Slow Mo MoFo

[Morgan Jones]

Thank you. Those last two references are much clearer. The writing of the TE White Paper was clearly subcontacted to Stanley Unwin. I have the kit to do the measurements and convince myself. And I've seen decaying pulse currents like that elsewhere so I'm predisposed to believe the counterintuitive.

[Nick]

Well, while I am not trying to convince anyone, it will be interesting if you do some of your own tests.

[MistyBlue]

I have always had a faint hum with my SissySit project; it has a big Toroidy transformer with a 230V primary…

I had another look at Toroidy last night (I did consider them originally, due to favourable comments by others), but went with Airlink partly to support UK manufacturing, and avoid any surprise import duties that I have had in the past whenever it seems Federal Extortionists get involved as couriers.

Just to be safe, I would have then tried it direct to the mains just to make sure the variac wasn't limiting current and making the waveform even worst.

I’m pretty sure I did this last night, but testing again (just in case I slipped a cog last night due to illness, probably some mutant flu/covid no doubt picked up from Torbay Hospital on a family emergency earlier in the week) Today, Monday 16:46 mains is 242.6V. Amplifier connected direct to mains, with soft start, result: buzzzzzzzz.

They have elected to plug 230V into their transformer design software and the result is that it has told them to reduce primary turns. What they should have done is to plug 230V +10% into their software. What you are hearing is the result of operating the core too close to saturation.

Apologies in advance for what could be construed as veering towards a slight rant, but I feel entitled. And “miffed”.

Hopefully my comments in the paragraph below will be taken as read and purely apolitical.

After researching this a bit more last night, as Morgan says, the issue is largely one of the way the standardisation for mains voltage has been implemented. More of a paperwork exercise designed to accommodate the differences in mains supplies between EU members all within the guidelines, rather than to impose changes on EU members. They just widened the tolerance range. Hence, the UK carries on as normal. Since we left the EU, it has not changed either.

Like a warm, fluffy legislative blanket. I like warm fluffy blankets, and, incedently, we both voted to stay in the warm fluffy blanket.

As I understand it, and as Morgan has pointed out, transformer manufacturers seem to have taken this as a legislative opportunity to then “standardise” everything at 230V and turn a blind eye to variations. That’s fine, I understand. I love standardisation. I probably would do the same if I was in their position as a manufacturer. Makes things simpler on the shop floor.

Unfortunately, back on planet MoFo, for me as a DIY user of end products, I am unlikely to be able to wield any influence for a one (or even two) off prototype for a domestic amplifier. Maybe, for a run of 100, I might hold some sway.

I could have a go at vacuum impregnation of the windings - I have a large(ish), welded steel vacuum chamber that we built years ago and suitable rotary vane pump for potting experimental guitar pickups (amongst other things) and general moulding activities. Even still, this won’t take into account core saturation (the root cause) or appropriate primary winding.

But part of me says “Why the hell should I?”. It offends my sensibilities that when I buy an end product it is literally because I have neither the time, skill, confidence, or equipment to do the job myself, and I pay a professional outfit to do this properly by purchasing their products.

I could give Airlink a call and have a polite moan. It is unlikely to achieve anything. I am sure that they are as just as capable as any other of designing and making a suitable transformer (with the appropriate oversizing of the core, appropriate primary windings or whatever). However, they may not be tooled up to do the impregnation etc (i.e. to make sure, sure and damn sure that the buzzing doesn’t happen, within justifiable limits), and may be unwilling to entertain me, who after all, is just a lone DIYer, building amplifiers at home.

I did have a look at Canterbury windings, but it did seem to me that he had retired (thanks Nick for confirming that).

I had a look at Rod Elliott’s piece on bucking transformers. Very interesting. But again “Why the hell should I have to do this?!” comes to mind.

Another option I have come across is Tiger Toroids in Norfolk. Interestingly, a few examples of their work notably have multiple primary taps ranging from 220V to 250V and have displayed sympathies (as in product design considerations) to DIY audio constructors. A few others have mentioned them favourably over on DIYA and report that they are willing to have a friendly chat about their products over the phone. I check out many of the outfits I deal with before I get involved (companies house check etc), especially small ones, to avoid the worst operators. I might give them a go.

