I'll agree if the problem is blown fuses but it's also good to have a softstarter if you want better protection using smaller fuses.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=74022&highlight=
http://www.diyaudio.com/forums/showthread.php?s=&threadid=74022&highlight=
softstart for poweramp?
I have done some simulatins using Tina Spice and YES, you do need softstart.
Dependent whether you switch on during zerocrossing of the mains or at the peak, the inrush currents can momentarely go up from 100 to 400Amps.
A good starting point therefore is to make sure your softstart circuit switches mains on as close to the zerocrossing as possible.
The circuit I simulated used a 500VA torroid with 2x55V sec.
and 2x 10000u caps.
By switching mains on at zerocrossing and a series resistor of 22Ohms for 0.5seconds made sure that the mains inrushcurrent
was limited to 15A peak at the first 50Hz cycle decreasing to nominal load in 7 to 8 cycles.
Reason for these simulations is that I decided to build a softstart circuit using zerocrossing softstart switch-on.
I am using TRIACs to switch the 22Ohms resistor, that gets bypassed after 0.5seconds with a second TRIAC.
I also have an audio sense circuit foreseen, where the system switches ON and OFF after 5 minutes without audio.
In this way I can remotely switch on my active speaker systems, directly from my pre-amp.
The one thing I noticed is that at full load simulation (10A continuous load) there are momentary current peaks to charge the 10000uF caps.
These are peaks of 30A with for some 1-2mS.
Such short current peaks are no problem for most TRIACs but I was wondering if in such cases you have to use Power Factor Correction as the load current does not have a sinus shape.
I added the zerocrossing and peak switch on currentwaveforms FYI.
I have done some simulatins using Tina Spice and YES, you do need softstart.
Dependent whether you switch on during zerocrossing of the mains or at the peak, the inrush currents can momentarely go up from 100 to 400Amps.
A good starting point therefore is to make sure your softstart circuit switches mains on as close to the zerocrossing as possible.
The circuit I simulated used a 500VA torroid with 2x55V sec.
and 2x 10000u caps.
By switching mains on at zerocrossing and a series resistor of 22Ohms for 0.5seconds made sure that the mains inrushcurrent
was limited to 15A peak at the first 50Hz cycle decreasing to nominal load in 7 to 8 cycles.
Reason for these simulations is that I decided to build a softstart circuit using zerocrossing softstart switch-on.
I am using TRIACs to switch the 22Ohms resistor, that gets bypassed after 0.5seconds with a second TRIAC.
I also have an audio sense circuit foreseen, where the system switches ON and OFF after 5 minutes without audio.
In this way I can remotely switch on my active speaker systems, directly from my pre-amp.
The one thing I noticed is that at full load simulation (10A continuous load) there are momentary current peaks to charge the 10000uF caps.
These are peaks of 30A with for some 1-2mS.
Such short current peaks are no problem for most TRIACs but I was wondering if in such cases you have to use Power Factor Correction as the load current does not have a sinus shape.
I added the zerocrossing and peak switch on currentwaveforms FYI.
Attachments
Well, if you all don't mind, I have a soft start related question -
I have been building UCD/Tripath/gainclone amps without a soft start for a while - I figured as long as I don't dim the light or blow fuses I was fine. The amps all vary from 150VA 50VCT Gainclones to dual 400VA 80VCT UCD400. All without soft start. Even the Hypex website indicated they may not be necessary...
Anyway, I am building a new UCD400 and it will be 750VA 40VAC Dual secondaries and worry about hurting amps or speakers in an "event"- there are dual mains fuses to single monster toroid, and each +V/-V rails will have fuses also (6 fuses total).
Anyway, is there a external soft start I can buy to plug between the socket and the amp so I don't have to stuff an internal soft start into the amp chassis?
Is there such an animal? Would be perfect if there was a combo soft start/surge protector/line filter available that I could buy (cheaply)...
I have been building UCD/Tripath/gainclone amps without a soft start for a while - I figured as long as I don't dim the light or blow fuses I was fine. The amps all vary from 150VA 50VCT Gainclones to dual 400VA 80VCT UCD400. All without soft start. Even the Hypex website indicated they may not be necessary...
Anyway, I am building a new UCD400 and it will be 750VA 40VAC Dual secondaries and worry about hurting amps or speakers in an "event"- there are dual mains fuses to single monster toroid, and each +V/-V rails will have fuses also (6 fuses total).
Anyway, is there a external soft start I can buy to plug between the socket and the amp so I don't have to stuff an internal soft start into the amp chassis?
Is there such an animal? Would be perfect if there was a combo soft start/surge protector/line filter available that I could buy (cheaply)...
