I'd like to start a discussion about K6 SMPS.
HERE is the original link. I think you know it already.
I put HERE a more clear schematic made by me.
And HERE what i have done so far.
I can't connect this on the main power because i don't understand all schematic, and the most thing i hate is the IC2b integrator
I know , i'm a beginer, but what can i do now ? I must finish this.
So, I'll be very happy if anyone can explain me. At last tell me if it's right so i can go on
HERE is the original link. I think you know it already.
I put HERE a more clear schematic made by me.
And HERE what i have done so far.
I can't connect this on the main power because i don't understand all schematic, and the most thing i hate is the IC2b integrator
I know , i'm a beginer, but what can i do now ? I must finish this.
So, I'll be very happy if anyone can explain me. At last tell me if it's right so i can go on
I think Tony from Russia a few posts back hit the nail on the had. I've used this approach and the results are great - ear up against speaker cones and there is zero noise from tweeters, mid or woofers (yes, the amp is powered up).
Bonsai said:
yeah, worked for me all the time....
btw, i am from manila working in the sahkalin island build an LNG refinery....
😀
HAPPY EASTER TO ONE AND ALL......😀
AKSA said:
Hugh, are you talking "pi" filter here, or using resistor/inductor on the ground sides?
torroidal power transformers
I recently substituted a torroidal power transformer in an amplifier where there had been a conventional one. While doing some detailed distortion analysis, where THD and IM products were hovering in the -100 dB range, I noticed a spray of 120 Hz harmonics disguising themselves as spectrum analyzer noise floor (out to several kHz or more). I'm pretty sure I hadn't seen that before. I also noticed that the position of the input and output wires to the amp under test had a big influence on this. Has anyone else seen this effect with torroidal power transformers, or am I just imgaining things. Admittedly, I was just using an ordinary 30A bridge rectifier.
Bob
I recently substituted a torroidal power transformer in an amplifier where there had been a conventional one. While doing some detailed distortion analysis, where THD and IM products were hovering in the -100 dB range, I noticed a spray of 120 Hz harmonics disguising themselves as spectrum analyzer noise floor (out to several kHz or more). I'm pretty sure I hadn't seen that before. I also noticed that the position of the input and output wires to the amp under test had a big influence on this. Has anyone else seen this effect with torroidal power transformers, or am I just imgaining things. Admittedly, I was just using an ordinary 30A bridge rectifier.
Bob
Re: torroidal power transformers
I have had the same effects when "standard" toroidal trans was in use. After replacing of last one with "special" trans the problem has gone. "Special" means reduced induction (down to 1.1-1.2Tl instead of 1.4-1.5Tl), which was achieved by rising turns amount (about 40-50% above "standard"). A drawback is, you need a wire with bigger diameter, or be ready a trans will be warm and "soft" - I mean secondary resistance. Very simple.
Bob,Bob Cordell said:
I have had the same effects when "standard" toroidal trans was in use. After replacing of last one with "special" trans the problem has gone. "Special" means reduced induction (down to 1.1-1.2Tl instead of 1.4-1.5Tl), which was achieved by rising turns amount (about 40-50% above "standard"). A drawback is, you need a wire with bigger diameter, or be ready a trans will be warm and "soft" - I mean secondary resistance. Very simple.
Re: Re: torroidal power transformers
Very good observation. Are these "special" toroidal transformers readily available anywhere? Do toroidal transformer makers normally spec the amount of induction? I am certainly not much of an expert on transformers.
Finally, can I get a relative measure of the amount of induction by knowing the volts-per-turn of the toroid (i.e., wind ten turns on it, measure the voltage, and divide by ten)?
Thanks,
Bob
anli said:
Very good observation. Are these "special" toroidal transformers readily available anywhere? Do toroidal transformer makers normally spec the amount of induction? I am certainly not much of an expert on transformers.
Finally, can I get a relative measure of the amount of induction by knowing the volts-per-turn of the toroid (i.e., wind ten turns on it, measure the voltage, and divide by ten)?
Thanks,
Bob
Re: Re: Re: torroidal power transformers
Bob,
I'm not trans-guru also. I (and many other Russian DIYers) order toroidal transformers in a firm which accept single-amount orders and all customer's demands (Vn, I, P, ternary steep and so on), and know how to achieve acceptable inductance (the firm's fellows call such orders as "audio" 😉 ). So, I'm a customer here only. The firm is Moscow-located, I think it isn't handy for you. At any case any decent firm MUST can to calculate such trans.
Of course, there are DIY-fellows which can calculate such trans, but I prefer to design schematics rather transformers
🙂
Bob Cordell said:
Bob,
I'm not trans-guru also. I (and many other Russian DIYers) order toroidal transformers in a firm which accept single-amount orders and all customer's demands (Vn, I, P, ternary steep and so on), and know how to achieve acceptable inductance (the firm's fellows call such orders as "audio" 😉 ). So, I'm a customer here only. The firm is Moscow-located, I think it isn't handy for you. At any case any decent firm MUST can to calculate such trans.
Of course, there are DIY-fellows which can calculate such trans, but I prefer to design schematics rather transformers
🙂
I had an Tektronix SG505 oscillator (which is a really excellent test oscillator) -- the SG505 has its own power supply (unlike the SG502 which relies upon backplane DC from the TM5xxx) -- I found that by putting a piece of mu-metal shielding (which I had insulated) around the transformer I could get the THD down below 0.001% -- on a newer model SG505 which replaced the old one the THD was 0.0008% -- I don't know if there was an intentional fix by the folks in Oregon or if it was just random variation and I happened to get a better unit.
