I've identified two blown outputs on the same channel of my Marantz 2285B. The first, Q001 (PNP 2SD557), is conducting in both directions between b-e and b-c. The second, Q003 (NPN 2SD427), is shorted from e-c. First question, do I replace all four transistors on this channel or just the bad ones? I've been suggested to use On-Semi MJ21193G and MJ21194G, Digikey has them. Second, is it apparent from the two damaged ones what might have caused them to fail? Do I need to pull all the other transistors on that channel or are some ruled out by inspection? Third, when I do replace the outputs, can I just solder them onto the replacements or do I have to wire wrap? The original installation was wrapped, but the brittle wire broke off while unwrapping. I do know I will need thermal grease and mica insulators, I'll get these with the transistors.
The schematic is here, but you have to be registered to gain access:
http://www.hifiengine.com/manuals/marantz/2285b.shtml
I do have a 400K version of the schematic I can email.
The schematic is here, but you have to be registered to gain access:
http://www.hifiengine.com/manuals/marantz/2285b.shtml
I do have a 400K version of the schematic I can email.
DreadPirate said:
I would buy a new transistor and compare the resistance of the faulty transistors with a new one.
I have had MOSFETS fail fully but some failed with lower than normal resistances so I just binned them and bought new ones.
You need to investigate why the amp blew otherwise you could power it up just to blow a new set of transistors.
1/ Did you drive too low an impedance from your amp ?
2/ Did the amp run very high power for a long time ?
3/ Has a speaker fried, shorted and blown your amp ?
4/ Has your speaker cable failed and shorted out ?
5/ It could be some otehr component has gone in the amp and has applied volts to both output sets and caused them to blow.
My first step when fixing an amp is to remove the output transistors completely and feedback the driver output into the LTP. Only when I am convinced the driver is OK do I even consider adding the output transistors.
I enjoy fault finding, it can very challenging but very rewarding too.
This is why I tend to keep my own amp designs simple so they are easier to fault find.
EchoWars said:
Unfortunately, I did unwrap the connections to the socket rather than remove the sockets from the heatsink. I've made a bit of extra work for myself.
I believe the transistors to be checked are the heatsinked ones? I'll get these checked and all replaced and startup with a dim bulb setup.
How does one torque the transistors properly? Are there beam wrenches for this, they'd have to be awfully small...
one useful trick is to temporarily attach 1.5k/2w resistors in place of the B-E junctions of the output devices and power the amp up with no load. this will provide an indication of whether something else is wrong with the amp. it will also provide negative feedback to the diff amp and you can actually feed a signal to it to see if everything up to the output devices is functioning properly. you can also measure the quiescent bias voltage between the base connections of the output devices. this only works for emitter follower outputs, not CFP or Quasi-Comp.
if you let the magic blue smoke out of one of the resistors you will instantly know you still have a problem, but you only lost a resistor not an expensive output debice.
if you let the magic blue smoke out of one of the resistors you will instantly know you still have a problem, but you only lost a resistor not an expensive output debice.
"what might have caused them to fail? "
Nothing lives forever, it's 30 years old.
"conducting in both directions between b-e and b-c."
The driver transistors need to be replaced as well as all the outputs. Check the b-e and b-c of the drivers, if bad you must replace the bias transistor, Vas, and CCS for the Vas.
Check the emitter and base resistors for the outputs, the resistor betweem the emitters of the drivers, the bias transistor and diode string, and the bias pot.
Check the current limiter transistors.
Due to the age of this piece I would do the outputs and drivers on both channels, cheap insurance (and easier to do before it blows up).
Use a 75W lightbulb in series with one side of the line cord when you fire it up again.
Nothing lives forever, it's 30 years old.
"conducting in both directions between b-e and b-c."
The driver transistors need to be replaced as well as all the outputs. Check the b-e and b-c of the drivers, if bad you must replace the bias transistor, Vas, and CCS for the Vas.
Check the emitter and base resistors for the outputs, the resistor betweem the emitters of the drivers, the bias transistor and diode string, and the bias pot.
Check the current limiter transistors.
Due to the age of this piece I would do the outputs and drivers on both channels, cheap insurance (and easier to do before it blows up).
Use a 75W lightbulb in series with one side of the line cord when you fire it up again.
