hey guys,
i have a little doubt, due to space constraints, i have decided to leave out my LM series of regulators and its counterparts, i have a very well developed 317/337(http://audio.gotroot.ca/minipow/) supply for project 88 preamp but due to space constraits,it uses up a lot of space.
can something simpler like this work? or will it defeat the whole purpose of HIFI system by injecting noises??
i have a little doubt, due to space constraints, i have decided to leave out my LM series of regulators and its counterparts, i have a very well developed 317/337(http://audio.gotroot.ca/minipow/) supply for project 88 preamp but due to space constraits,it uses up a lot of space.
can something simpler like this work? or will it defeat the whole purpose of HIFI system by injecting noises??
Attachments
If the transformer is 2x12V it gives the desired output for ~15mA.
The ripple rejection isn't that good.And zeners produce noise.
You could use the TL431 (LM431) with a 10k/39k dividers in the place of the zeners.
They have a much lower Ri and noise.
Mona
The ripple rejection isn't that good.And zeners produce noise.
You could use the TL431 (LM431) with a 10k/39k dividers in the place of the zeners.
They have a much lower Ri and noise.
Mona
the output of the rectifier will be about 17Vdc.
The drop across the 220r will be ~5V. Pdiss ~ 110mW
The maximum output current will be ~22mA.
If the output is not connected the Zener/s will dissipate ~ 270mW
The circuit must draw an average current of between 1mA and 21mA to keep the zener voltage near it's nominal value.
The maximum current cannot exceed 22mApk.
One single channel opamp that has a 4mA quiescent current and a maximum output of 10Vpk into 1k loading (load//feedback) will draw ~12mApk.
Your circuit may struggle to power a dual opamp.
The drop across the 220r will be ~5V. Pdiss ~ 110mW
The maximum output current will be ~22mA.
If the output is not connected the Zener/s will dissipate ~ 270mW
The circuit must draw an average current of between 1mA and 21mA to keep the zener voltage near it's nominal value.
The maximum current cannot exceed 22mApk.
One single channel opamp that has a 4mA quiescent current and a maximum output of 10Vpk into 1k loading (load//feedback) will draw ~12mApk.
Your circuit may struggle to power a dual opamp.
power supply
What about using the LM7812?
Would this offer less noise?
Thanks ang good luck with your build
What about using the LM7812?
Would this offer less noise?
Thanks ang good luck with your build
My thoughts for what they are worth...
the circuit in post #1 is fine in principle but it really needs more "compliance" a 15volt tranny would be far better together with an increase in the resistor values. A simple R/C filter after the zeners would reduce any noise to essentially non existent.
Its not that far removed from the scheme I use in my amp although that feeds the zener regulated supplies from a -/+35 volt pre-stabilised rail.
the circuit in post #1 is fine in principle but it really needs more "compliance" a 15volt tranny would be far better together with an increase in the resistor values. A simple R/C filter after the zeners would reduce any noise to essentially non existent.
Its not that far removed from the scheme I use in my amp although that feeds the zener regulated supplies from a -/+35 volt pre-stabilised rail.
@andrew i can do with lesser use of space. your calculations are concrete and given i am using two dual op amps. its gonna be one tough cookie for the zeners to power it.
@bfg as much as i have all the positive rail regulators with me, i barely have any negative rail regulators except 317 and 337. these regulators need more parts so gonna be tough in putting in small places. most probably going to build that only..
@mooly i have 36vct tranny. but my zeners only rated at 1w 🙁
@bfg as much as i have all the positive rail regulators with me, i barely have any negative rail regulators except 317 and 337. these regulators need more parts so gonna be tough in putting in small places. most probably going to build that only..
@mooly i have 36vct tranny. but my zeners only rated at 1w 🙁
you did not understand what I posted.
1.) the resistor is dissipating only 110mW
You can change the resistor to get more current.
2.) the Zener is dissipating only 270mW
you can increase the current a LOT for a 1W Zener.
3.) you NEED to know the MINIMUM current draw and the MAXIMUM current draw to design your PSU.
so should i reduce the value of the resistor to increase the current?
if i reduce the resistance, the zener will dissipate more heat..maybe too much for the little soul to handle.a 1w rated zener shouldnt be put through hours of torture and high power dissipation.
Interestingly, i found this Power Supply for Preamps which uses simple 78xx/79xx linear supply(ive used this before) the advantage it gives me is that i dont need a pot/trimmer to precisely set the voltage when 317/337 is used.hereby again reducing space, but on the finished board, i see, that he says the HF noise is massively annutated.
I also see in the finished board,a resistor and a extra caps. can it be that he is using an RC filter towards the output?
if i reduce the resistance, the zener will dissipate more heat..maybe too much for the little soul to handle.a 1w rated zener shouldnt be put through hours of torture and high power dissipation.
Interestingly, i found this Power Supply for Preamps which uses simple 78xx/79xx linear supply(ive used this before) the advantage it gives me is that i dont need a pot/trimmer to precisely set the voltage when 317/337 is used.hereby again reducing space, but on the finished board, i see, that he says the HF noise is massively annutated.
I also see in the finished board,a resistor and a extra caps. can it be that he is using an RC filter towards the output?
Rod Elliots project 88 uses two OPA2134. Current per package, 8 milliamps typical value. Forget worst case loading at full output because its just never going to happen.....
You say you have a 36 volt CT tranny. That gives -/+25 volts dc. You can run the opamps on -/+ 18 volts dc using 18 volt zeners.