I think slower than I type...ebay ebay ebay
http://cgi.ebay.com/SOFT-START-SIX-...3QQihZ004QQcategoryZ73369QQrdZ1QQcmdZViewItem
http://cgi.ebay.com/SOFT-START-SIX-...3QQihZ004QQcategoryZ73369QQrdZ1QQcmdZViewItem
Looks like the only option is internal - hows this for a Soft Start?
http://cgi.ebay.com/Power-Soft-Star...6534752QQihZ015QQcategoryZ14980QQcmdZViewItem
http://cgi.ebay.com/Power-Soft-Star...6534752QQihZ015QQcategoryZ14980QQcmdZViewItem
not enough info. doesn't specify the current it can handle. 5-8 seconds seems like a VERY long time to be in circuit, especially for a power amp. could burn out resistor if not adequate, unless it is a high value resistor to really drop the current, and therefore require a really long time constant.
Hi Max,
what is your circuit connected to?
if the connections on the left come from the soft started transformer then they are going to sit low and delay the time constant.
Why is c1=10mF, this should work with less than 100uF.
Why is the RC of C2*R1=22seconds? how will this give 1 second delay? 200mS to 300mS is OK for a soft start delay.
When the voltage does come up then a higher voltage is going to appear on C2 & R4. will the high voltage be OK?
are the diodes on the right connected to a pair of relays?
Why two diodes?
What are the two relays doing? One for soft start presumably, and the other?
what is your circuit connected to?
if the connections on the left come from the soft started transformer then they are going to sit low and delay the time constant.
Why is c1=10mF, this should work with less than 100uF.
Why is the RC of C2*R1=22seconds? how will this give 1 second delay? 200mS to 300mS is OK for a soft start delay.
When the voltage does come up then a higher voltage is going to appear on C2 & R4. will the high voltage be OK?
are the diodes on the right connected to a pair of relays?
Why two diodes?
What are the two relays doing? One for soft start presumably, and the other?
Hi AndrewT,
my circuit is connected to another transfo 9-0-9Vac and the 10mF cap is for dc filter.
The delay is 1 second because the cap must be charge at 0.6V for the transistor Q1 turn on, not 12V.
Yes, you are right I could have put one diode. I have 2 relay one by side because i have 2 transfo. Thank You! Maxpou
my circuit is connected to another transfo 9-0-9Vac and the 10mF cap is for dc filter.
The delay is 1 second because the cap must be charge at 0.6V for the transistor Q1 turn on, not 12V.
Yes, you are right I could have put one diode. I have 2 relay one by side because i have 2 transfo. Thank You! Maxpou
I have just finished a PCB for a soft start and switch on circuit. The soft start uses 4 x 10R/5W resistors in series to limit the inrush current to about 5.5A, which are then taken out of the circuit by a relay which operates after about a half second or so. The other part of the circuit is a remote switch on, allowing any switch to be used as the power switch (I am using the anodised aluminum push button - TH Contact). There are also 2 LED connections - standby and power on, with trimpots to allow adjusting the LED intensity. Board size is about Board size is 168x97mm. Just tested and works beautifully.
I've been using the soft start circuit alone in my power amps for about 8 years now with no problems to report.
Pic of completed PCB attached. I can make the boards available (bare), with components or completed if anyone is interested. Sorry, but PCBs have been designed for 220-240V use only.
I've been using the soft start circuit alone in my power amps for about 8 years now with no problems to report.
Pic of completed PCB attached. I can make the boards available (bare), with components or completed if anyone is interested. Sorry, but PCBs have been designed for 220-240V use only.
Attachments
softstart for poweramp?
John,
Maybe you can check if Hypex is working on SMPS for their UcD's.
What is the kick of an efficient amp if the powersupply is bulky?
In that case I asume the PFC, hence inrush limiting will be included.
Alternatively, Have a look at Per Anders' website.
This is by the looks of it a relay based limiter.
Pro: its simple
con: high probability it switches at the peak of the mains sinewave, causing a massive bang in the torroid.
You can hear it buzzing for several seconds if that happens.
Last option: be patient, I am working on a microcontroller controlled TRIAC version that can switch up to 3 transformers, switches on at the zerocrossing with a delay between each transformer.
Sounds like overkill but if you spend so much on your torroids, high power rectifiers and high voltage caps, 10 bucks for a protection isn't gonna bankrupt you.
John,
Maybe you can check if Hypex is working on SMPS for their UcD's.
What is the kick of an efficient amp if the powersupply is bulky?
In that case I asume the PFC, hence inrush limiting will be included.
Alternatively, Have a look at Per Anders' website.
This is by the looks of it a relay based limiter.
Pro: its simple
con: high probability it switches at the peak of the mains sinewave, causing a massive bang in the torroid.
You can hear it buzzing for several seconds if that happens.
Last option: be patient, I am working on a microcontroller controlled TRIAC version that can switch up to 3 transformers, switches on at the zerocrossing with a delay between each transformer.
Sounds like overkill but if you spend so much on your torroids, high power rectifiers and high voltage caps, 10 bucks for a protection isn't gonna bankrupt you.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.