totally. I can only pass on what I've found by trial and error.

HMV had a habit of making sets properly & one of the weirdest is the R53B R43B the Radio Club has R54 . As a moot point it uses 6N8 as the first IF & detector a second as AGC (I would suggest delayed as its coming off of the previous plate) & first AF then a third 6N8 as the driver for the #807 Output valve.

This pumps quite a bit of noise into the display area when used there.

Turntable is a horror story and a design abomination.

I agree that you are the engine room Fred. I did a one year course in electronics (including valves) back in the the late sixties or early seventies, a couple of hours a week. That is all very well if you then go ahead and apply it and get the knowledge into your bones, so to speak. But that wasn't the case. If the testing machine I was responsible for in those days had trouble with its valve circuitry, I would call the specialist electronics bloke to come and fix it. Really, for most people, that study was to learn which specialist to call.

We also had a specialist mechanical devices bloke and a glassblower. So if someone needed a specialist bit of chemical or measuring gear these specialists would put there heads together with the "customer" and come up with something. All of these blokes are gone decades ago.

So I have some basic knowledge but not at my fingertips - have to look at bibles like the Radiotron Designer's Handbook to fix things in my mind again.

This handbook mentions somewhere that the final stage of development of a radio design was finessing various capacitor and resistor values to tame whatever problems that it has. So even experienced designers could not guarantee that a design from scratch would work. Inevitably time would catch up with them and probably the circuity of many radios is a compromise forced by commercial considerations. Fred can sort these out, but I can't.