Heat-sinked devices on-board:
Pre-Drivers
Q706 2SA914 $1.35
Q707 2SC1953 $1.24
Power Amp
Q708 2SC1568 $.91
Drivers
Q709 2SA913 No Longer Available, Expensive and Rare
sub available: MJE15030G $1.30
Q710 2SC1913 No Longer Available, Expensive and Rare
sub available: MJE15031G $1.30
Varister
Q711 STV3HY $10.95
no sub found
I'll go ahead and purchase all of the above and substitute as I test for failure and also replace them on the good channel (but keep original outputs on that side). Does this sound like a reasonable plan or should I also spring for outputs on that side? Also, once subbed in I'll do a dimbulb startup will any adjustments prior to the startup be necessary or should I just go with it and take measurements after everything is working (well, hopefully working...)?
Resistors check out ok.
Pre-Drivers
Q706 2SA914 $1.35
Q707 2SC1953 $1.24
Power Amp
Q708 2SC1568 $.91
Drivers
Q709 2SA913 No Longer Available, Expensive and Rare
sub available: MJE15030G $1.30
Q710 2SC1913 No Longer Available, Expensive and Rare
sub available: MJE15031G $1.30
Varister
Q711 STV3HY $10.95
no sub found
I'll go ahead and purchase all of the above and substitute as I test for failure and also replace them on the good channel (but keep original outputs on that side). Does this sound like a reasonable plan or should I also spring for outputs on that side? Also, once subbed in I'll do a dimbulb startup will any adjustments prior to the startup be necessary or should I just go with it and take measurements after everything is working (well, hopefully working...)?
Resistors check out ok.
"Q711 STV3HY $10.95
no sub found"
Just a three diode stack. Check it on the diode range of your meter, it should meacuse about 2V forward drop if it's OK. If not you will have to kludge one with discrete diodes and glue.
I like the MJE15xxx parts as drivers for the MJ21xxx outputs.
(but keep original outputs on that side)
I always replace them. That's like replacing one bad tire when the other three have 100,000 miles on them.
Start-up with a lightbulb is generally safe. If the relay pulls in (and they usually do) you can drive it at low levels and listen to it or look at it on a 'scope. If the bulb stays bright you have a problem (it should flash when you turn it on and then glow dim). I let the unit run for quite a while before plugging it straight into the wall and then adjusting the bias (and midpoint if it has one).
Sometimes the bulb can flash briefly after several hours of playing at low levels. This is usually a problem in the front end or bias circuit. Freeze-mist can help here.
Chances are it won't come to the worst case as outlined above.
If the amplifier has no or low gain, the cap in the feedback loop to ground is bad. I usually replace this, and input coupling cap (if electrolytic) and add some power supply bypass caps while I'm at it.
no sub found"
Just a three diode stack. Check it on the diode range of your meter, it should meacuse about 2V forward drop if it's OK. If not you will have to kludge one with discrete diodes and glue.
I like the MJE15xxx parts as drivers for the MJ21xxx outputs.
(but keep original outputs on that side)
I always replace them. That's like replacing one bad tire when the other three have 100,000 miles on them.
Start-up with a lightbulb is generally safe. If the relay pulls in (and they usually do) you can drive it at low levels and listen to it or look at it on a 'scope. If the bulb stays bright you have a problem (it should flash when you turn it on and then glow dim). I let the unit run for quite a while before plugging it straight into the wall and then adjusting the bias (and midpoint if it has one).
Sometimes the bulb can flash briefly after several hours of playing at low levels. This is usually a problem in the front end or bias circuit. Freeze-mist can help here.
Chances are it won't come to the worst case as outlined above.
If the amplifier has no or low gain, the cap in the feedback loop to ground is bad. I usually replace this, and input coupling cap (if electrolytic) and add some power supply bypass caps while I'm at it.
I've finally gotten around to wrapping this one up. I've enclosed a schematic of the amp section. I replaced all the heat sink mounted devices and they are shown in bubbles on the schematic. Did this for both sides. I checked all the emitter and base resistors for the drivers and the outputs and the bias pot as well, all checked out ok.
The faults were as follows:
Q706 Driver: e-c shorted
Q001 Output: b-e and b-c shorted
Q003 Output e-c shorted
Have I pretty much covered all the bases here?
The bias pot is set at 60 ohm and goes to 300 ohm. Am I good leaving it where it was, or should I turn it all the way to 300 ohm for a startup?
The faults were as follows:
Q706 Driver: e-c shorted
Q001 Output: b-e and b-c shorted
Q003 Output e-c shorted
Have I pretty much covered all the bases here?
