Greetings,
I recently built the LM3886 kit from Chipamp.com, and I blow the mains fuse every time I turn it on. I'm using a single power supply (Chipamp.com's snubberized PS, see diagram). I'm using a 300 VA torroid with dual 18V secondaries (Antek AN3218, see other file). I've tried 3 and 4 amp slow blo, and 3 amp medium blo fuses. Amp boards are not hooked up yet. If the rectifier & filter caps are not attached, I measure ~18VAC from both secondaries - no blown fuses - so I'm pretty confident I've got the transformer wired correctly.
However, when I have the rectifier and filter caps attached and press the power switch:
So my problem is that I don't know whether I'm using too low rated fuses, or if there's a short in my board. I'm afraid to simply try higher rated fuses in case it's the latter. My multimeter tells me I don't have a short. Current flows between the inputs for each pair of windings very briefly - I guess it's charging up the filtering caps (pair of 10,000uf 50V). The fact that the indicator LED dims slowly after the fuse blows also suggests the filter caps were charging up.
So . . . Do I try higher rated fuses? If they keep blowing, when do I stop? Is there another way to test/troubleshoot the board?
Any advice is greatly appreciated!
I recently built the LM3886 kit from Chipamp.com, and I blow the mains fuse every time I turn it on. I'm using a single power supply (Chipamp.com's snubberized PS, see diagram). I'm using a 300 VA torroid with dual 18V secondaries (Antek AN3218, see other file). I've tried 3 and 4 amp slow blo, and 3 amp medium blo fuses. Amp boards are not hooked up yet. If the rectifier & filter caps are not attached, I measure ~18VAC from both secondaries - no blown fuses - so I'm pretty confident I've got the transformer wired correctly.
However, when I have the rectifier and filter caps attached and press the power switch:
- The on-board LED lights up very briefly
- one of the mains fuses (there are 2, one for live and one for ground) blows - usually live.
- the on-board LED slowly dims.
- Nothing on the board appears fried, or feels hot.
So my problem is that I don't know whether I'm using too low rated fuses, or if there's a short in my board. I'm afraid to simply try higher rated fuses in case it's the latter. My multimeter tells me I don't have a short. Current flows between the inputs for each pair of windings very briefly - I guess it's charging up the filtering caps (pair of 10,000uf 50V). The fact that the indicator LED dims slowly after the fuse blows also suggests the filter caps were charging up.
So . . . Do I try higher rated fuses? If they keep blowing, when do I stop? Is there another way to test/troubleshoot the board?
Any advice is greatly appreciated!
Attachments
National has an app note for an inrush circuit to limit the high current that occurs at power up.
AN-1849
http://www.national.com/an/AN/AN-1849.pdf
AN-1849
http://www.national.com/an/AN/AN-1849.pdf
18VAC from an unloaded 18VAC secondary? That doesn't sound right. Set aside your confidence and make sure the transformer is wired correctly. Same for the rectifiers and filter caps.
The simple power calculation of 4 amps at 120VAC is 480 watts. Something is obviously wrong with your circuit. Nothing feels hot because power is dissipated by the fuse - in short order. Don't use a larger fuse, use a light bulb in series with the hot mains for current limiting. Do that only after you've thoroughly checked the circuit. Then checked it again.
The simple power calculation of 4 amps at 120VAC is 480 watts. Something is obviously wrong with your circuit. Nothing feels hot because power is dissipated by the fuse - in short order. Don't use a larger fuse, use a light bulb in series with the hot mains for current limiting. Do that only after you've thoroughly checked the circuit. Then checked it again.
Id hook up a 150 watt light bulb in series. If it stays lit while the ps is on ( i assume the amps arent hooked up) there is something wrong with the ps.
With 300va and 20,000uf of caps, odds are the surge current is too high on turn on and you will need to install a soft start.
For a soft start id use this relay on the power supply outputs:
T7NS5D1-48 Tyco Electronics / P&B General Purpose / Industrial Relays
to bypass a CL-80 or CL-90 on the primary side
CL-80 GE Sensing / Thermometrics Thermistors
Your whole soft start would be a relay, a diode for the coil, and a ntc.
With 300va and 20,000uf of caps, odds are the surge current is too high on turn on and you will need to install a soft start.
For a soft start id use this relay on the power supply outputs:
T7NS5D1-48 Tyco Electronics / P&B General Purpose / Industrial Relays
to bypass a CL-80 or CL-90 on the primary side
CL-80 GE Sensing / Thermometrics Thermistors
Your whole soft start would be a relay, a diode for the coil, and a ntc.
Thank's everyone for the suggestions.
Thanks All,
I read a little article on the "light bulb" tester, but was hoping to avoid building one.
(lazy). I also know that things such as inrush current limiters exist, but I was hoping to avoid building one of those too.
I'm going to go up to a 5amp fuse, and if that blows, I'll build a light-bulb tester.
SofaSpud: I don't understand what you think is wrong. I measured 18VAC from the secondaries, which are supposed to be 18VAC. This is without the rectifier attached.
-Byron
Thanks All,
I read a little article on the "light bulb" tester, but was hoping to avoid building one.
SofaSpud: I don't understand what you think is wrong. I measured 18VAC from the secondaries, which are supposed to be 18VAC. This is without the rectifier attached.
-Byron
Yeah, I'm very probably wrong to misread your measurements. But you can with reasonable care use a lamp that's handy and some gator clip test leads to do the light bulb test. No need to sidetrack into another project.
You can also try swapping the 10000uF caps with smaller values to get the time constant within the fuse's slo-blo time. This will at least tell you if you have a working circuit.
You can also try swapping the 10000uF caps with smaller values to get the time constant within the fuse's slo-blo time. This will at least tell you if you have a working circuit.
