Burnt component - MTR2000 500W power supply

[wavetar]

Looking for a component identification. I have a bad MTR2000 500-watt power supply, with a burnt component in the A/C input circuit in position "R3" on the board. The component looks like a large circular ceramic capacitor and is stamped "SG-7" on it's face, although I believe it's some sort of thermistor. I can provide a pic later on if needed. Anybody deal with replacing these & have a source? Thanks,

Todd

[Bill_G]

It's a MOV. I cut them out and run without them.

[kcbooboo]

There are some MOVs in the supply, but R3 is in series with one of the fuses. If you just remove it, the station won't work.

I believe it's an NTC thermistor. The one in my PS measures 3 ohms when cold and as the supply could draw up to 8 amps at 120VAC, the GE CL-30 seems to be the correct replacement.

Bob M.

[wavetar]

Thanks kcbooboo, that makes a little more sense, as the supply definitely doesn't run without it!

Todd

[kcbooboo]

I've got a dead station on the bench right now. I located the L1 and L2 leads leaving the input filter circuit board that go through the casting to the electronics sealed on the other side. I get 0.0 ohms from the IEC input receptacle to one of the lines leaving the board, and 0.3 ohms from the other input. From all the switching PS design schematics I could come up with, there's always a thermistor in series with the input bridge, and its resistance goes lower as the supply draws more current. It also gets hot, which is why there's some protective wrap around it. The other components in there are for RF/EMI filtering. The toroidal coil is two separate windings and each one is in series with one side of the line. As the notice inside that filter assembly states, both sides of the line are fused, mainly because this is a wide-range input voltage supply and neither side of the line would be considered neutral if you feed it with 240VAC.

While not identical, the input circuitry is similar to what's used in the upper left corner of this PC supply:
http://img.docstoccdn.com/thumb/orig/24830169.png
This one uses a 120/240V jumper/switch. The MTR2000 does the same thing electronically with a Triac, which turns the circuit into a voltage-doubler for 120V, while it works as a standard bridge for 240V. The MTR2000 supply has half a dozen 680V 200uF caps.

While we're discussing these supplies, can anyone measure the 28VDC line on a working station and tell me what voltage it is when the station is transmitting? The battery-revert terminals are a convenient place to measure that, or the red wires on the Molex plug feeding the 100w PA. I've got one that drops to 24V or lower and I don't think that's right. I would hope that the 28V line would stay up near 28; if it doesn't I suspect the switching part of the PS. "My kingdom for a schematic!"

Bob M.

[wavetar]

While we're discussing these supplies, can anyone measure the 28VDC line on a working station and tell me what voltage it is when the station is transmitting? The battery-revert terminals are a convenient place to measure that, or the red wires on the Molex plug feeding the 100w PA. I've got one that drops to 24V or lower and I don't think that's right. I would hope that the 28V line would stay up near 28; if it doesn't I suspect the switching part of the PS. "My kingdom for a schematic!"

Bob M.

I don't have one handy to measure, but I can tell you your supply is bad. We've had to replace a few over the years with voltage drop such as you're seeing. They were under Motorola Maintenance Agreement coverage, so we just swapped supplies & shipped for repair, so I can't say what the problem was. If you go into the 'metering' screens, you'll see the voltage readings from many sensors, including the main DC, and you'll see that the station displays it in red, as it knows it's not good.

[kcbooboo]

The supply I've got doesn't always start up, so looking at the metering screen can't be done. When it does power up, several fields show up in yellow but not in red. Sometimes the station will transmit at reduced power then jump up to near the set power level, while throwing PA VSWR pass/fail error messages.

This station is 12 years old and not under any maintenance agreement. I'm considering just attaching a 27V 14A switcing supply to the battery terminals and running it that way.

Bob M.

[wavetar]

This station is 12 years old and not under any maintenance agreement. I'm considering just attaching a 27V 14A switcing supply to the battery terminals and running it that way.

Bob M.

That is probably the easiest thing to do. We've ordered stations in a 'DC only' configuration (ie: minus AC supply) for certain installations where good DC was available and AC wasn't & things work fine that way.

[kcbooboo]

A follow-up to my PS problem: I tried removing the main circuit board. Not possible. There are a dozen high-power devices underneath it that are mounted to the aluminum heat sink. Even after loosening all the screws on all those things, the pcb still would not budge. So I started unsoldering them (easily identified by their proximity to the screws plus they all had flux on them). After undoing half a dozen, I realized that I couldn't get to the rest without unsoldering a bunch of thick wires going to coils/transformers or getting a desoldering tool that was about 1/4 inch outside diameter. So I gave up on it. (Yes I could have slid a big pry-bar underneath and broken the components out, but that makes it a lot harder to repair.)

They may have used some form of cement - rather than ordinary replaceable silicone-based thermal compound - under all those devices, which would explain why the board won't just lift straight up. There are two power resistors adhered to the edge of the chassis that will not budge, even though the compound seems to be rubbery and flexible.

I was able to identify the 120/240V selector IC and its triac, so I shorted out the triac, making it a 120V-only supply. That didn't do a thing. I also identified a pair of bleeder resistors that should be across the two banks of filter caps, but I was not able to measure continuity from all four ends to anything else. I think a foil under the board has been eaten or burned up, but since I can't get the board out, I'll never know.

So I bought an Astron (Uniden-branded ARX330-28B) LSRM-28A 28V 18A linear power supply, connected the battery cables to that, plugged the cable into the station, and turned things on. The station came up with zero errors and is working just fine. It draws 11ADC at 28VDC when making 70 watts. So all of my problems were related to the original (ABB) bad 28V power supply.

Now I could get mine depot-repaired or just buy a new one, but the chances of another one going bad are high, and since they can't be repaired, it would be a waste of between $600 and $800. I'd rather fix a linear Astron any day.

Bob M.