At any rate, a little mu-metal cures some transformer woes --- just remember that it doesn't like sharp bends.
Jim Williams at Linear Tech had an application note (which I can't find) in which the radiation patterns of various transformers were illustrated.
At any rate, a little mu-metal cures some transformer woes --- just remember that it doesn't like sharp bends.
Jim Williams at Linear Tech had an application note (which I can't find) in which the radiation patterns of various transformers were illustrated.
Re: Re: torroidal power transformers
implies lower core losses, magnetising current is likewise smaller, the bigger wire is needed to keep the copper losses down, you can do the same thing with EI's...
anli said:
implies lower core losses, magnetising current is likewise smaller, the bigger wire is needed to keep the copper losses down, you can do the same thing with EI's...
Hi,
Can I assume that the extra turns on the primary give a different solution than adopting a higher VA rating.
It seems that the thicker wire and more turns solution is going back to the low regulation transformers that were more common 20 or 30 years ago.
Maybe all we need to do is select transformers on the basis of regulation (plus electrostatic screen).
Any thoughts?
Can I assume that the extra turns on the primary give a different solution than adopting a higher VA rating.
It seems that the thicker wire and more turns solution is going back to the low regulation transformers that were more common 20 or 30 years ago.
Maybe all we need to do is select transformers on the basis of regulation (plus electrostatic screen).
Any thoughts?
jackinnj said:
The SG505 was indeed a great oscillator, designed by my friend Bruce Hofer at Tek before he co-founded AP.
Bob
OK, suppose we are stuck with a given transformer, which may often be the case. How can we reduce this tendency to radiate a spray of 120 Hz harmonics? Is there some kind of better filtering we can do at or near the ac side of the rectifiers? Will soft-recovery diodes help this? I've tried snubbing and shunt capacitance with limited improvement. Its just surprizing that the toroid was so much more prone to this than a conventional transformer.
Bob
Bob
Bob,Bob Cordell said:
EM-radiation depends little on current consumption. The only way to reduce this negative effect for given transformer is to cover last one with ferrous "bowl". 1mm thickness reduces radiation ~8-10db. Just try in real-time (steal suitable iron bowl from your kitchen) and see a magic 🙂
Bob Cordell said:
Just thinking aloud Bob.
From what you post it seems you are not certain whether the spray is conducted all the way from the rectifier, or whether it is radiated from the transformer and picked up.
Perhaps removing the transformer and connecting with a long thick cable may help, moving it around, to check if something changes.
If it is conducted - which seems more likely - perhaps the culprit is the lower leackage inductance inherent in a better confined magnetic circuit as is the case for a toroid in comparison with a I-E or E-E lamination. If that were the case, then a small input choke will also make for a measurable difference, just to make sure.
Just a cent, and keep us posted on your results.
Rodolfo
I filter my rectifiers, fast or slow, and haven't had a problem with radiated or conducted HF stuff. As for LF magnetically induced noise, I've had little success with mu-metal for power applications. IMO, things need to be fully encased, and bending and joining the stuff reduces the performance unless you can re-anneal it. It's been ok for small signal shielding, but so has regular steel.
Here's a crazy trick that has worked for me. Shield your magnetic fields with aluminum! Foil won't work. 1/4" thick plate won't work. 3/8" plate starts to work, but isn't great. But, 1/2" plate has so much eddy current loss that it can be more effective than steel or mu-metal. It isn't even that expensive, but you need a way to cut it (band saw). I was working on some very sensitive amplifiers (tunneling microscope signal conditioning using OPA627s) and kept picking up transformer fields. A thick aluminum plate between the amps and tranny did the trick, when little else was effective. Being a commercial product, they said "no way" due to cost. The ultimate solution was to mount the transformers diagonally, redirecting the fields away from the sensitive amps. Take your pick.
Here's a crazy trick that has worked for me. Shield your magnetic fields with aluminum! Foil won't work. 1/4" thick plate won't work. 3/8" plate starts to work, but isn't great. But, 1/2" plate has so much eddy current loss that it can be more effective than steel or mu-metal. It isn't even that expensive, but you need a way to cut it (band saw). I was working on some very sensitive amplifiers (tunneling microscope signal conditioning using OPA627s) and kept picking up transformer fields. A thick aluminum plate between the amps and tranny did the trick, when little else was effective. Being a commercial product, they said "no way" due to cost. The ultimate solution was to mount the transformers diagonally, redirecting the fields away from the sensitive amps. Take your pick.
Because of its very very high relative permeability, the mumetal easily saturates, and that is the problem.
Skin depth of aluminium is some 11.5 mm for 50Hz, so we have to make it thick 😉 . Actually, it does not saturate.
I am designing measuring instruments for high power labs. My last case/housing is milled from 8mm Al plates (outer case), and there is further Al or Fe inner case, depending on application.
Skin depth of aluminium is some 11.5 mm for 50Hz, so we have to make it thick 😉 . Actually, it does not saturate.
I am designing measuring instruments for high power labs. My last case/housing is milled from 8mm Al plates (outer case), and there is further Al or Fe inner case, depending on application.
- Status
- Not open for further replies.