SUCCESS!
Who would have thought that the back surface of the MUR860 diodes are not insulated. I put tiny heat sinks on all the bridge diodes (MUR860). I figured it couldn't hurt. Well, the heat sinks and screws were touching each other. I removed them and . . .
! I hooked everything up, and I'm listening to Yello's album Zebra right now. I put the 3 amp medium blo fuses in, as that's all I have left.
Obviously, this is my first amplifier. It's surprisingly quiet bolted to a wooden board with no ground or shielding. I think it's got more punchy bass and detail than the Sansui receiver I was using. Of course I'm biased. Wait until I put it in a proper chassis with shielding and grounding! I'll post photo's in the gallery forum when it's all done.
Cheers!
Byron
Who would have thought that the back surface of the MUR860 diodes are not insulated. I put tiny heat sinks on all the bridge diodes (MUR860). I figured it couldn't hurt. Well, the heat sinks and screws were touching each other. I removed them and . . .
! I hooked everything up, and I'm listening to Yello's album Zebra right now. I put the 3 amp medium blo fuses in, as that's all I have left.Obviously, this is my first amplifier. It's surprisingly quiet bolted to a wooden board with no ground or shielding. I think it's got more punchy bass and detail than the Sansui receiver I was using. Of course I'm biased. Wait until I put it in a proper chassis with shielding and grounding! I'll post photo's in the gallery forum when it's all done.
Cheers!
Byron
Cent88, does it blow the fuse while playing, or at turn on? The fuse pop is normaly because you are using a toroid combined with a nice size cap bank, it might but sucking 50 amps or more at turn on.
You have a 225va transformer. that means its rated at 225watts, so you are looking at 2 amps tops. Why a slow blow 2a makes it sound soft sounds impossible considering its just a short.
You have a 225va transformer. that means its rated at 225watts, so you are looking at 2 amps tops. Why a slow blow 2a makes it sound soft sounds impossible considering its just a short.
A 225VA transformer at 115V should draw about 2A. It is usual to specify a fuse at 3 X operating current.
Try 5A Slow Blow. This will protect the transformer but wont protect the delicate electronics beyond that. You may need to fuse the supplies on the mains input at 5A as suggested and then after the caps to protect the amps. The charging current to the caps can be a fuse killer. If you specify 25A diodes in the rectifiers which are cheap and common then the fuses should prevent disastor. In the end, a faulty PSU component is going to cause some damage. If the PCB or wires are meaty enough then the damage will be limited to the faulty component. If a cap fails with 25A diodes and 5A mains fuse, the diode will normally survive long enough to blow the fuse.
Try 5A Slow Blow. This will protect the transformer but wont protect the delicate electronics beyond that. You may need to fuse the supplies on the mains input at 5A as suggested and then after the caps to protect the amps. The charging current to the caps can be a fuse killer. If you specify 25A diodes in the rectifiers which are cheap and common then the fuses should prevent disastor. In the end, a faulty PSU component is going to cause some damage. If the PCB or wires are meaty enough then the damage will be limited to the faulty component. If a cap fails with 25A diodes and 5A mains fuse, the diode will normally survive long enough to blow the fuse.
Fuse selection is more complicated than that.
UL recommends 25% derating at 25C.
I would not go more than 50% over rating of transformer primary wiring...
I hate to nitpick further
, but.....note that the PEAK DC voltage is 1.4x. If you consume the full DC design current the AVERAGE DC voltage will be close to 1x.Current also does not 'go down to 0.62x'. The transformer needs to be sized for secondary amps (RMS) of about 1.8x (Plitron) or 1/0.62x (Hammond?) maximum expected continuous DC amps to compensate for increased I^2R heating. Numbers are manufacturer's de-rating, valid for bridge rectifier use only. Typically the transformer will happily push more current, but this de-rating will prevent smoking the transformer.
Eh ill just put a reactor in my amp to correct the power factor
Still alittle on subject, are you saying that a 100va 50vac 2amp unit, rectified and capped, is going to go from 70vdc at idle, down to 50vdc when drawing 2 amps? Or are you saying that even though i rectified the vac, and the load is purely resistive, the real power going into the resistor is going to be 2 amps at 50 volt?
There is no free lunch, with just a resistor and no rectifier, its real power will be 50 volts at 2 amps. If you rectify and cap you just moved your voltage up to 70, you cant also pull the same 2 amps.
Still alittle on subject, are you saying that a 100va 50vac 2amp unit, rectified and capped, is going to go from 70vdc at idle, down to 50vdc when drawing 2 amps? Or are you saying that even though i rectified the vac, and the load is purely resistive, the real power going into the resistor is going to be 2 amps at 50 volt?
There is no free lunch, with just a resistor and no rectifier, its real power will be 50 volts at 2 amps. If you rectify and cap you just moved your voltage up to 70, you cant also pull the same 2 amps.
Pretty much this, except you cannot draw 2A DC continuously without overheating the transformer. It is the ^2 in the I^2R that you have to consider. If the capacitor charge pulse is 4x the RMS current, the I^2R heat generated during the pulse is 16x. The duration is short, but it still makes a big difference.
It can get complicated, depending on the application.
-In general, fuse should be at rating of equipment (e.g. primary rating of transformer).
-In general, fuse should be de-rated 25% (or more, if above 25C ambient, using maximum expected continuous current as a baseline) to prevent nuisance blowing.
-An application with in-rush current needs a T-type fuse. T-type fuses do not need to be de-rated.
Electronic Fuse Selection Guide and Definitions.
For audio amp requirements, Rod's site is always a good reference:
Elliott Sound Products - Linear Power Supply Design
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.