The Beehive

[BusyBee]

This week I have been doing a little testing and have also made an amplifier module which includes some band pass filtering (see photo). Basic tests on circuitry indicate that the modules are working so my next step will be to put all this together. I have allocated an enclosure (similar to that of the attenuator mentioned earlier). Before I rush into this though I am going to re-examine noise sources to determine whether I have the tidiest solution (noting that the circuitry at present will need two different voltage supplies).

One other thing I have thought about this week is my flight simulator access. My full system takes 10 or 20 minutes to setup but I often need to just check or test things.It would be useful to have an easy setup for this purpose. With that in mind I am thinking of building a new interface. This would be much simpler than previous with hopefully just one USB plug in. No rush and not before Christmas. Just thinking it out at the moment.

   

[BusyBee]

After a few checks I realised I needed to get the main amplifier of my noise generator together so that I could test things properly. I appeared to be getting uncertain operation. Today I put things together, apart from the noise generator board which I am still uncertain about. Things came together fairly well but I still got uncertain operation. The amplifier circuit appeared to have some form of parasitic oscillation. Although I have not used those particular components before I could not really see why I was having such problems. To cut a long story short, along with a lot of fiddling, the issue I finally found was a loose connection of the 0V socket on my power supply, resulting in  a poor / intermittent ground. With the original noise generator circuit I finally got an acceptable output. It was in the region of lower tens of millivolts (the thing about noise is that specific amplitudes are indeterminate and, at this stage, I'm not getting into detail).

The next step, apart from sorting my PSU, is to finalise on the noise generator circuit. I think it better to use only one supply voltage so maybe the 24 ish volt noise diode will have to go.

Otherwise, in the past week, I have been looking into Spice models for valves. I came across this thread / site in which lots of information is discussed: https://www.diyaudio.com/community/threa...ls.243950/ . I have found models which seem to work as they should and others which may be iffy. There are other sites but most models that I am aware of are mentioned in this thread. Implementation is helping my learning of LTspice which cannot be a bad thing.

[BusyBee]

This week I have jumped a little. The noise generator is aggravating a little as the circuit seems unstable and I am a little unsure about the noise source, as I mentioned. I have been doing a little on valve simulation but decided to look into testing an audio amp I have built.

Sometime around the middle of the year I resurrected an old project that I thought about from time to time. Back in the teens I bought a push pull audio output valve transformer with the intention of using it in a radio with gram facility. I finally got round to doing some work on this. I will probably go further into this in the future, after I have done some further tests and finalised things somewhat. The reason I mention it now is that I have put together the output and driver stages but have not, until now, got round to the preamp / input. As this is intended for a radio it did / does depend on the detector somewhat as I was thinking of using one of those double diode triodes as detector and first audio, as I have some, but that is where recent things about filter testing come in as it really needs thought about the IF stages too. As an interim, to enable better testing of the audio stages I have decided to build an audio input stage which may or may not be final in the circuit. Up to yet I have tested the power stages, setting up the output and testing with single frequencies but I think I would be happier to know things sound good with more complex signals. All has been good up to yet though.

The amplifier is relatively ordinary, using two 6p30b pentodes and one 6n16b double triode Russian valves. It's not a new approach as there are circuits online although I am using an ultra linear arrangement (perhaps a one off as I see the transformers have tripled in price since I bought mine).

To get to the present though. I decided to use either a 1j24b or a 1j29b Russian rod valve. I have also had ideas on a different construction for the tubes. I am not sure whether I have gone into any detail in the past about my assembly / layout so I thought I would take a few pictures while I was working as they may give a few ideas. I will describe the pictures but they should be fairly self explanatory.

Pictures
--------

1: The first part of things is to construct the screening tube. I use sheet tinplate I have bought but I imagine any tinplate would be Ok. With both of the valves I am thinking of a square 3cm x 3cm is suitable, which I cut with shears. I cut a line of these yesterday, as shown.

