Hi Dave, when I look at the circuit, yes, the bottom end of the diode load is hooked to the cathodes.
That is a feedback loop of the diode DC voltage into the bias circuit implying some sort of AGC control.

The concept of the set seems to be.
1/ the KT61/EL33 with its high inherent gain would give full output with mV audio input, that sorted the audio out.
2/ the 6AR7 while its not a high slope valve (like a 6EH7) , is said to be high gain by radiotron so is "better" than say a 6U7/6G8 as an IF amplifier.

Thinks: this is the last gasp of the octal valve design with the large envelope, example the X61M convertor. Radiotron was heading towards phasing out octal and moving to the noval base designs.

3/ the coil set is critical, those three coils would determine the gain and selectivity more than the valve type.
Radiotron would have been trying to sell their valve type to S/C who probably would have prefered to import the european EL/EF types of valves being cheaper.

A very interesting design and a preamble to sets using 4 novel valves like 6BE6, 6X4 ect.

If I get enough motivation I'll hook up a 6A8 and a 6AR7 to some coils to that circuit and see what they do!

Fred.

One thing to note is that the volume control is a special job that leaves some resistance in circuit even at max. If you short it out, the radio stops working. Sneaky!

I compared the gain of various output valves and yes there's quite a difference, and using a 6V6 is a mistake.

6F6 = 2500 umho (at 250v)
6V6 = 3700
EL33 = 9000
KT61 = 10500

However in practice, to my ears the 6V6 sounded much the same as the EL33, which goes to show that numbers aren't everything. Fortunately all 4 use the same connections, so swapping is a quick easy job. A permanent change will require bias adjustment though, with the high-slope types needing less bias. For example, plugging in a 6V6 where a EL33 should go could cause the 6V6 to overheat unless the bias is modified.

The only common valve to beat the KT61 figure is a 6BV7 (11000), while the humble 6M5 = 10000, but of course they don't fit in.

Hi Rob, and then I looked at the circuit for the 4A20.

What a difference!
The Radiotron guys must have got into the ears of the S/C designers because the 6AR7 is now configured as a full blow IF/AF amplifier using the screen as the AF "plate" load as you see in the Radiotron Designers Handbook.
They also changed to using back bias from the 6X5 to provide a bias pedestal for all the valves and used a seperate diode to generate an AGC to loop back to the 6A8. They even have a AF feedback loop from the speaker back to the 6AR7 to modify the AF response (bass boost?).
In other words it is the 4A19 all fixed up and proper.

Possibly they used the 4A19 as a price point seller, the "economy" model?
Then the 4A20 would be a higher sale price point with "5 valve " sound, for those that could afford the bit extra.

Fascinating looking back and trying to second guess what designers and sales were doing from our viewpoint today!

Fred.

Rob, re post #62 and valve type.

That all makes sense as manufactuers were heading toward the 3 valve line up using higher gain types like the 6BE6/6BA6/6BV7 to get the required gain with silicon diodes for the HT. Sets did not need an external antenna wire as ferrite rods were used for the tuning coil.
A real "Hot" 3 valve set of the era can drag in stations from all over Australia and match an older 5 valve octal set for Sensitivity easily.

All these sets around 1950 to 1960 show the change from octal to novel and nearly getting to using the new sub minature valves being developed.
Then side by side came transistors and by 1965 or so, the valve was DEAD!
Oh well.

Just like that.
The transistor was a heater less triode that could get hundreds of hours of use on a small battery.
Into the tranny/ chip era we went and the rest as they say was history.

But we like our history!
Fred.

If you search for stromberg carlson 4a19 in google, you'll see some photos. Mine looks like the brown one with brown knobs.

The 4A20 diagram (freely available this time) shows the way to improving the 4A19. Of course I can't do anything about the tapped volume control or the 2nd IF, but the reflex, AGC and back-bias all look interesting.

Re post #65.

