Haigh (Matthews) Wobbulator

[Bazz4CQJ]

Hi guys, here is my preliminary offering by way of a BOM. I've shown it as a PDF which I've created in Excel. If anyone wants to chip in with spotting any errors or anything else, I'm happy to act as gate-keeper for the BOM.

EDIT; 4 mins later and I'm seeing the need to list the 400V caps as "notables"!
V2 shows it.

B

[Electronpusher0]

Originally Posted by Bazz4CQJ

Yes, having bought the 10-turn pot for the main tuning, I'm now very aware that the fine tuning pot will really need to be a good quality to complement it. That is leading me to check out my junk box for some nice old vintage pot; I suspect that much of the modern stuff may not be adequate. I found one old pot which feels silky smooth, but somebody cut the spindle pretty short, so it will need a knob that can cope with that .

B

Do you actually need a fine tune pot if you are using a multiturn pot for the main tuning?
A fine tune pot is used because a single turn main tuning is too coarse to set the frequency sufficiently accurately. With a 10 turn pot it should be possible to set the frequency accurately enough.

Peter

[keland_uk]

I suspect a few of us have made our own BOMs. My attempt attached.

1) R1 Doesn't exists on new schematic, but it is where the duplicate R30 is, so that may be cause of confusion. Its also 47K pot rather than a 22K, suspect either will do.
2) Use ceramic where its part of the original cct since HF, film ones for the new parts.
3) Adding a 10nF 400V+ to o/p of 3.3pF will protect that cct. I see you've put C8 in the same boat (Frequency o/p) Suspect that won't do any harm, and have added that to my BOM
4) I'm probably going to use a LDO voltage regulator, means C101 & C102 will change.

Hope this helps, and I haven't stepped on any toes

Will add example components when I get the time.



I've saved file as .xls so I can upload here (latest file version of Excel not supported by forum by the look of it)

Originally Posted by Bazz4CQJ

Hi guys, here is my preliminary offering by way of a BOM. I've shown it as a PDF which I've created in Excel. If anyone wants to chip in with spotting any errors or anything else, I'm happy to act as gate-keeper for the BOM.

EDIT; 4 mins later and I'm seeing the need to list the 400V caps as "notables"!
V2 shows it.

B

[Bazz4CQJ]

Originally Posted by Electronpusher0

Do you actually need a fine tune pot if you are using a multiturn pot for the main tuning?
A fine tune pot is used because a single turn main tuning is too coarse to set the frequency sufficiently accurately. With a 10 turn pot it should be possible to set the frequency accurately enough.
Peter

Yes, I guess that may well be right. I suppose I could make the initial build just using a link (made so that it can easily be replaced) in place of the fine control and see how that works?

Thanks to Keland for his BOM. I'll look at it later in the day and adopt any new info in to mine.

B

[Dylan85]

Not sure if this is the right place to post this thread but as there seems to be a fair amount of interest in the H-M Wobbulator I thought perhaps that comparing test results / frequency ranges etc might be useful reference.
Mods please move this if necessary.

I have got as far as a fully populated pcb which is fully functional but not in its case yet.
I have checked each of the 4 ranges of the oscillator and this is what I've found so far:-

Range 1. For this range only, I found the secondary winding wasn't required so have removed link JP2 and connected the lower end of the primary to 0V. With this arrangement the tuning range is approx 390kHz to 600kHz.

Range 2. Link JP4 left in place. Tuning range approx 1.2MHz to 2.0MHz.

Range 3. Link JP6 left in place. Tuning range approx 3.2MHz to 5.2MHz.

Range 4. Link JP8 left in place. This range is a little touchy. At the moment the frequency control potentiometer works from approx 10 to 15% to 100% of its rotation. In other words the lowest 10 to 15% causes the oscillator to start going unstable and then to stop. However, outside of this limitation it is working and the tuning range is approx 8.9MHz to 14MHz.

