RGD 1046 amplifer repair

[mindburner]

Hi yes the anode to chassis

voltage on R47 is 18v (that's after another 10k in parallel)

yes 1 more fitted

the anode voltage has increased to 97 (again with extra 10k) it slightly falls over time

with 4 10'ks

its 109v anode

and 21v for r47

[Radio Wrangler]

Once you get to the point of suspecting disconnected welds inside the valve, then you could use a good capacitance meter to compare inter-electrode capacitances on the valves.

There are capacitance ranges on many multimeters these days, but they're really set up to measure larger capacitances than these. The Peak LCR meter is good, though.

David

[PJL]

With the increased current you MUST move the resistors onto the centre tap of the transformer so we are not putting all the DC current through one side of the transformer.

[PJL]

Move the resistors as above and leave it on for 10 minutes. Is there any change in the anode voltage reading? I am looking for it to increase over time.

We are still only putting 6W or so into the valve but it should help raise the temperature a bit to activate the gettering.

Is there a blue glow from the valve? It would be inside the anode cavity.

[mindburner]

Originally Posted by Radio Wrangler

Once you get to the point of suspecting disconnected welds inside the valve, then you could use a good capacitance meter to compare inter-electrode capacitances on the valves.

Hi

I have an LCR so will get to that, anode to grid?

[mindburner]

Originally Posted by PJL

Move the resistors as above and leave it on for 10 minutes. Is there any change in the anode voltage reading? I am looking for it to increase over time.

hi righty o will do

just to double check is that the red centre tap between the 2 anodes in the picture

[mindburner]

I think I may have found a physical problem. A break in grid? this is the strong valve

[Radio Wrangler]

Do all three capacitances anode to grid, grid to filament, and anode to filament. And then compare the two bottles.

Check the capacitance reading with the probes held but not connected to be able to subtract the strays, then someone else can measure another PX4 for comparison... but I think you're looking for a fairly large fault.

David

[mindburner]

on the strong valve.

Anode to Grid - 2.2pf
Anode to L filament - 7.8pf
Anode to R filament - 8.4pf
Cathode to L&R fil 1.5pf

on the weak.

Anode to Grid - 13pf
Anode to L filament - 8pf
Anode to R filament - 8.4pf
Cathode to L&R fil 10.8pf 18pf

[Radio Wrangler]

There is a difference, but it's the opposite way round to what I was expecting.

Anode to filament seems similar enough between the pair.

Anode to grid on the 'strong' valve makes it look like the grid might not be connected, and that fits with post 67. Without a connected supply of negativity to the grid, the grid will float and the current will increase to whatever space-charge and the filament emissivity limit it to.

So it sounds like you have a problem with the strong valve hogging all the current, and the weak one gets biassed off by it via the cathode resistor. there might still be a separate problem in the 'weak' valve but you won't see it with these two run as a pair.

I'm assuming the "Cathode to L&R fil' is a mistake and you meant grid to filament pins. These fit with the grid to anode capacitance difference, so reinforce the open grid suspicion.

The filament runs at a certain temperature, and that throws off electrons at a rate dependant on the temperature and the Work Function of tungsten, thorium or whatever the filament is coated in. So there is a limited current flow available from each unit of area of the filament surface.

The electroons would just hang around the neighbourhood and do nothing without the positive attraction of the anode to pull them away. The electron population around the cathode acts to repel other electrons and discourage them from ever leaving the cathode. So far we have a diode valve which shows a current saturation limit imposed by emissivity, surface area and space-charge. If we surround the cathode by a grid, we artificially increase the space-charge effect and reduce the emission of electrons. to cut the natural diode current right down, we need to make sure the grid creates a uniform field over the whole hot area of the filament, otherwise electrons will get through.

If an area of grid became disconnected (Very unlikely, it's supported by welds and bits would drop off if they separated) then parts of the filament wouldn't be controlled, and there would be saturated current density in that area. If a connection tape from the pinch to the grid support wires comes loose, that leaves the grid physically supported, but electrically floating, and the valve's current is full-on.

With directly heated wire filaments as the cathode, it's difficult to get a big area so the current capability isn't very high by today's standards, so to get power, you need high-ish anode voltages. The introduction of indirectly heated cathodes suddenly increased the emissive area a lot, and got around the awkward relationship of wire diameter to area which forced low voltage high current heaters. the development of oxide coatings put up the emissivity a lot as well. Cathode to heater insulation suddenly gave the circuit designer freedom to make biasing a lot more individual.