Hi Rob, yes, if you keep your 6V6 as the audio, then reflexing the 6AR7 could make up about the 3 times audio gain to get to a "EL35" gain.
The IF gain would be down a bit as a negative.
I would still go for it as you could ditch the cathode control and wire in the volume control as the 6AR7 reflex ciruit.
Not sure about how much AGC voltage would be available to control the 6A8 but then again if the set is in a city or suburban area probably does not matter that much.
At least you would have a sane volume control.
On that point I have used a 3 leg pot and if I wanted to restore the bass boost feedback I used a pair of resistors to fudge the tapping.
The other thing is to forget the bass boost feedback than you need no tap.
Only way to find out is to rewire and workshop the changes.
Fred.

One of the things I have done for a 2m log pot was close couple it to a multi position switch. The original one is missing & it had the wrong pot; Along with a myriad of other issues. Info on the tapping is certainly not on the circuit diagram or easily found to date.

I stepped the switch, measuring the resistance at each step then added resistance to match that point; Thus giving me a close tapping point.

Marc

Hi Rob, I dug out my reference to reflex receivers, my 'Radio Designers Handbook' (Radiotron). 7th impression 1967, page 1140.
A lot of the book is way over my head, but is a gold mine of radio information coming straight from the horses mouth so to speak.

In chapter 28 Fritz Langford-Smith discusses the pros and cons of the design, the background and the nuts and bolts of the circuit.
But more than that there are circuit diagrams of real receivers shown as examples, one of them the Radiola 517-M and another the SC 4A20.
The 4A20 is obviously taken straight from Radiotron's design work and why not, Radiotron were the gurus and made the valves!

The 4A20 circuit is the one to use as a guide.
Modifying a 4A19 would be a challenge as there may be no room to squeeze in a 2nd IF.
The first IF they used may be a different very high "Q" unit to get best gain and selectivity.
I would leave the R/C network that replaces the 2nd IF as is and work with just the single IF coil.
The tapped volume control could be ignored at first, being the point for the AF feedback.
That feedback was done for technical reasons L-Smith goes into to complete the design efficiency.

You would not attempt the change unless you treated it as a technical exercise, and prepared to put it back the way it was if unsuccessful.
I would do it as I am pretty confident in getting myself into and out of trouble! (called over confidence!)

Very very interesting.

Fred

Down to the dregs, radios that have been hiding under boxes and beds, unseen for ages. Unfortunately almost all of them have problems. Once these are dealt with we're into the large consoles, not too many luckily, not a fan of them.

1. Philips, unknown model, large table radio with 3 SW bands. Most of the insides is just space. Valves: 6N8, 6AN7, 6N8, 6N8, 6M5, 6X5GT. As mentioned last week, no reception unless my hand was placed in a strategic area. After poking here and there, decided to measure the 2 series-connected aerial coils, to find one of them was open circuit. It was coated with black wax which appears to have eaten through most of the windings. The wire from the aerial connection into the coil was missing entirely. It's a special fitting which means there nothing much I can do about it. So, I did something unexpected and not recommended, which I won't describe. Suffice to say it now works on MW but not SW (not sure if SW ever worked), and there's one less valve in the radio. I'll let the smart cookies work it out.

2. Precedent World Range Six (don't know the actual model number). Wooden cabinet. Valves: 6V4, 6M5, 6BD7, 6BH5, 6AN7, 6BH5. After turn-on there was nothing. Turning up the volume produced a squeal then an ominous click and then utter silence. When the click happened, there was a purple spark on the underside of the 6V4 base. The voltages seem ok and all the caps (except the electros) were replaced when I first got the radio. The 6V4 gets hotter than the 6M5. It looks like the heat of the 6V4 and 6M5 has damaged the cheapo valve sockets, and I couldn't be bothered replacing them. Might just put it aside for now.

It does not appear to matter if is US or our black tar / bitumen or whatever. I get a strong feeling that it has an undesirable amount of sulphur in it , or is just plain corrosive.

