I'm building ESP's project 36 and capactive multiplier power supply. I've completed the power supply but am still waiting for my current inrush limiter boards from ESP.
I'm still struggling to understand what the inrush current limiter is protecting. Is it protecting my diodes? What risks am I running wiring up the supply and measuring it's output voltage without protection?
I'm using a 400VA toroid with each secondary feeding a gbpc3501 35A rectifier.
I'm still struggling to understand what the inrush current limiter is protecting. Is it protecting my diodes? What risks am I running wiring up the supply and measuring it's output voltage without protection?
I'm using a 400VA toroid with each secondary feeding a gbpc3501 35A rectifier.
You need to be patient and wait for the current limiter.
A toroidal transformer sucks a lot of amps in the first split second its magnetic is energized. Plus, power supply caps look like a dead short until they are charged. A lot can happen in the first 1-2 seconds when you turn on a power supply.
Some find that a thermistor in series with the incoming AC is sufficient, something like a CL-30. Others, like me, use a time delay relay and a 50 watt 50 ohm resistor to allow 5-7 seconds for the power supply to "soft start." After the initial few seconds the relay closes and removes the power resistor from the circuit and puts the AC directly to the amp.
Of course, some find their diode bridge and house branch electrical circuit husky enough to withstand initial turn-on abuse. Most don't, especially if your cap bank in the amp is of any good size.
A toroidal transformer sucks a lot of amps in the first split second its magnetic is energized. Plus, power supply caps look like a dead short until they are charged. A lot can happen in the first 1-2 seconds when you turn on a power supply.
Some find that a thermistor in series with the incoming AC is sufficient, something like a CL-30. Others, like me, use a time delay relay and a 50 watt 50 ohm resistor to allow 5-7 seconds for the power supply to "soft start." After the initial few seconds the relay closes and removes the power resistor from the circuit and puts the AC directly to the amp.
Of course, some find their diode bridge and house branch electrical circuit husky enough to withstand initial turn-on abuse. Most don't, especially if your cap bank in the amp is of any good size.
Darn, more waiting. Your explanation makes sense.
My supply has a pair of 4700uF 63V caps before the capacitance multiplier.
Where's the cutoff point when deciding whether or not to use an inrush limiter?
My supply has a pair of 4700uF 63V caps before the capacitance multiplier.
Where's the cutoff point when deciding whether or not to use an inrush limiter?
It's been some time since I last looked at the ESP cap mult. circuit, but most cap multipliers use a husky pass transistor with its base tied to a fairly large capacitor -- the one that get "multiplied." Before and after the pass transistor can also be sets of fairly large capacitors. One must consider the circuit as a whole. You don't want to fry the pass transistor or the diode bidge or the electrical branch circuit in your home. Without a circuit diagram and component values it is difficult to be more precise, but if Rod Elliot says to use an inrush current limiter then I would do so -- or make a circuit similar to his.
Are you really impatient? Then wire a 100W light bulb in series with one leg of the mains feed. It will glow nicely at power on, then fade out as the caps charge up. And it will severely limit the inrush current to a safe value.
If the light bulb stays lit, power down immediately and figure out what you wired wrong.
If the light bulb stays lit, power down immediately and figure out what you wired wrong.
Yes, I am really that impatient. I'm just getting excited to finish this project. This has been a very challenging project putting all the pieces together.
I must confess, I actually tested one of the power supplies before posting my question but then decided to not to proceed further. I was delighted that it worked.
Actually, the inrush limiter was recommended by someone on Rod's board due to the size of the transformer I'm using. I thought I had seen Rod mention it but going back through the project 36 articles, I don't see any mention of it. Given that his softstart circuit is project 39, that may be why.
In the project 36 construction article Rod recommended a 300VA 30-0-30 transformer. I already had a 400VA 25-0-25 without a home so I decided to use it. All other component values are identical to what Rod published. Here's the schematic I'm working from:
I must confess, I actually tested one of the power supplies before posting my question but then decided to not to proceed further. I was delighted that it worked.
Actually, the inrush limiter was recommended by someone on Rod's board due to the size of the transformer I'm using. I thought I had seen Rod mention it but going back through the project 36 articles, I don't see any mention of it. Given that his softstart circuit is project 39, that may be why.
In the project 36 construction article Rod recommended a 300VA 30-0-30 transformer. I already had a 400VA 25-0-25 without a home so I decided to use it. All other component values are identical to what Rod published. Here's the schematic I'm working from:
An externally hosted image should be here but it was not working when we last tested it.