In a way, these valves are museum pieces, though rather interesting ones. Unfortunately someone, somewhere, somewhen said that directly-heated triodes sounded better than anything else, and a whole cult snowballed. It begins to look like you need to find at least one good PX4 to get that amplifier working, and the cult status has put the prices higher than anyone not a cultist could justify.

I'm just glad it didn't cost you anything.

David

[Radio Wrangler]

Figures for Osram balloon era PX4:

anode to grid 13.3pF

Anode to filament 5.8pf

grid to filament 9.3pF

from data sheet at r-type.org

David

[mindburner]

Originally Posted by Radio Wrangler

I'm assuming the "Cathode to L&R fil' is a mistake and you meant grid to filament pins. These fit with the grid to anode capacitance difference, so reinforce the open grid suspicion.

yes sorry should have been Grid to left/right filament.

Having a closer look at the strong valve, it does seem to have a physical disconnection from the grid pin. There is a distinct gap compared to the other PX4. So essentially the grid is just doing nothing.

I supposed this issue could have potentially damaged the weak valve. You explanation of operation is very helpful and fascinating. I love how these valve bottles are full of electrons being coaxed to amplify signal. How designers used better coatings and ultimately higher voltages to increase gain.

It would seem that I have been both simultaneously lucky and unlucky. With a nice 1046 amp and 2 possibly, one certainly defective valve. Well that's not too bad I suppose.

I would be concerned in this amp that voltages seem to be high at 419v coming off the UU8 rectifier. Is this possibly too high as I believe the PX4's operate at 300V

I have also decided in a fit of madness to buy a 'matched' pair of PX4's, tested working from eBay. The seller has good feedback and hopefully will deliver as advertised, special delivery next day etc, with fantastic packing. Concerning if a small weld pops off. This is a bit of a gamble as I am now going down the route of this endeavor costing me hard earned cash. £400 However, I have gone so far and I really would love to see this amp working as intended. I was saving up to buy a new camera lens but hey...This is where I need to be extra careful not to do anything to damage these new valves, especially as I will not have other kit to test their operation before trying them on the 1046.

This is one of the reasons I mentioned the rectifier voltage. It will drop with load but anode voltage is still too high?

I must say this lamp limiter is a fantastic idea. I had used one way back but with this amp it has been very useful so far. A must if trying any new valves.

[PJL]

So the strong one is definitely unserviceable.
Unfortunately, that is also true of the weak valve which does not conduct at all.
Odd to have two faulty valves like this and you have to wonder if someone swapped them out at some time in the radio grams history.
You would do best to look at the thread related to alternative valves.

[Radio Wrangler]

It's unfortunate.

It's a lot of money to tie up in a mono amplifier and going stereo would be very expensive, still, a woking amp with the original valve types is a currently fashionable and valuable thing.

In the reverse twist, I'm going to have to stump up about that much money to have a Nikkor repaired. My camera bag rolled out of the car when I opened he tailgate. and the lens on the camera bent the lands of its bayonet. The lens cost the equivalent of a pair of fashionable amps, and the repair will cost the equivalent of a pair of fashionable bottles. So other hobbies can be comparably costly, which puts this in perspective. I just didn't get the damaged lens given...

david

[mindburner]

Originally Posted by PJL

Odd to have two faulty valves like this and you have to wonder if someone swapped them out at some time in the radio grams history.

Hi yes it's a bit bizarre. I was reading that these valve were pretty reliable so it does seem that this odd pair have been placed in the amp at different times.

Originally Posted by Radio Wrangler

It's unfortunate.

yes certainly is. I don't actually mind too much as I sort of collect kit. I may never actually use it that much but like to know it's there. If the new valves work as expected, then it will be a reasonable investment. I shall spend some time going over the components and double checking.

That's a disaster about the Nikkor. Hopefully the repair will bring to what it was. Was it one of the larger lenses. I hear the 300mm f2.8 prime is very nice. I like the longer focal lengths as they are basically small telescopes that can be used well in astronomical imaging, if the CA is controlled and lens stopped down a bit. A heavy lens hitting the deck will certainly put some stress on the mounting.

I recently put about 15 amps through a specialist CCD camera, through messing with a poor power supply connection at 3am in the dark. Not cleaver, luckily repair was not as expensive as I expected.

Thank's very much for your patient and detailed help guys. I really would be still head scratching without it.

I shall get the new vales do a heater 4v check on them and perhaps capacitance check. try them on the amp with bulb limiter and see how it goes...