Sometimes you can get away with blue murder if its not a resonant coil. Often it is the plate coil that fries and that can be caused by shorting a top trimmer on an IF. Where its an oscillator coil tickler, or a non resonant primary, a few turns of thin winding wire, will get things going. Do however, be aware that you can wire it "S" or "Z" and hook it up out of phase. You therefore do not call a fail until you have reversed the wires of the coil you just wound on.

I even admitted this sort of dodgy on the RF coil of an STC in Radio Waves. 6D6 had no B+ as both BC & SW primary coils were open. I fixed BC which was an external mouse chomp through 2 wires. I fixed the SW which would have taken a month of Sundays to get out, by winding a few turns of modern winding wire around a chunk of ferrite & dropping it into the centre of the tubular former.

Marcc

Fixed the Precedent. The capacitor (a modern-looking plastic one) that is wired across the speaker transformer was a dead short. After this was replaced, still no sound. The new cap (rated at 1000V) had also shorted, but not a dead short. The speaker transformer was now also open.

After looking at the transformer with a magnifying glass, I could see that one of the wires of the winding had come off the tag. Joined this back up, used another cap rated at 1200V. Working. Put the radio back together - no sound. Arghh! I had been well and truly caught out. The speaker plugs into one of those little 2-prong sockets of yesteryear that were once so common. But, there's 2 of these sockets - I'd used the one for the gram input. Put it in the right place - working - at last !!

The shortwave on this radio (despite the World Range moniker) only covers 4 to 12.5 MHz, so there's nothing to listen to during the day. I did find a couple of stations in the 25m band (11.5-12MHz), and it's quite sensitive. So finally this radio is a good 'un.

Getting it back together was another trial. On the first attempt, a dial light popped out, so tried again and again. On the third try the bulb shattered, so I said fine, no light for you. Lining up the screws on the bottom took a while too.


The next radio is a Murphy 674, made in Welwyn Garden City in England. The valves are mostly oddball Mazda types. EZ80, 6P15, 6LD12, 6F18, 6C12, 6L12, EM84. The 6L12 (in the FM unit) has been replaced by a 6BQ7A.

Not sure what is wrong with it, but the EM84 display is quite dull and indicates nearly full scale all the time. Tuning in a station barely makes a difference. On my 10-valve set the EM84 is bright and the shadow moves quite a bit. I've never encountered a dud EM84, but we'll see about this one next week.

The plate bypass on the output tube, has always proven to be the most likely to fail. The transients in the audio are like those of the buffer cap on a vibrator unit & 1600V caps are not unusual, often I will use 2kV or a line cap on a vibrator. Often the data sheet will specify a voltage and woe betide those who transgress, and go below it.

Dull EM84 cannot be assessed until the PSU is overhauled along with cap replacement. If there is instability and an oscillating stage that will affect it. It is not wise counsel to power a set before assessing it. I know as I have seen the damage and am expected to repair it.

Spent some time on the Murphy.

Equivalents:

6P15 = EL84
6LD12 = EABC80
6F18 = 6CE7 (EF89 with slightly different pinout)
6C12 = ECH81
6L12 = ECC85

The 6F18 has the same pinout as a 6EH7, so one of those went in for extra gain.

Replaced a sticky wax cap. The transformer is producing much less volts than the diagram says, so the HT is correspondingly lower, at 190 volts instead of 286. Still, the radio works perfectly well on MW. It makes a lot of buzzing on LW, but I think that's normal too, as my comms radio also makes the same kind of noises. Only on FM is there a problem - utter silence. The voltages are present on the ECC85 anodes, but that's about all I can check. I swapped the 6BQ7A that was there for a 6ES8, but still nothing. All the valve radios that I've come across have dreadful performance on FM, so even if it worked it wouldn't be worth it.

