 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Good money after bad - Solar upgrade time
  Page 4 of 4 |
|||||
| Author | Message | ||||
wiseguy Guru  Joined: 21/06/2018 Location: AustraliaPosts: 1193 |
I have seen and experienced this exact phenomena, I consider it to be a nuisance type issue. I put it down to random probably nano/pico second differences in the nano just executing the code. If enough of these slight perturbations occur in the same direction ie always on the positive cycles the Toroid Flux wanders towards (very mild) saturation on one side of the half cycles but eventually it slowly wanders back the other way. Sometimes there is no noise for multiples of minutes, even tens of minutes but it is never gone. The corresponding slight increase in current during the mild saturation would also be expected. The reason this does not continue on to become a real issue is that the slight error causing it means there is not enough energy in the error to put it into greater saturation. Note a toroidal transformer in an amplifier or similar device can make a hell of a "boing" sound followed by a few cycles of hum that decreases rapidly, if you catch it on the wrong part of the cycle after a turn off turn on event. I mention this to illustrate that the tendency of a transformer is to try to reach a level of equilibrium after an upset, maybe the AC has a natural tendency to demagnetise & remove any saturation naturally. The Power PCB FET holes were made to accommodate a standard 3.6mmOD TO220 Bush, and with the long necked variety yes the bush shoulders continue into the TO247 FET hole. Something similar to the Digikey part 12SWS0400, it is better not to have to drill out the circuit board plated through holes.... Edited 2025-04-30 00:10 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
oreo Regular Member  Joined: 11/12/2020 Location: CanadaPosts: 97 |
Interesting. Thanks for the update Mike! Greg |
||||
| analog8484 Senior Member  Joined: 11/11/2021 Location: United StatesPosts: 131 |
Thanks for the explanation. I was wondering because I've seen quite different ratings from different manufacturers for similar size/weight transformers. |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 131 |
Thanks for the video. It sounds like some kind of harmonic resonance. Especially odd that it happens under load. Not sure if it makes a difference but I remember seeing the timer resolution in the Nanover is relatively low and accuracy for 60Hz is not as good as the 50Hz config. |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 131 |
Not sure if it's still used in the latest Nanoverter code but I remember seeing a simple fixed correction for half-cycle balance that is supposed to be tuned for each inverter? Perhaps that's at play here. |
||||
KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1816 |
You are going back to the early code development days and "per inverter" would imply a kludge, obviously nothing like that here. I assume you are referring to stability, so no, both are stable and adjustable slightly in frequency (in mine) There will be small differences between 50Hz and 60Hz with a given Toriod, however if it's wound correctly you should see no real difference in operation, maybe slightly lower idle at 60Hz. NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1193 |
Something else that affects the stability of each half cycle of the sine is how well the timing variations are matched between the 4 FET drive optocouplers. There are propagation delays (for both, high going and low going events) + a distortion component in the optocouplers. These are specified at 25 Deg C and no doubt can also vary with temperature, along with the capacitor stability for the dead time capacitor in parallel with the Opto LEDs. Now consider the 4 combined FETs on/off rise & fall time for each leg of the bridge as well. So is it any wonder that there may be slight differences between the integration of the total switching on and off times for each half cycle? I think its actually surprising that there are only minimal issues. Perhaps prior matching of each sides of the bridge H & L Opto couplers timing would pay good dividends - maybe some builds fluke the matching better than others? If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1816 |
Thanks, I had not thought of most of those, the list is growing longer. Multiply by two for Dual power boards and Toroids, or maybe they cancel out  NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| oreo Regular Member Joined: 11/12/2020 Location: CanadaPosts: 97 |
Toroid Transformer Noise I spent about 25 years working for a prominent Canadian speaker company who used transformers in many of their products. Audio customers did not want to hear any transformer noise, so we spent a fair bit of time working to reduce transformer noise. Skip to the recommendations in the next post if you just want to know the recommendations. Most of the transformers we used, were common EI style transformers and we worked with just 2 manufactures to determine what worked best. One of the suppliers built a transformer audio testing chamber, so they could accurately judge the effect of process changes. In a nutshell, for EI transformers it was found that putting them through vacuum impregnation varnishing twice was the most cost effective way of getting a particular design as quiet as possible. Of course we also used isolation mounting techniques and cabinet/chassis designs which reduced the transmitted noise and the flux density used in the design influenced things significantly. While most powered subwoofers used EI transformers, nearly all the receivers and preamp/amplifiers we built used toroidal transformers. While toroid