• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

switching power supply help needed

gadget73

gadget73

junk junkie
Subscriber
Do any of you guys understand switching power supplies? I've got a B&K 1692 40 amp supply that died on me. I opened it up and found 2 puffy caps, and the 4 large ones for the rectified AC line current also looked a little swollen so I replaced those too. It still does absolutely nothing. I can confirm I have about 330 volts DC after the bridge rectifier, but it seems I'm getting nothing into the power transformer. The digital display and the fan also do not operate. I'm totally at a loss on this one. I can't find a schematic, and I don't understand these nearly enough to even know where to go next. Help?
 
Register to hide this ad
gadget73 said:
Do any of you guys understand switching power supplies? I've got a B&K 1692 40 amp supply that died on me. I opened it up and found 2 puffy caps, and the 4 large ones for the rectified AC line current also looked a little swollen so I replaced those too. It still does absolutely nothing. I can confirm I have about 330 volts DC after the bridge rectifier, but it seems I'm getting nothing into the power transformer. The digital display and the fan also do not operate. I'm totally at a loss on this one. I can't find a schematic, and I don't understand these nearly enough to even know where to go next. Help?
Click to expand...

I don't know that supply in particular, but I think chances are that you lost the caps first and they brought down the main switching device(s). These would be attached to a heatsink, either a chassis hs or an individual device hs. Most often, these are TO-220 packages, although T0-3, 126, 247, 202, etc are possible. I would check this next.

-k
 
I pulled the main switcher mosfets out and checked them according to this:

http://www.4qdtec.com/mostest.html

They don't show as dead or shorted, so I guess they're OK. They don't appear to be getting any drive to the gate according to the scope. Unfortunately the ground lead popped off and grounded against something it ought not have and the fuse blew on me. I don't have another of that size and type, so I guess I'm done troubleshooting it for a bit.

I guess I'll have to get my old 25 amp non-switcher back from my friend and run that for a while. I really would like to fix this, its a nice power supply.
 
Its not all surface mount stuff if thats what you mean. Just standard DIP and through-hole components on this one. Its actually got 3 DIP16 type chips in there. One is an LM324N, one is TL494CN, the other is TC4069UBP it looks like. I'm guessing its one of the latter two thats the controller but I'm not sure which. The TL494CN is described as a "pulse width modulation controller", which sounds like a switching controller to me. The TC4069UBP is a hex inverter, whatever the heck that is. The LM324N is a quad op-amp.

Looking at the board, it looks like the TL494CN wires to the small transformers ahead of the large output MOSFETs, so I'm assuming thats the switch controller ?
 
A hex inverter is simply six (hex) independent inverters in a chip. It has fourteen pins. Six inputs drive six respective outputs. That makes twelve pins. The other two pins are power and ground.
 
gadget73 said:
Its not all surface mount stuff if thats what you mean. Just standard DIP and through-hole components on this one. Its actually got 3 DIP16 type chips in there. One is an LM324N, one is TL494CN, the other is TC4069UBP it looks like. I'm guessing its one of the latter two thats the controller but I'm not sure which. The TL494CN is described as a "pulse width modulation controller", which sounds like a switching controller to me. The TC4069UBP is a hex inverter, whatever the heck that is. The LM324N is a quad op-amp.

Looking at the board, it looks like the TL494CN wires to the small transformers ahead of the large output MOSFETs, so I'm assuming thats the switch controller ?
Click to expand...

Yes, the TL494CN PWM controller is the heart of the switching supply and drives the switching transistors, usually then into a toroidal DC-DC switching transformer. Here's the datasheet:

http://www.datasheetcatalog.org/datasheet2/a/0a91owp2owcksesx812wr26kdjyy.pdf

See circuit on page 9 for typical Push-pull power converter supply:
http://www.datasheetarchive.com/dl/Datasheets-21/DSA-409130.pdf

For your specific model, here's the B&K website page:
http://www.bkprecision.com/products...dc-40a-switching-digital-dc-power-supply.html

And their documentation:
http://www.bkprecision.com/downloads/datasheets/1692_datasheet.pdf

http://www.bkprecision.com/downloads/manuals/en/1692_manual.pdf

You may need to contact them for a schematic and parts list or service manual.
 