[Techman]

I had a cathode connection open circuit like that on a 6KD6 RF output valve. I'd bought it on line from a seller who was selling his fathers old valve stock, so knew nothing about valves and couldn't test them. The valve semed to be brand new and still boxed but it looked like there had never been a connection to the cathode when viewed through the glass - so near, yet so far, if only I could just open the valve and make the connection! The chap was very good about it and offered a refund or another valve of the same type, but a definitely used one. I took the replacement used one as I was just pleased to get a working valve at the time.

That's good to get a pair of valves for £400. I haven't been following the prices for these for a long time, so it's good to see that the prices may at last be coming down. This gives us hope that we may be able to keep some of these old radiograms running and they may not be quite so atractive for the amps to be pulled out after all.

I was playing with valves yesterday. I was thinking about the heater resistance you'd measured at 2.7 ohms and although it was quite some time ago that I checked one, I seemed to think that was a bit high, even though your filament/s looked to be glowing normally. I had a PX25 to hand (the bigger version of the PX4). I measured the cold resistance of the heater and it was 0.5 ohms. On a 4 volt supply the heater current was 1.9 amps, It should really be 2 amps, but heater was glowing normally, so probably just down to the long thin test leads from the power supply that I was 'bodging' with. Having seen this, I decided to fetch a PX4 type and check the cold resistance of the heater of that and it measured around 1 ohm. The spec for the PX4 heater is 4 volts at 1 amp. On the power supply at 4 volts, the current was 1 amp - the connecting leads were giving a truer current reading with the lower current heater of this valve than the higher current PX25. Obviously, ohms law tells us that the hot resistance of the heater is quite a bit more than when the heater is measured cold. I found that if I measured the heater resistance quickly after disconnecting from the power supply I got a brief higher reading which quickly dropped down to the normal cold reading. Did you make sure your valves were cold when taking the resistance reading, AND also, what are the readings now after your pin re-soldering? This may not matter much now, but if the readings are now down to 1 ohm (ish), it would be a good indication of the success you've made of the other electrode pin connections. It's also a good comparison to check your new valves when you get them. The valves I did the checks on were the Mazda balloon versions, but I checked in my 1954 data book and the specs are the same as the Osram version.

I don't think I have the means to be able to check those low capacitance values, which is a shame as it would have been useful.

Quick edit to say - When using a DVM to measure resistance, make sure it reads 'zero' when the probes are connected together, as a lot don't. If there's already a reading when the probes are shorted together, then you'll have to subtract this from your actual reading, or you'll get a false, slightly higher reading.

[PJL]

A look at your pics and the 'strong' valve does seem to have overheated at some time as the plate is darker and the pins have blued. Whether the weld failed first or as a consequence of overheating we will never know.

The 'weak' valve looks OK. Don't throw it away as it is just possible that it can be made to work again, possibly not well enough for your amp, but maybe for a radio.

[mindburner]

Originally Posted by Techman

Did you make sure your valves were cold when taking the resistance reading, AND also, what are the readings now after your pin re-soldering?

hi I just remeasured the resoldered valves filament to filament resistance. The valves are cold and I was using my new Fluke meter but I now get

0.9Ω on both valves

much more like it.

tried with my old meter and it's 0.9-1, jumps around a bit, meter is not really great quality,Maplins special)

On another note, 2 'new' PX4's in Osram boxes arrived this morning.

I shall check valves capacitance and resistance and post back

[mindburner]

Originally Posted by PJL

The 'weak' valve looks OK. Don't throw it away as it is just possible that it can be made to work again, possibly not well enough for your amp, but maybe for a radio.

hi

That would be most interesting to get the 'weak' valve to work at some level.

[mindburner]

Well today I got the eBay valves. Packed to perfection in a sturdy box with lots of shredded paper for protection. They came in Osram boxes, although not original.

The valves look in ok condition. There is more silvering on the back of one of the valves than the other.

Tried both on a 4v bench supply and heater filamets showed nicely on both valves

resistance
Filament resistance was 0.9Ω on one valve and 1.0Ω on the other

capacitance
V1

anode to filament 10pf / 10.4pf
anode to grid 14.9
grid to filament 3.5 / 12.2

V2

anode to filament 9.4pf / 9.5pf
anode to grid 15.7
grid to filament 11.2 / 11.6

seller stated
43ma@-45v / 43ma@-45v

the only one that seem off is V1 grid to filament at 3.5 but this may be ok

I have rechecked the amp again as best I can so I shall now proceed to power up with lamp limiter and measure anode voltage, fingers crossed and prayers to the valve gods