The pass transistors will not begin to conduct until the big cap attached to their base charges so you have a sorta' soft start already available. The initial 4700 uF caps are not that large so you should be able to just turn on the PS without a problem. The part that might create a problem is the big toroid transformer sucking a huge amount of initial current until its magnetic field is established. Do some research in the search engines here. There is a complete thread (or two) devoted to the use of an inrush current limiter and one message has the URL for a table of values to use to select which thermistor to use. The CL-30 is a good safe choice for you, I believe. Many amps that use toroidal transformers use this CL-30 inrush current limiter.
In the meantime use the light bulb hook up suggested by SY and you will be safe.
In the meantime use the light bulb hook up suggested by SY and you will be safe.
I don't want to hijack ultrachrome's thread, but it seems pointless to start a new one since I have the same question to ask - only my components are different.
I'm making dual single ended power supplies, one for each t-amp, with the following components:
(1) 120VA toroidal transformer with dual 15VAC secondary
(2) 25A bridge rectifier
(2) 22000uF filter cap
The supplies for the t-amps only need to be single ended, so each secondary will be connected to a bridge rectifier and a filter cap. The unregulated DC will be fed to a voltage regulator (LT1083) closer to the amp PCB for good voltage regulation.
While the VA rating of the transformer is quite below the 300 VA or so rating that I've read would require a soft-start circuit, the filter caps are quite large. Do you guys think this supply would require a soft-start circuit or will the inrush current be kept to an acceptable level?
back to the original topic - ultrachrome, did you ever get/install the soft-start circuit?
edit: I forgot to mention that there will also be a 6A rated EMI filter before the transformer, if that makes any difference. I know that they filter out the EMI going back into the line but I don't know if they would have any impact on inrush current.
I'm making dual single ended power supplies, one for each t-amp, with the following components:
(1) 120VA toroidal transformer with dual 15VAC secondary
(2) 25A bridge rectifier
(2) 22000uF filter cap
The supplies for the t-amps only need to be single ended, so each secondary will be connected to a bridge rectifier and a filter cap. The unregulated DC will be fed to a voltage regulator (LT1083) closer to the amp PCB for good voltage regulation.
While the VA rating of the transformer is quite below the 300 VA or so rating that I've read would require a soft-start circuit, the filter caps are quite large. Do you guys think this supply would require a soft-start circuit or will the inrush current be kept to an acceptable level?
back to the original topic - ultrachrome, did you ever get/install the soft-start circuit?
edit: I forgot to mention that there will also be a 6A rated EMI filter before the transformer, if that makes any difference. I know that they filter out the EMI going back into the line but I don't know if they would have any impact on inrush current.
I received the boards for ESP's inrush current limiter yesterday but I'm actually going to start testing without it this week.
I went back and reread the project and construction articles for the project and Rod does not mention inrush current protection anywhere. Additonally a recent thread suggested I wouldn't need protection unless my capacitance was dramatically higher than what it currently is.
I went back and reread the project and construction articles for the project and Rod does not mention inrush current protection anywhere. Additonally a recent thread suggested I wouldn't need protection unless my capacitance was dramatically higher than what it currently is.
The main purpose is not to blow the fuses in the wall.ultrachrome said:
The second reason is to prolong the lifetime of the mains switch.
That's what I am now understanding.
It sounds like I'll be okay in this case but I think I'll add a thermistor for the long term.
It sounds like I'll be okay in this case but I think I'll add a thermistor for the long term.
There have been several threads floating around lately about thermistors, inrush limiting, time delay relays and soft start circuits.
Thermistors are best used with a class A amp that has a steady current demand. For an AB amp a thermistor is not the best solution. This is where a time delay relay circuit would be best.
Another consideration is how the current limiting circuit gets reset and how quickly. If there is a series of intermittant electrical service interrutions a thermistor that is still hot will conduct a lot of current and not provide any current inrush limiting. The PS caps will still be mostly charged which is good but the toroid transformer will have to rebuild its magnetic field and will suck a lot of current, which is bad without current limiting.
The best solution I have found is to use a time delay relay that runs from the incoming AC and shorts across a big power surge resistor in a matter of several seconds. I use the one below. It can be set for a delay over a wide range depending on the value of a small attached external resistor. I use a 50 ohm 50 watt resistor in series with incoming AC and the relay can handle up to 20 amps at its switch contacts. All of this is set for around 7-8 seconds and gives me a nice gentle start to an amp with even a huge amount of PS capacitance. Plus, it will reset almost immediately in case of power interruption and begin the initial 7-8 second delay period all over again thus protecting the amp from sudden power interruptions and power restarts.
Thermistors are best used with a class A amp that has a steady current demand. For an AB amp a thermistor is not the best solution. This is where a time delay relay circuit would be best.
