Kriesler Beehive radio

[WaveyDipole]

I recently obtained this from a member and have started working on it. The radio does not appear to have been touched in a long time. On opening the case, there were two Electrolytics under the chassis. One had become detached from a large solder joint connecting two other components and a wire. The second one had a broken wire at the chassis end.

I replaced the electrolytics and "that capacitor" and gently brought the radio up on an isolation transformer and the variac. This is the first opportunity I have had to do so after recently restoring the variac in question. It has already proved useful for two other tasks.

The dial lamp lit but there was no sound whatsoever. I have bit by bit now replaced the remaining wax paper capacitors including the two Metalmec types. I raised a question earlier in the Components section of the forum regarding the capacitors within the tuning section of the circuit here:

https://www.vintage-radio.net/forum/...d.php?t=167862

Since I now have other questions I decided to start a more general thread in the radio section and despite my rather embarrassing oversight, I am happy for the mods to move that component information to the more general thread here.

The wiring to the speaker and electro-magnet had become brittle and the insulation was breaking up, so these were rewired and the cable slightly re-routed to keep it away from the heat of the mixer valve. A short piece of wire from the variable capacitor to the grid of the mixer valve had suffered the same problem so was also replaced.

I have so far been rewarded with an operational broadcast band, but the short-wave band remains quiet. The problem seems to be no LO output when switched to the short-wave band. I have cleaned and checked the band switch contacts and they all seem fine.

Of note is the fact that the mixer valve, which is specified as a 6J8G on the circuit diagram, has been replaced with a Philips ECH35 at some point. According to the National Valve Museum, the ECH35 is designed "for broadcast applications" while the 6J8G is "designed for stable operation on the broadcast bands including Short Wave". Could this difference be the reason why the LO is not oscillating on SW?

http://www.r-type.org/exhib/aam0014.htm
http://www.r-type.org/exhib/aag0141.htm

One other issue I have at the moment is how to clean the dial glass. There are two layers of glass sandwiched together with approximately 3mm gap between them for the dial pointer to travel (see photos - note the camera lens has made the second one look a bit distorted). I have cleaned the front and back surfaces, but there still seems to be a fair bit of grim in between the two pieces of glass and I have not been able to remove the glass from the metal frame. The separators are rigid and the metal looks slightly crimped at each end so that the glass will not budge. There is some possible evidence of either adhesive having being used or the padding melting slightly at some point. I suspect the former. How do I get in between the two pieces of glass to clean them? A Q tip doesn't fit and I have gently tried an air duster without success. Also how much risk is there of the print coming off?

[ms660]

The ECH35 is good to 31MHz at least, not saying it should work in your receiver though.

Lawrence

[Herald1360]

May be suck eggs but have you checked continuity of the SW oscillator coils and switching?


The valves are pin compatible so I wouldn't expect "the ECH35 is designed "for broadcast applications" while the 6J8G is "designed for stable operation on the broadcast bands including Short Wave"." to stop the LO per se. ECH35 should certainly work at SW broadcast band frequencies.

[ms660]

Pin 1 of the ECH35 is the connection to the valves metalization, I had an Eddystone 358X receiver that used
the ECH35 as original, 30Mhz no problem, the mixer section is a hexode so not ideal for sliding screen circuit.

Lawrence.

[WaveyDipole]

The radio is a model 11-4 and the SW band ranges from 16 to 50 metres, or 6 - 18 MHz so theoretically at least, the frequency range seems to be more than adequately covered. I have attached the diagram of the tuning/LO stage in case it helps the discussion. Since the valve ought to be fine for this application (thank you for the confirmation), I am now going to check a few more things such as shorted trimmers or tuning capacitor plates and the right hand (i.e thos connected via C19A to anode 2) oscillator coils.

[ms660]

According to some service information I've just found the 6J8G was replaced by an ECH35 in some versions of the 11-4:

https://www.kevinchant.com/uploads/7...sions-ver2.pdf

Lawrence.

[WaveyDipole]

Originally Posted by ms660

According to some service information I've just found the 6J8G was replaced by an ECH35 in some versions of the 11-4:

https://www.kevinchant.com/uploads/7...sions-ver2.pdf

Lawrence.