In my entire television engineering career, I've never seen anyone fix a switching or switch mode power supply of any type. Cut your losses and just replace it, would be my advice. Maybe you could return it to the manufacturer for repair?

Dave
 
I think since it doesn't use a proprietary controller, it could be fixed without too much trouble unless there are some other components that are broken and irreplaceable. A schematic would definitely help though.
 
gadget73 said:
I pulled the main switcher mosfets out and checked them according to this:

http://www.4qdtec.com/mostest.html

They don't show as dead or shorted, so I guess they're OK. They don't appear to be getting any drive to the gate according to the scope. Unfortunately the ground lead popped off and grounded against something it ought not have and the fuse blew on me. I don't have another of that size and type, so I guess I'm done troubleshooting it for a bit.

I guess I'll have to get my old 25 amp non-switcher back from my friend and run that for a while. I really would like to fix this, its a nice power supply.
Click to expand...

FWIW- I am not convinced by those test procedures. Time to get the schematic.

-k
 
gadget73 said:
...I can confirm I have about 330 volts DC after the bridge rectifier, but it seems I'm getting nothing into the power transformer.

On most switchers they take the AC input and run it into a bridge rectifier. Assuming the input voltage is 120 VAC, the rectified DC should be in the 170V range. The DC voltage then feeds into the center tap of the xfmr.

Unless the rectifier is a voltage doubler, the 330VDC seems to be to high.
Click to expand...
 
mrdbdigital said:
In my entire television engineering career, I've never seen anyone fix a switching or switch mode power supply of any type. Cut your losses and just replace it, would be my advice. Maybe you could return it to the manufacturer for repair?

Dave
Click to expand...

I'd consider just buying another, but thats nearly $400 that I really don't want to spend right now. I'd like to try fixing it myself first, if at all possible. I loathe the idea of paying people to fix things for me. Not because of the money issue, but because it doesn't help me understand why it broke or how to fix it or a similar one in the future. Once I know how and why it broke, and how to theoretically fix it, I'll pay people to do that work but I dislike having to rely on others to do things for me unless I absolutely have to.



I'll try to contact the mfg to see if I can get a schematic. I'll also re-check the DC voltage. It does have the big caps in series, so it may be a doubler. I'll have to look at it more closely. I'd also agree that the quick and dirty mosfet check doesn't prove they aren't bad, but they aren't dead shorted and aren't totally open either. I'm getting no input at all to the gate, so I'm thinking the problem is upstream of that point. I'll see if I have voltage input to the main switch controller and the hex inverter as well. It may be an internal supply problem that isn't getting voltage to those parts.
 
In my entire television engineering career, I've never seen anyone fix a switching or switch mode power supply of any type.
Click to expand...

Really? Wow. They take a bit more time than a linear supply, but it isn't rocket science.

Be pleased if the unit uses mostly discreet components, SMD based switchers are a royal pain.

I'd say its the PWM controller, especially if you haven't blown the main fuse. If the main fuse is blown, it's the high speed chopper transistor/MOSFET, or one of the associated diodes (usually a zener). Check the small electrolytics on the primary side even if they don't look swollen or leaky, and be certain to use an analyzer that can check ESR values on those caps (very important). Many times, these timing caps will be placed close to a heat-sink or other heat generating device that dries them out quickly - when I replace them, I'll always use a 105 degree C rated device and sometimes even insulate with Teflon and stand them proud from the board. Also, be careful - the frequency at which these PSUs operate can leave a nasty RF burn, and the HV can jump to ground quite easily at that frequency. Caution is indicated.

Good luck,

-Richard
 
palasr said:
Really? Wow. They take a bit more time than a linear supply, but it isn't rocket science.
Click to expand...

Broadcast equipment is full of switch mode power supplies.

We had a Grass Valley 300 video switcher run by these and one of the supplies died. It was a Hewlett Packard supply, so we took it out to the regional service center in Atlanta. They worked on it all day and still never fixed it. They finally admitted they couldn't fix their own stuff and they usually just returned the customer a new power supply of the same model.