25 - 18 gives 7 volts to "lose" across the resistors. A tiny 0.6 watt 18 volt zener will withstand just over 30 milliamps. A 1.3 watt zener, 80 milliamps.
Lets go with the tiny 0.6 watt. A 270 ohm resistor would allow for 25 milliamps to flow in the zeners. The resistor would need to be a 0.25 watt component. When the opamps are connected they would draw typically 16 milliamps. That means the zener current falls from 25ma to just 9 milliamps. That 9 milliamps is available for the opamp output current that may be needed. Doesn't sound like much... how many preamps deliver 9 ma to the next stage. If you want more headroom then you can lower the resistor. Won't that overload the zener ? No, because as long as the opamps are present (and the supply isn't running unloaded) then the opamps are drawing a minimum of 16 ma away from the diodes. So you could make the resistors 180 ohm. That's nearly 40 ma but remember 16 of those always goes to the opamps leaving 24 "free" for the load to use and still only giving a dissipation of 0.43 watts in the zener.
Its easy and reliable. Put your own numbers in. Try it for your 12 volt zeners or your lower voltage tranny
You say you have a 36 volt CT tranny. That gives -/+25 volts dc. You can run the opamps on -/+ 18 volts dc using 18 volt zeners.
25 - 18 gives 7 volts to "lose" across the resistors. A tiny 0.6 watt 18 volt zener will withstand just over 30 milliamps. A 1.3 watt zener, 80 milliamps.
Lets go with the tiny 0.6 watt. A 270 ohm resistor would allow for 25 milliamps to flow in the zeners. The resistor would need to be a 0.25 watt component. When the opamps are connected they would draw typically 16 milliamps. That means the zener current falls from 25ma to just 9 milliamps. That 9 milliamps is available for the opamp output current that may be needed. Doesn't sound like much... how many preamps deliver 9 ma to the next stage. If you want more headroom then you can lower the resistor. Won't that overload the zener ? No, because as long as the opamps are present (and the supply isn't running unloaded) then the opamps are drawing a minimum of 16 ma away from the diodes. So you could make the resistors 180 ohm. That's nearly 40 ma but remember 16 of those always goes to the opamps leaving 24 "free" for the load to use and still only giving a dissipation of 0.43 watts in the zener.
Its easy and reliable. Put your own numbers in. Try it for your 12 volt zeners or your lower voltage tranny
Zeners seem to quite long lived when operated at disipations that approach their Pmax.
A 12V 1W Zener can pass 83mA
If you select a resistor that passes about 70% to 80% of that then the Zener should be reliable.
That resistor current will heat the resistor and the Zener. Check Pdiss for both. Select a resistor Pmax that is at least double Pdiss.
The maximum output of the supply is now resistor current minus 1mA.
Check to see if that is enough assuming each opamp needs ~8mA to 20mA for the loads you have to drive.
It isn't as simple as that.
For 1W dissipation to be accepteble there are some conditions.
Short leads to a copper plane ,not to small so the temp.will not go over 50°C.If not the device will overheat.
Mona
Attachments
is that the copper plane, or the leadouts, that are @ 50°C?
Never seen/noticed that value before.
Zeners seem to operate with a long life even when dissipating a lot of power.
now have the datasheet.
Still not sure what they mean.
But reading yet again, I think they mean that the Lead temperature at 3/8ths" (TL) must invoke or not the temperature de-rating.
If the temp is at or below 50°C then the full 1W capability is allowed. There appears to be no Ta de-rating.
If the 3/8ths away temp is above 50°C then de-rating must be applied.
Eg temp at 3/8ths away is 65°C then 65-50 * 6.67C/W is applied resulting in 900mW,
if it rises to 80°C @ 3/8ths away, then reduces to 800mW
Is that the way you would read it?
Still not sure what they mean.
But reading yet again, I think they mean that the Lead temperature at 3/8ths" (TL) must invoke or not the temperature de-rating.
If the temp is at or below 50°C then the full 1W capability is allowed. There appears to be no Ta de-rating.
If the 3/8ths away temp is above 50°C then de-rating must be applied.
Eg temp at 3/8ths away is 65°C then 65-50 * 6.67C/W is applied resulting in 900mW,
if it rises to 80°C @ 3/8ths away, then reduces to 800mW
Is that the way you would read it?
The max. chip temp. is 200°C. If the LT = lead temp is 50°C and you dissipate 1W the chip reach 200°C ,that's why it is the maximum.So you have 150°/W from lead to chip.Supose you could keep the leads at 20°C then dissipation can go as high as 1.2W !
Mona
Mona
i believe semiconductors dont perform too well when dissipating a lot of heat. but yes, zeners dont break down so easily.
@mooly thanks a lot for your help. ill try to understand the theory you posted 🙂
But i really want to know apart from basic zener theory, how do i calculate the RC filter to make the supply quiet?
@mooly thanks a lot for your help. ill try to understand the theory you posted 🙂
But i really want to know apart from basic zener theory, how do i calculate the RC filter to make the supply quiet?
Last edited:
Your welcome. If you are not sure then ask 🙂
Calculating something requires you to know basic facts and figures, in this case the intrinsic noise present and the final level of noise you would say was acceptable and I'll be honest, I've no idea what the actual pk to pk noise from the zeners you might use would be. It will be small though and wideband in form. So knowing that in general we could reasonably say that something like a 10 ohm and 100uf with a say the 100uf shunted by a say a 0.1 film in series with a 1 ohm should reduce the noise to essentially zero.
- Status
- Not open for further replies.