Please do bear in mind that the above figures are approximate and depend very much on where each coil is tuned along with potentiometer tolerances and also to some degree on wiring layout.

The current consumption on my unit when powered from 18V supply is 6.9mA. Please note that this does not include the 'Power On' led which I have not yet fitted.

I am happy with the initial results so far but I need to investigate Range 4 to see if that can be improved.

I hope these results are helpful.

Dylan

[Electronpusher0]

Thank you Dylan, the link arrangement echoes what I found.
I did not observe the instability on the highest range that you have.

Peter

[Bazz4CQJ]

Thanks for that feedback Dylan. Re the instability, I wonder which FETs you used? I have some 3819's which I know to be good, so I'm going to use those.

I'm about half-populated at the moment.

B

[Dylan85]

Originally Posted by Bazz4CQJ

Thanks for that feedback Dylan. Re the instability, I wonder which FETs you used? I have some 3819's which I know to be good, so I'm going to use those.

I'm about half-populated at the moment.

B

Yes I used 2n3819.as I had those in my 'store'.

Dylan

[HG MICKE]

A warning, check L4 as Spectrum have supplied 1u2L it should be 1R2 as in Peters assembly notes. On their ebay page the picture is of the correct one but in the list it is incorrect, fitting the correct one cured the instability problems.
For the FET I used 2N3819 and all works fine.

Colin

[Bazz4CQJ]

Many thanks for that Colin. I imagine that once a coil has been mounted on the PCB, removing it must be a pain, unless you had added some short "legs"?

I've also got the 1u2L, but have not yet mounted it.

B

[HG MICKE]

A pain is some what of an understatement. Unfortunately I have assembled two boards (one for myself and one for a friend) and getting those coils out was a task in itself.
After all of these years I should have had enough common sense to check everything at least twice before proceeding.
Colin

[Bazz4CQJ]

I think that, over time, the coil replacement issue has become a bit of a fuzzy area, so you've probably been unlucky.

I was already thinking that I would solder short leads on the bottom of the coils so that if there were problems, they could be got out much more easily than with the body flat against the board. I suspect that it is most unlikely that I will ever use the unit on anything other than 455kHz and I may just install that one coil to begin with. There are just so many projects I'd like to progress before spring when other things then need my time.

B

[Dylan85]

Originally Posted by HG MICKE

A warning, check L4 as Spectrum have supplied 1u2L it should be 1R2 as in Peters assembly notes. On their ebay page the picture is of the correct one but in the list it is incorrect, fitting the correct one cured the instability problems.
For the FET I used 2N3819 and all works fine.

Colin

Well now things are beginning to make sense. I had noticed the 1u2L ident label on L4 but assumed that those who supply have more knowledge on these things than I.

This morning I looked a bit more closely at the L4 issue and logic was telling me the problem had to be related to L4. Having taken a closer look at L4 I decided not to attempt to remove it from the PCB until all other options had been tried. The close fit of the pins in the pad holes would have made it very difficult to remove with out damaging it.

Next step was to raid the junk box. I have a dozen or so I.F. coils mostly new but some reclaimed, but none of the identifiers seemed close to what was needed. One clue was the core in the supplied 1u2L coil was above the top of the coil former but the inductance could not be increased further. So I thought I wonder if a change of core might help?
To cut to the chase, I found a tuning slug that fitted, eliminated the instability and gave the tuning range I think is needed. Result.

So if anyone has the same problem, before you attempt to remove L4, it might be worth trying a different tuning slug. The tuning slug I used came out of a 10mm Toko coil with red marker on top of the former and on the side of the screening can the part no / Id 38930 90389. I have a hazey memory that this coil assembly may have been used on the 'simple spectrum analyser' that some of our members may remember. (Another project I never got around to! I could be wrong on this one though.
I hope the above fix is helpful to others with the same problem.

Thanks to Colin for the info above.