I checked the remaining coils which seemed fine. I checked the trimmers and variable caps for shorts, broken links, etc but could find no other problems. This exercise has enabled me to map out the entire tuning section so was worthwhile. The oscillator, when it was running, ran from almost exactly 1MHz to 2MHz which isn't quite wide enough to cover the band from 540 to 1650kHz assuming a 455 or 450kHz IF product added on top. Would alignment have provided well over 100kHz of correction? In any case, the LO has now stopped working on the broadcast band as well so that taken along with not being able to find any other fault leads me to suspect a faulty ECH35 valve. Since I don't have another to substitute or any means to test the valve, I can't be sure. Would a weak valve suffer an impeded frequency range and would it suddenly stop oscillating?

There seem to be a few ECH35's on eBay at the moment so I'm going to have to buy a spare to test I think.

[Herald1360]

Have you got a 6K8 to try?

[Nickthedentist]

Re. the dial glasses: I'd leave well alone. Trying to separate the two could end in tears, as could trying to clean the printed side of the glass.

If you must, I'd wrap an almost dry piece of soft material around a lolly stick and carefully do what you can, but there is a real risk that it could go horribly wrong.

I think it looks fine as it is, personally.

[WaveyDipole]

Originally Posted by Herald1360

Have you got a 6K8 to try?

Sadly no, but thanks for the suggestion. I don't have any octal valves to hand. I have posted in the wanted section, but will resort to eBay if needed, its just that is always a bit of a gamble and I would rather the money go to a member.

Originally Posted by Nickthedentist

Re. the dial glasses: I'd leave well alone. Trying to separate the two could end in tears, as could trying to clean the printed side of the glass.

If you must, I'd wrap an almost dry piece of soft material around a lolly stick and carefully do what you can, but there is a real risk that it could go horribly wrong.

I think it looks fine as it is, personally.

Nick, point well taken. I have no intention of trying to apply any more force than I have already. I can't remove the assembly from the mount never mind trying to separate the two plates. I tried gently levering to a point, but have reached a point, I think, there any further force is likely to cause the delicate dial glass to splinter, or even worse, to crack and break. I have heard that old decal can sometimes come off even with a dry wipe or a blast of air, never mind when using water. As you say, it doesn't look too bad in the photos. The downward angle on the second photo makes it look much better than it actually is. In reality, it does have a rather dirty appearance, but at least the decal is intact and all clearly is visible if looking a bit dimly lit. I think most of the dust is probably on the frosted surface and I did take a risk with the air duster which was not successful although no harm done. I figure poking something down there could be risky so was hoping there might be some other clever way. Maybe improving the back lighting will help - there is only one solitary bulb.

Anyway, in the mean time I have had an idea about testing the operation of the valve which I will try tomorrow. If I can determine the IF frequency (as it does not seem to be stated) and then feed a modulated signal into g1H and a sine wave at the sum of the modulated signal and IF frequency into gT to represent the LO, then I should presumably get the IF product in the same way as accomplished when tuning the radio at the anode/1st IF and be able to decode the audio signal at the output. If the test is performed at various frequencies between 550kHz and 18.7MHz, this would determine the response of the valve across the frequency range of the radio. Of course, that would only test whether the valve is capable of generating a mixed IF product at those frequencies, but presumably if it can do one it should be able to do the other at the same frequency? I could do with a better understanding of how the oscillator actually works and will do some research on that aspect.

[PJL]

Measuring the DC operating condition of the ECH35 triode and checking against the datasheet will give you a rough idea of it's condition.

I guess you have already replaced any decoupling capacitors? A common problem is losses due to leakage in the waveband switch so you could try hard wiring around the switch.

[WaveyDipole]

Originally Posted by PJL

Measuring the DC operating condition of the ECH35 triode and checking against the datasheet will give you a rough idea of it's condition.

I guess you have already replaced any decoupling capacitors? A common problem is losses due to leakage in the waveband switch so you could try hard wiring around the switch.

Thanks. I haven't replaced any mica caps. but the wax paper ones have been replaced.

Voltages around the valve:

Heater		5.3Vac
aH		205Vdc
g2,4		77Vdc
g1H		0V
gT,3H		0V, (no signal)
aT		120V

Voltages from the data sheet for a Mullard ECH35 are:

Heater		6.3
Anode (mix)	250
Anode (osc)	100
Screen		100
Grid		-2.0

For whatever reason the heater supply is 5V in this radio, but on that basis does g2,4 seem high?

My test did work and I was able to determine that the IF operates at approximately 455kHz and I did get output from the mixer up to around 19MHz. The sig gen I was using to synthesize the LO only goes up to 20MHz, but certainly enough to prove that the valve works across the frequency range.

[ms660]

5.3 volts heater voltage is too low.

Lawrence.