transformers are normally quieter than EI designs, we had more issues with toroidal transformers than we did with EI transformers. When you turned on the product, the toroid would make a louder buzz than normal, which would make the customer hyper sensitive, listening for any noise. At one point, we were dealing with 2 Canadian suppliers, 3 Chinese suppliers and Noratel which built product in Sri Lanka and had issues from all our suppliers. From any particular production batch, there would be some noisy transformers. Some suppliers were worse than others. We took apart product and found that no-one was varnishing the transformers. We approached all our suppliers and asked them to varnish the transformers and not one of them said they would do it. Our local supplier with the sound chamber were more open, and said they would do some testing for us. They tested varnishing the core, and varnishing the windings. The results were not intuitive. Varnishing the core increased the noise. Varnishing the windings had no measurable effect. Keep in mind that these results are for production transformers, with tight windings and production controls that enforce good build practices. This testing was done with low distortion 50 and 60 Hz power sources. Greg |
||||
| oreo Regular Member Joined: 11/12/2020 Location: CanadaPosts: 97 |
Toroid Build recommendations 1. Low Flux Density. Many are doing this already to get low standby power draw, and it has a major impact on noise. (my transformer is a commercial design, which I have improved upon slightly, but probably has a relatively high flux density compared to most of you) 2. Isolate the transformer from the case. Most transformers are provided with neoprene pads. Adding further isolation will help as these are pretty thin and IMO not the right durometer. Rubber isolate hold down screws and don’t tighten them too tightly. (right now I have my transformer resting on nylon plastic bars. This is making the noise worse) 3. Make chassis with dense sound absorbing material. KeepIS has a chassis made of wood/MDF which is a good choice for noise control. My chassis is aluminum/steel and could benefit from some damping. (I ordered some damping material for a car project and plan to use a bit of it in my Inverter) I will report on it's effectiveness. Of course since this is DIY, other factors will come into play. If you make your windings loose, then you may benefit from Varnish and as long as you don't varnish the core, it may be a good idea as an insurance measure. Greg |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 131 |
Kinda of surprising result. Good to know no varnishing needed/recommended is another plus for toroidal transformers. |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 131 |
It seems the presence of DC voltage is the most likely cause of the transformer noise. I wonder if the DC voltage of the power board output could be measured to show noticeable increase when the noise occurs? |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1816 |
Cabinet is 19mm hardwood ply, the Base is a separately constructed Heavy "Box sectioned" frame, containing two 100mm heavy cast framed 240Vac fans, one under each Toroid, fans completely isolated from the frame which is skinned with thin 3mm hardwood ply and supported by five 200kG wheels. You would think it would be sound deadening, but the base is partly hollow, it has resonance and easily transfers any noise from the effects of unbalanced AC loads in the Toriods. I doubt that anything you do will stop a Toroid CORE from ringing when abused. The potted ones still hum in GTIs. However I understand that you have noise at Idle. I can reproduce a similar noise on the bench test unit, I'm not suggesting it's the cause of your problem, but the results are similar. I think I know why this happens in the test unit, it may have something to do with what I found when trying to set static resonance in a Toriod via the Secondary CAP value, with a simple test to prove what I was seeing, I realized it was like trying to heard cats, it's why we have soft ramp down, as well as Soft ramp up. Wiseguy suggested flux walking and getting closer to saturation, either by Toroid design decisions or inducing flux to move towards saturation and this certainly appears to be a likely cause. To your point on Flux density, my Toroids have an effective 280Vac secondary. With a 14 turn primary they are still Quite when "AC->AC" testing at 275Vac output, the saturation power curve knee hits hard past 280Vac, hum/buzz/hi current appear quite rapidly as I move into saturation past 280Vac. They are silent 99.9% of the time at 240Vac in the Inverter, they are silent and 240Vac regulated just above 40Vdc. DC voltage on the Power boards should not change at idle with a few watts of power cycling. I have a permanent DSO monitor output on the front of the Big inverter, connected to the output of the Main 700A DC current Hall sensor for current waveform monitoring. And I just remember the following because I tested for it. It was Poida who suggested a current sensor on the DC input could be used to identify the beginning of saturation, with small current peaks appearing on the DC current waveform, the peaks coincide with Zero crossing of the AC waveform, I may look at the Test setup again and see if I can see that. BTW: There is a reason I don't use metal underneath these two x 46Kg Toroid stacks, it has to do with purposeful design choices I made for this build and part of the reason these Toroids and Chokes run so cool. . Edited 2025-05-02 12:59 by KeepIS NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| Page 4 of 4 |
|||||
 Print this page |
 |
The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2025 |