Now to the EM84. The display seemed to be ok after all, it's only the shadow angle that is the issue. So I experimented. There's usually a 470k resistor from B+ to the triode's anode. Making this resistor larger increases the range of control, magnifying the shadow change. A 22M resistor gave lots of movement. Another unlikely experiment is to try out other EM84's - there's considerable variation between specimens. And lastly, the EM87 is much more sensitive than the EM84 and can be plugged straight in and looks the same. To get the proper range with the original setup requires a greater variation of the AGC - this radio only varied from 0.5 to 2.5 volts, which is nowhere near enough. I decided in the end to leave it alone with the original resistor and valve.

So the main issue is the transformer voltage, and the non-working FM. I'm finished with it now.

EDIT: measurement of the filament voltage showed only 4 volts, so it really is a transformer problem. All voltages measured with the rectifier unplugged. The transformer gets warm after about an hour, but not excessive, no smoke or smells. There's no voltage taps, so I can only guess either it has developed some shorted turns that don't cause heat, or it was faulty from the start. If it wasn't for the FM problem you'd never know there was anything wrong.

It seemed like I would be repairing 2 radios today, but in the end only one had to be done.

1. Philips Radioplayer, unknown model number. I didn't take note of the valves but it seemed to be probably ECH35, 6U7G, 6B6G, 6V6G, 5Y3GT. It was to be repaired because of no sound, but when I put it on the bench it was working well. There was a bit of a smell at the start, but it went away and everything was good. I shrugged and put it back on the shelf.

2. Kriesler 11-7E. Kevin Chant doesn't have anything for the 11-7 which surprised me greatly, however google sent me to a thread from 2014 on this very forum. The attachment has details for every version which is wonderful. So the valves are ECH35, 6SK7GT, 6SQ7GT, EL33A, 6X5GT. The ECH35 had been replaced by a ECH33 at some time. The problem, again, was no sound. Touching the Gram jumper at the back produced a healthy hum, so that ruled out half the radio in one go.

The backbias circuit only produced 1.8 volts, so it was studied and found to only be for the AGC - the output valve has its own arrangements. The grid of the 6SK7 was glowing, so it seemed the AGC voltage wasn't working. This is provided by R146A and R146B, both 1 meg. R146A was found to be open circuit, while R146B wasn't found - it was missing. 2 new resistors fixed that, but still no sound. After pondering for a bit, and thinking of a previous repair, I measured the aerial circuit which turned out to be open circuit. Surely not another ruined aerial coil. 'Fraid so - the SW coil was open. Since SW is of no use these days I soldered a capacitor across it, and now MW is working perfectly well. Put the radio back in its case, repair complete.

Again I have come to the end of the queue for now.

I tested all the consoles apart from 2 which I know are not working. One of those never worked, it was waiting for a dynamic speaker which never arrived, so I'll just get rid of it. The other of those is an octal Radiola which used to work but just faded away. I might dig it out from under a pile of stuff and take a look. Of the tested consoles, the Peter Pan clone works, the unnamed radio with 2.5 volt valves now only works at the bottom end of the dial, it might have a blown aerial coil, and my ancient regenerative radio (24A, 47, 80) is totally dead. It used to work though, but seems the regen stage (around the 24A) died. Maybe I'll try fixing these, I don't know yet.

After all that, there's a console in the laundry with 2.5 volt valves, but it's missing a 2A5. It used to work (after borrowing the 2A5 from the other 2.5v radio), but with a few squeals. It's been a bit exposed to the weather, so it has quite possibly died since then.

The death of an electrodynamic speaker, is not really a good excuse to get rid of a redeemable set. 90% of the stuff that comes here is commercial repair. A rare fail, more often a chomped cone, is a minor inconvenience that can be dealt with.

Some hefty WW resistors, or a panel mount one and a Permanent Magnet type speaker sort that out. When a tube (not rectifier) is looking a bit embarrassed & red in the plates; the first place to look is the control grid. Some times its the coupling cap, & others an open circuit control grid. The latter as its becomes a diode.

Often its the tickler coil, or a non resonant plate coil that's gone open. In such circumstances I will run a hand wound coil onto the bobbin over it & it works; especially desirable if its feeding the plate of a TRF stage prior to the pentagrid DC.

Marc