Dylan

[Nanozeugma]

I am not aware whether any other contributor has approached Spectrum Communications regarding what appears to have been an error in supplying 1u2L instead of 1u2H/1R2 (?)
I have examined their website (not eBay) and seen that the two coils have their own entries and appear to be different.
The packing slip supplied by Spectrum with the coils from the eBay listing suggests that 1u2L is the correct substitute for TOKO KXNK3767EK whereas the eBay listing clearly illustrates 1R2 will be supplied.
In view of this, I have messaged Spectrum asking them to respond to what would appear to be their error.
I will update the thread as and when I receive a response from them.
Fortunately, I have not yet populated my board(s.)

N.B. Spectrum appear to be away from the business until Feb 26th.

[Nanozeugma]

I have received the following response from Spectrum Communications...
(Copied verbatim from their email reply.)

"This wasn't a mistake. The Haigh Wobbulator uses a Hartley oscillator for which the normal feedback is 1/4 to 1/5 of the main windind, as was the original coil type 3767 with 2t pins 1-2, 6t pins 2-3 and 1 5t secondary not used.

The 1R2 is the first of a new batch of coils which was not marked by the manufacturer as requested and should have been masrked 1u2H. It is a centre tapped coil 4t 1-2, 4t 2-3, and 2t secondary. It is NOT the equivalernt of the 3767.

The 1u2L is 2t 1-2, 6t 2-3 and 1t secondary. It IS the direct equivalent of the 3767. If the circuit is unstable using it then it is a function of the new board."

[HG MICKE]

That is rather odd, in Peters original notes he replaced the TOKO 3767 with a TOKO 3335 which obviously worked. On Spectrums site they list the 1u2H/1R2 as the replacement for the TOKO 3335.
Fitting the 1u2H/1R2 in place of the 1u2L that was supplied by Spectrum in two boards and both boards function perfectly. The 1u2H/1R2 coils that I had in stock were purchased from Spectrum last year.

Colin

[Electronpusher0]

Very Interesting, the first batch of boards I supplied I suppied as complete kits.
I have just checked back on my e-bay orderes and I ordered and supplied -
"Spectrum 10mm RF/IF transformers, type 1u2H (TOKO 3335R) for 15-35MHz, pack of 5"
ebay item-
https://www.ebay.co.uk/itm/282388294942

These all worked perfectly.

I had not realised that Spectrum had made a different subsititution for L4

Peter

[Nanozeugma]

Further correspondence from Spectrum received today...

"Further thoughts. If you look at the wiring diagram you will see that the switch wiring is very long and it is in series with the main winding. So if it is too long, or is close to other switch wires it will significantly add to the inductance above the tapping point. It will also be the cause of instability. Keep the switch wires as short as possible and clear of each other."

[Bazz4CQJ]

Given the sort of use I suspect most people will make of their wobbulators, and certainly my expectation of using it purely on 455kHz, the idea had already occurred to me to make the coils "Plug-in" on to a single base (maybe B9A?).

This sort of thinking is probably the outcome of owning an HRO for well over 50 years .

However, perhaps it wouldn't be the worst way forward. I think my initial build will simply have L1 installed.

B

[keland_uk]

Gosh it’s been quiet in here. How is everyone doing with their constructions?
Anyway, finally got round to finishing the soldering et al. Mine doesn't work. the 555 isn't doing anything. There was a previous thread (https://www.vintage-radio.net/forum/...46#post1294746, #17) that described this sawtooth generator, when it was using diodes rather than pseudo darlington pair, that implied R5 aka R26? should be 10R rather than 10k, to support the older 555's. All of mine are older 555's. Any pointers.... Old diagram attached, and gives some ideas where to look at. Sweep C21 just has a DC offset of around 3.4V

I am getting a nice sine wave at the cathode of D1 (varactor). Also getting o/p from the frequency output CN2. (Range 4 is giving me nothing at the moment)