Another consideration is how the current limiting circuit gets reset and how quickly. If there is a series of intermittant electrical service interrutions a thermistor that is still hot will conduct a lot of current and not provide any current inrush limiting. The PS caps will still be mostly charged which is good but the toroid transformer will have to rebuild its magnetic field and will suck a lot of current, which is bad without current limiting.
The best solution I have found is to use a time delay relay that runs from the incoming AC and shorts across a big power surge resistor in a matter of several seconds. I use the one below. It can be set for a delay over a wide range depending on the value of a small attached external resistor. I use a 50 ohm 50 watt resistor in series with incoming AC and the relay can handle up to 20 amps at its switch contacts. All of this is set for around 7-8 seconds and gives me a nice gentle start to an amp with even a huge amount of PS capacitance. Plus, it will reset almost immediately in case of power interruption and begin the initial 7-8 second delay period all over again thus protecting the amp from sudden power interruptions and power restarts.
Attachments
Maybe I should make a WIKI page and link all this stuff on there. As always seems the case, this flurry of interest in the subject will die down shortly and a couple months from now someone will ask the same questions.
That looks like a simple straightforward solution. The only downside is that those relays seem to cost about $100. I searched digikey for "relay 120VAC time" and clicked on "Undefined Family."
What about the solution mentioned in the other thread by Mrupp:
That looks like a simple straightforward solution. The only downside is that those relays seem to cost about $100. I searched digikey for "relay 120VAC time" and clicked on "Undefined Family."
What about the solution mentioned in the other thread by Mrupp:
Check out these relays:
http://search.ebay.com/time-delay-r...5AIPQ2FPostalQQftrtZ1QQftrvZ1QQfsopZ1QQfsooZ1
There must be something there useful and for a lot less than $100!
http://search.ebay.com/time-delay-r...5AIPQ2FPostalQQftrtZ1QQftrvZ1QQfsopZ1QQfsooZ1
There must be something there useful and for a lot less than $100!
Re: amp delay relay circuit question Quote: Originally Posted by oaktsh Hi Dick I am interested in implementing your time delay relay using the same delay timer you used in your old post. [url said:
Dick, Thanks for the information.
Can you describe your experience regarding how you decide on the proper delay time. In reading other posts, I have seen delay time set ranging in the 100ms to 3-6 second delay time. My transformer is rated at 1.3kva with cap bank of 132000uF.
Can you also describe how you implement your fault indicator using the your AC time delay relay? Also did you have to build a transient suppressor circuit using Resistor+Cap in parallel with your coil?
Thanks
If I were a trained EE I probably could crank out some equations to determine the best time span to allow those really big caps of yours to charge. I found that 7 seconds was enough for me to charge up around 130K of uF -- without the lights dimming very much; and with 50 ohms resistance in the AC between the relay and the amp/s. Also, you have to consider the stress placed on the diode bridge as the large caps are charging. Is your transformer an old type IE or a toroid? But, it is generally agreed that any amount of capacitance in a power supply beyond ~50k uF a suppression circuit of some kind is required so as not to fry the ON/OFF switch or blow out the diode bridge.
If you look at my schematic in the previous messages you can see how I connected a small AC bulb to light up while the delay contacts were NOT tripped and that when the circuit does trip the bulb would be disengaged. I have my relay and other parts of the delay circuit mounted up in a metal case (from an old AT computer power supply). So, I watch what happens when I flip its master switch to engage the relay circuit and if the AC bulb does not go off as the delay time expires I know the relay has "stuck" or is not functioning properly. If it were to do so the surge resistor would still be in the AC circuit and it would continue to get hotter and hotter.
I suppose I could mount a thermal switch close to this resistor that would cut off the AC if it got too hot. Check the schematic for a Hafler DH-200 on how these thermal switches are used.
I never bothered with a transient suppression circuit as I never sensed the relay contacts were chattering.
Well, whatta' hobby. Right?
If you look at my schematic in the previous messages you can see how I connected a small AC bulb to light up while the delay contacts were NOT tripped and that when the circuit does trip the bulb would be disengaged. I have my relay and other parts of the delay circuit mounted up in a metal case (from an old AT computer power supply). So, I watch what happens when I flip its master switch to engage the relay circuit and if the AC bulb does not go off as the delay time expires I know the relay has "stuck" or is not functioning properly. If it were to do so the surge resistor would still be in the AC circuit and it would continue to get hotter and hotter.
I suppose I could mount a thermal switch close to this resistor that would cut off the AC if it got too hot. Check the schematic for a Hafler DH-200 on how these thermal switches are used.
I never bothered with a transient suppression circuit as I never sensed the relay contacts were chattering.
Well, whatta' hobby. Right?
- Status
- Not open for further replies.