But this time, instead of being hopelessly naive, I think I will go armed with a decent idea of what I am looking for in a toroid for this application, instead of assuming an off the shelf product should do. I will do this in the next few days and report back.

Thanks for the links to the info regarding soft starts and toroids. I will read up some more later on this evening if I’m feeling up to it.

[simon]

Entirely reasonable and pretty restrained I reckon. I hope you get a good 'un from somewhere.

[Nick]

They have elected to plug 230V into their transformer design software and the result is that it has told them to reduce primary turns. What they should have done is to plug 230V +10% into their software. What you are hearing is the result of operating the core too close to saturation.

Actually, I think it is even worst that that. Its 230+10% + not a perfect sine, often with the top of the wave cut off + not always a symmetrical sine, so there is a net DC offset due to half wave loads.

I have had better results with Vigotronix than Airlink, but the best I know of now is Toroidy. I suspect if you ask them they would wind for 250v core size. But then you are paying for a custom product.

[Nick]

Just had a thought. These are the same transformers that you had the problems with Quazimodo? Yes? I wonder if the current spikes of that caused any problems, creating a shorted turn in there.

[MistyBlue]

Just had a thought. These are the same transformers that you had the problems with Quazimodo? Yes? I wonder if the current spikes of that caused any problems, creating a shorted turn in there.

Actually, those were Vigortronix (used on the new PSU for the active crossover) that I had the problem with IIRC. But I certainly had problems "ringing" the Airlink. I put it down to not having enough excitation voltage (but I'm hardly qualified to make that assertion).The R core for the MoFo valve front end responded as expected.

I also note that Rod Elliott also has his own ideas about snubbers :https://sound-au.com/articles/psu-snubber.htm

[Morgan Jones]

I doubt that the Quasimodo has enough grunt to damage anything. And toroids tend to have quite robust insulation.

I had noticed that Rod Elliott doesn't believe snubbers are worthwhile. Just wait until he encounters a split bobbin EI transformer having significant leakage inductance...

[Greg]

It’s been a while but in the past I have dealt with Airlink and found them very helpful. They always used to offer a bespoke winding service. If they still do, there might, under your circumstances be a deal you can do.

[MistyBlue]

It’s been a while but in the past I have dealt with Airlink and found them very helpful. They always used to offer a bespoke winding service. If they still do, there might, under your circumstances be a deal you can do.

Thank you Greg for reminding me of the importance of due process, fair play and the opportunity for redemption.

I just had a chat with Mike Norman at Airlink Transformers and discussed my situation with him. He was nice enough to give me a very reasonable amount of time to discuss options for a solution, and point to a way forward. Airlink do indeed still do custom winding and a very reasonable offer was made, which I have accepted.

Further reports to follow.

[MistyBlue]

While I’m waiting for the new toroid to arrive from Airlink, I thought I would try some measurements on the 6Z9PE front end.

Big thanks to Nick for making this possible with the sound card interface.

6Z9PEFrontEndMeasurements001.jpg (112.22 KiB) Viewed 5188 times

6Z9PEFrontEndMeasurements002.jpg (115.4 KiB) Viewed 5188 times

I managed to replicate Andrea’s measurements at around 0.6VRMS input (checked with MM), giving slightly higher distortion at 1.72% for 19.8VRMS output, compared to his at 1.59% for 17VRMS, but still near enough to confirm his results.

At 2VRMS, things start to look nasty with higher order harmonics becoming visible above the noise floor at 5.41% THD, but hardly unexpected. This may explain some of that sharply etched HF hash I was hearing.

6Z9PEFrontEndMeasurements003.jpg (101.25 KiB) Viewed 5188 times

Currently, I have 3 different options for the front end, this one (6Z9PE), the 6072 and the FE2022 JFET. The plan is to measure each one, listen to them and decide on a winner for each application (LF/HF sections of the speakers).

[Nick]

Cool, glad its being used again.

[Mike H]

[Morgan Jones]

That's interesting hum you have on your spectrum. 50Hz and 150Hz suggests a saturating transformer.

[Nick]

I assume you can add some averaging to the FFT, that will clean up the HF noise and show any harmonics that are lurking inside it. The second trace is basically clipping, I would not make any interpretations from that how it will sound.

Also, I would ground the input to the interface box and see how close to a flat line you can get before doing any measurements. Getting the test kit right is always the first step, those hum signals may or may not be from the kit under test.