[WaveyDipole]

I thought I had posted the schematic earlier. Here it is.

It seems that only the rectifier heater and speaker magnet are fed from the 5V heater supply. I guess the remaining ones should be connected to the 6.3V supply. I will check that out. I should mention this was with the variac set to 220Vac on the mains input, and all readings done on the AVO, but it still seemed proportionally rather low. I set the variac to 240V and measure again and this took it to just below 6Vac, so better but still perhaps on the low side.

BTW, Lawrence thank you for that link back in #6. Nice to have some layout information and dial cord arrangement drawing. Also the text is of a much more readable size compared to what I had, so much appreciated. I see the ECH35 was introduced in the 11-4U version onwards.

[WaveyDipole]

Just an update: it seems the LO is very sensitive to mains voltage variation, particularly on the SW band. At a full 240V it sprung to life on SW at least for some 80% of the band, but there are still significant gaps at either end of the tuning range. Turn it back down to 220V and it stops oscillating again. A good sign is that with the LO running I can pick up the odd station on SW with my random piece of wire attached to the antenna.

I am going to check the resistors with the DMM. The AVO did indicate some of them may be high, but it was difficult to get a precise reading.

[Graham G3ZVT]

Originally Posted by WaveyDipole

It seems that only the rectifier heater and speaker magnet are fed from the 5V heater supply. I guess the remaining ones should be connected to the 6.3V supply. I will check that out.

Just a slight correction, the speaker field coil is in series with the HT from the directly heated cathode of the rectifier. The voltage energising the heater has no real relevance to this.

Certainly check that the other valve heaters are connected to the correct 6.3v supply,

[WaveyDipole]

Originally Posted by rambo1152

Just a slight correction, the speaker field coil is in series with the HT from the directly heated cathode of the rectifier. The voltage energising the heater has no real relevance to this.

Indeed, my error. The speaker field coil is indeed in series with the HT.

Certainly check that the other valve heaters are connected to the correct 6.3v supply,

Checked this to confirm. The heaters of the remaining valves are connected to a separate tap on the transformer.

One resistor, R136 was out of spec. I temporarily placed a 100K in series which brought it into spec, but it made no difference to the LO on SW. I have another ECH35 on the way now, so we will see if that makes a difference.

I notice that one of the revisions was to remove the tone control pot and replace with a two position switch. Would this have been a cost saving measure, i.e a resistor being cheaper than a pot?

[PJL]

Without knowing the value of R101A, the circuit has very odd component designations, it is impossible to check the state of the triode. The ECH35 triode should be running at around 12mA anode current with 120V on the anode and 0V on the grid so I suspect it is a bit weak whilst the 6J8G runs with a slightly lower anode current.

The other thing to watch for is ensuring the earth wiper on the variable capacitor is making good contact with the rotating shaft.

[ms660]

So far as I can make out R103A or R105A? (Osc anode load) and R103B or R105B? (1st AF amp anode load) are 15k 1 Watt in the service info.

Lawrence.

[WaveyDipole]

Originally Posted by PJL

Without knowing the value of R101A, the circuit has very odd component designations, it is impossible to check the state of the triode. The ECH35 triode should be running at around 12mA anode current with 120V on the anode and 0V on the grid so I suspect it is a bit weak whilst the 6J8G runs with a slightly lower anode current.

The other thing to watch for is ensuring the earth wiper on the variable capacitor is making good contact with the rotating shaft.

Thanks. I will check the anode total current. Interestingly, with the anode at 120V (which is what was measured) a 15k resistor would limit the anode current to about 8mA. With a 40k resistor (see note below) this drops to about 3mA? Both are below the 12mA. A 10k resistor would have been spot on.

Originally Posted by ms660

So far as I can make out R103A or R105A? (Osc anode load) and R103B or R105B? (1st AF amp anode load) are 15k 1 Watt in the service info.

Lawrence.

Yes, it is difficult to make out, but as per circuit diagram and original BOM I agree that R103A/B are stated as 15k 1W. What is actually fitted measured 40.8k. Looking at the list of modifications, starting at model 11-4U, R103A was replaced with a 40k. I haven't looked at R103B yet but will check that out as well.

Also, R136 was changed from 30k to 40k but actually measured at 50.1k. The 100k that I temporarily placed in parallel reduced this to about 34k, making it much closer to the original value, however that made no difference.

I am ignoring the new designations for these parts (R153A/B) as this would confuse discussion when referring to the circuit diagram. However, given PJL's comment, I am curious as to the manufacturers intentions when making this modification?