We had a huge intercom system powered by 12 switching supplies. At any one time there were about 3 dead ones. (It required 6 to operate and featured dual redundant power supplies) We had a regular shipment routine with the manufacturer. Even they couldn't fix them and were just sending us new supplies. They finally abandoned the switching design and went back to analog because they couldn't fix them and they were their own in house design.

From our experience, there are just too many feedback paths and exotic parts to be serviceable. If you miss one bad part, the next time you turn it on to test it, all the previously replaced parts get fried, etc. We had one piece of equipment that used a 5V 250A supply. It was quite small for it's capacity, and you couldn work on it without taking it all apart, and it wouldn't work unless it was totally assembled. (A lot of interconnected boards and parts) We gave up on those and just bought replacements.

I'd be willing to bet that B&K offers a flat rate repair fee or else sells a replacement board that would be the best way to pursue this repair.

Of course, that's just my opinion and experience. Your mileage may vary.

Dave
 
I already switched out the obviously bad ones. I won't rule out the rest of them, but visually they look OK. They are the crappy CapXon caps that go bad in everything though, so I wouldn't be at all opposed to replacing them on general principals. I just don't want to go to the effort of shotgunning all of the little SOBs only to get no results. Tomorrow I'll poke at it some more and see if I'm even getting voltage input to the main controller chips, and then I guess I'll go from there. I also need to email B&K to see if they'll spit up a schematic for me.
 
Old TV circuits had a switcher in the horizontal output. When you turned on the TV it put out a "kick start" pulse. There was a low voltage secondary on the xfmr that was used to power the switcher. So at turn on if the kick start didn't power up the switcher then the power supply never got going.

If this switcher works the same way, it's possible that since the power supply is not working properly you won't get a power supply voltage on the chip.

You've got to be very careful probing around with a scope or meter. With the high voltages here a slip up can be painful. In addition, the capacitance in a scope probe can act as a signal generator on high impedance digital circuits. If at all possible, make test connections with power turned off then turn on power.
 
I've fixed a couple of DC-DC convertors, but they are not really the easiest things to tackle, especially if you're starting out.

To work on them, I'd strongly recommend:
An isolation transformer. Much of the circuit is hot and you may need to attach grounded equipment.
A scope. The signals are switching signals and you may have to probe the controller IC (if fitted), or discreet components.
A Schematic. There are many flavors of switching supplies. I've never found any two alike. Similar? yes. The same? no.

That would be the foundation for starting work on this. A generous portion of experience will help here, because sometimes subtle problems will make a switcher not oscillate. Knowing what to ignore or concentrate on will really help.
 
I've been fixing switching supplies for years. Just because no one you know knows how to do something, doesn't mean it can't be done. The ones that are run by a chip like this one are the easiest. I don't think it will require a kick pulse to start it, the chip should just start oscillating. But the chip requires a power supply to work. Look at the datasheet. You should have 41V on pin 12. Pin 7 is ground. In this circuit, the ground for the switcher is NOT the same ground as for the secondary supply or amp. In fact, it probably is one of the mains. So you need to measure the switcher supply with respect to the ground on the input cap. You must connect the circuit to an isolation transformer if you want to look at it with a scope, otherwise you will connect one of the mains to ground with the scope ground and sparks will fly! If your meter ground is not isolated you need to use the isolation transformer. If your meter is battery operated, you're ok.
The chip outputs are pins 8,9,11, 10. These come from transistors in the chip, with pins 8 and 11 being the collectors that may connect to the supply or to resistors to the supply, and the collectors may drive the switching FETs. If that is the case, pins 9 and 10, the emitters, will connect to ground. Most likely, pins 9 and 10 will drive the FETs. Ohm 9 and 10 out to ground. If they don't connect to ground they are the outputs. If you have an isolation transformer then you can scope the outputs. If there is no output and you have power supply to the chip, the chip is probably bad.
NOTE:, when replacing the caps in a switching supply, use only 105C caps in all locations.
 
You must log in or register to reply here.
Back
Top Bottom