2: To form the tube I utilise an old valve which had gone to air. I made a cardboard tube that slips over this so that when I bend the tinplate around with pliers I can crimp it without problem. Care is needed though. It may be more of an issue if this technique were used with anything other than these tough little valves. A close fit is reasonably easy to get. Once the tube is closely crimped then I run a soldering iron down the seam, soldering it (while the valve and cardboard tube are still inside). I usually spot solder each end first as I am holding things in my hand and it needs quick work else it burns The tinplate solders easily though (please note it may be not as straightforward with other metals). With practice a little pile of tubes can be quickly obtained.

3: Referring to my original method of circuit construction I used tie wraps and their sticky back clamps to hold the valve tube in place. I used a pad of double sided sticky foam between the tube and mount. What cannot be seen are the strips of rubber band in the tube which rest between the inner tube wall and the glass of the valve sided. These are intended for cushioning and damping. The board shown has the basic components which I generally use for supplies / bias etc. It has not been purposed. 1j24b as written on the mount. This method has worked well although I have varied things slightly from time to time - Predominantly for tests at higher frequencies.

4: To explain - One of the main points about the new arrangement is that the shielding tube will now play a major part in the mechanical supporting structure. My intention is to use two smaller circuit boards. One at either end of the valve. One for input and one for output. Another point is that the multiple grounded leads which these valves have can be better accommodated. Although it's not the intention here, and some minor changes would be needed, the arrangement has scope for high RF frequencies too.

The valve itself now has a heat shrink wrap, replacing the rubber bands. As they do not get hot it shouldn't be an issue but if I find problems a couple of small bands would replace that. The photographs show the 'end plate' boards. Ground wires are soldered directly to this at the cathode end and the non grounded leads are sleeved in coded colours (I have a limited colours but do work to a plan). The screening tubes have thick copper wire supports soldered to the boards (three each end). It's slightly untidy but it's the first time I have made them. I do think it may be better to have a longer screening tube so that each end could be soldered directly to (maybe) double sided copper board. Further thinking but it's one thought for RF work, although at present I am not going there. The two assemblies shown are one of 1j24b and one 1j29b. I have not got a naming convention at present but the lead configuration of each is different.


Today I did some basic tests to verify the valves are operating as they should. The next step is to construct amplifier modules. One disadvantage of this construction as the amplifier and tube assembly are more separated but I cannot have it both ways

               

Tracy

[Ed_Dinning]

Hi Tracy, see your friendly local plumber for offcuts of copper tube. Available down to 6mm and with various solderable fittings available to support the screening pipe

Ed

[BusyBee]

Thanks Ed. Yes, that could result in a tidier soloution with more precision which may well be useful, particularly if I do anything at higher RF frequencies.

I bought the tinplate for a broad range of screening useage as it is easy to solder.

Tracy

[Mike Watterson]

Tinned steel is what the Russians used. I wind it on a screwdriver. The Rod pentodes are very sensitive to magnetic fields. Easily cut with scissors. Mostly I've used cut up coffee tins.

[Ed_Dinning]

Hi Tracy, the very best stuff would be mu metal sheet; solderable and easily cut with scissors. Tight corners as in bending reduce magnetic screening properties.
Also terribly expensive!

Some small pieces here if you need to try it

Ed

[DrStrangelove]

IIRC mu metal's magnetic properties after mechanical working can be "repaired" by tempering in a hydrogen furnace.

2010 to 2160 deg F apparently.

I wonder how I knew that. Lost in the mists of time.

[Mike Watterson]

I think only critical for precision analogue 'scopes?

[BusyBee]

I'll carry on with the tinplate I have for now as I have enough for a while and it fits the purpose.. I have also fabricated enough tubes for the near future too.

I do have some mu metal, I think, which I picked up from somewhere. It is a shroud which came from an oscilloscope tube. I have not tested it but the metal is a bit like tinplate but with less shine, looking more like stainless steel. Similar to what Mike mentions.

Tracy