The bias pot is set at 60 ohm and goes to 300 ohm. Am I good leaving it where it was, or should I turn it all the way to 300 ohm for a startup?
just like bose said, set it to 300 ohms. after powering up with a 100 watt light bulb in series with the AC line, let the amp idle for a few minutes (10 or 15). then measure the voltage across one of the 0.75 ohm 5 watt ceramic resistors in the other channel. should read somewhere around 10-50 MILLIvolts. then, choose one of the resitors in the channel you just repaired and hook your meter across it. adjust the bias pot until you get about 50-75% of the voltage you had on the other channel. wait about another 5-10 minutes and adjust the bias again to match the other channel.
we missed something here
is this marantz really 30 years old ????????
cause if it is there is no chance that have working electrolytics inside .....
in this case all electrolytics from every section of the amplifier need to be triple double checked and probably replaced .....
often i ve seen resistors also change values either from overheating or to humid areas .....
use of better quality elctrolytics , be also generous with them , and also replace some of them with MKT also this will upgrade the sound of your amp
plus that this might be the reason that started the fire in the first place
is this marantz really 30 years old ????????
cause if it is there is no chance that have working electrolytics inside .....
in this case all electrolytics from every section of the amplifier need to be triple double checked and probably replaced .....
often i ve seen resistors also change values either from overheating or to humid areas .....
use of better quality elctrolytics , be also generous with them , and also replace some of them with MKT also this will upgrade the sound of your amp
plus that this might be the reason that started the fire in the first place
The service manual (available from hifiengine.com) has one measuring across J741 and J742 (circled in the schematic I provided earlier) to set the bias at 30mv. I will startup with pot at 300ohm, wait as indicated and initially set to 20mv.
May go ahead and also replace the smaller electrolytics on the board, cheap enough to do.
One thing I have noticed is that the wire wrap connections are a bit brittle, several have broken off.
May go ahead and also replace the smaller electrolytics on the board, cheap enough to do.
One thing I have noticed is that the wire wrap connections are a bit brittle, several have broken off.
i've seen a lot of those wire wrap connections. i usually cut the wire, strip off about 1/4", pre-tin the post (after a long time the metal gets a little oxidized) and solder the wire on to the post near the top. that way it's easy to get to if i need to remove the wire from the post, without melting the solder under the board. also check he soldering on the underside of the board, especially the collector pins of transistors and large resistors. thermal expansion stresses can break the solder loose and create "solder rings" where the solder fails in a little circle between the component lead and the pad.
2285B Setting Idle Current
I got the unit started up just fine using my dim bulb tester and am now running full voltage. I've hooked up some test speakers, I'm getting music out it. Having a problem setting the idle current, though. I have zero'ed out DC offset per the manual and moved on to setting idle current. The manual is available here, by the way, this procedure is described in it:
http://www.hifiengine.com/manuals/marantz/2285b.shtml
Requires registration.
The procedure is to hook up a meter to J741 and J742 and adjust trimming resistor R727 to get 30mV. These devices and locations can be seen in the schematic I posted in Post #12. I have no speakers hooked up, set the volume to zero, and have tried with both speakers engaged and disengaged and the reading from this location and for the other channel as well does not respond to my adjusting the trimmer resistor. I had both trimmers initially set at maximum (300 ohm) and have turned it back as much as 1/4 with no results at all, stays locked on zero. I'm using a Fluke 77 set to 300mV DC setting. If I turn up the volume, though, I do get an increase to 10mV and it starts bouncing around a bit.
This is strange, I keep thinking I'm missing something. Seems like an odd way to set idle, but its per spec. Any ideas?
I got the unit started up just fine using my dim bulb tester and am now running full voltage. I've hooked up some test speakers, I'm getting music out it. Having a problem setting the idle current, though. I have zero'ed out DC offset per the manual and moved on to setting idle current. The manual is available here, by the way, this procedure is described in it:
http://www.hifiengine.com/manuals/marantz/2285b.shtml
Requires registration.
The procedure is to hook up a meter to J741 and J742 and adjust trimming resistor R727 to get 30mV. These devices and locations can be seen in the schematic I posted in Post #12. I have no speakers hooked up, set the volume to zero, and have tried with both speakers engaged and disengaged and the reading from this location and for the other channel as well does not respond to my adjusting the trimmer resistor. I had both trimmers initially set at maximum (300 ohm) and have turned it back as much as 1/4 with no results at all, stays locked on zero. I'm using a Fluke 77 set to 300mV DC setting. If I turn up the volume, though, I do get an increase to 10mV and it starts bouncing around a bit.
This is strange, I keep thinking I'm missing something. Seems like an odd way to set idle, but its per spec. Any ideas?
What is 'per spec'? According to your post, you have no idle current. Turn up the volume with a signal (which you aren't supposed to do when setting bias) and you're modulating the bias current (or rather, faking it) with a music signal.
Both channels the same? I'd take a look at supply voltages, or my meter (or even my connection points...easy to get confused there).
Both channels the same? I'd take a look at supply voltages, or my meter (or even my connection points...easy to get confused there).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.