That's it for tonight
Check all those passives john... hopefully this can be a success story.
Do we assume the amp proper is Ok... have you checked the audio outputs just for interest to make sure they are not short.
Well I was assured it was all working fine until that PSU gave up and it was turned off very fast. It's very complex in there but everything LOOKS ok at any rate. Not sure how to test the outputs I'm afraid (eeek!).
Anyways enjoy your evening - I'm just about to have a pint of Crabbers - um ummm!
Cheers!
I'm going lol
Yes but bulb across the DC output rather than the windings themselves.
Low value resistors usually check OK in circuit... yes they do interact so if you are unsure over any then lift one end and confirm. Many will read OK in circuit though.
If R622 has failed open or gone high that may have caused a bit of other damage... so we need to sure exactly what parts have failed.
It's a case of pieceing all the evidence together first and rebuilding carefully rather than just replacing any thing obvious.
Goodnight
Yes but bulb across the DC output rather than the windings themselves.
Low value resistors usually check OK in circuit... yes they do interact so if you are unsure over any then lift one end and confirm. Many will read OK in circuit though.
If R622 has failed open or gone high that may have caused a bit of other damage... so we need to sure exactly what parts have failed.
It's a case of pieceing all the evidence together first and rebuilding carefully rather than just replacing any thing obvious.
Goodnight
That 0.22-ohm resistor reads 1.6 on the 200-ohms setting on my multimeter. Doesn't sound right so I'll lift a leg on it and measure again tomorrow.
It's labelled 0.22 then the ohm (omega) sign, then a K afterwards. Hope it hasn't been "repaired" with a 0.22k instead!
Incidentally here's the two halves of the PSU in better quality than that manual scan (hope the original poster of these on Audio Karma doesn't mind my using them here to fault find):
Will replace with fresh scanned images from a proper copy of the user manual once mine arrives from Canada.
- John
It's labelled 0.22 then the ohm (omega) sign, then a K afterwards. Hope it hasn't been "repaired" with a 0.22k instead!
Incidentally here's the two halves of the PSU in better quality than that manual scan (hope the original poster of these on Audio Karma doesn't mind my using them here to fault find):
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Will replace with fresh scanned images from a proper copy of the user manual once mine arrives from Canada.
- John
Last edited:
That's a bit clearer...
The 0.22 ohm is critical as it's used to sense current in the switching transistors to provide overload protection. You need to be sure on this one... and measuring low values isn't always easy. Does your meter read 0.00 with the leads shorted ? 0.22 ohms is getting near the resistance of the actual meter leads and contact resistance so I would expect a reading of around 0.3 or even a little higher on a normal meter and leads.
I suspect better to replace it tbh as it will have been stressed by the transistors failing.
The 0.22 ohm is critical as it's used to sense current in the switching transistors to provide overload protection. You need to be sure on this one... and measuring low values isn't always easy. Does your meter read 0.00 with the leads shorted ? 0.22 ohms is getting near the resistance of the actual meter leads and contact resistance so I would expect a reading of around 0.3 or even a little higher on a normal meter and leads.
I suspect better to replace it tbh as it will have been stressed by the transistors failing.
Morning Mooly!
Yeah definitely better to be safe than sorry - a 5W 0R22 has been added to my online Farnell list. Later today I'm going to remove the 6 dodgy transistors too (Q601, Q613, Q609-Q612 and - possibly - Q602 as that's right next to the two red ones that blew) as well as all electrolytics, to be replaced with 109C low ESR types. I'll also test the remaining resistors.
Shall I go ahead and order 6 BUT11A transistors or the TIP150s?
Cheers,
- John
Yeah definitely better to be safe than sorry - a 5W 0R22 has been added to my online Farnell list. Later today I'm going to remove the 6 dodgy transistors too (Q601, Q613, Q609-Q612 and - possibly - Q602 as that's right next to the two red ones that blew) as well as all electrolytics, to be replaced with 109C low ESR types. I'll also test the remaining resistors.
Shall I go ahead and order 6 BUT11A transistors or the TIP150s?
Cheers,
- John
Last edited:
Hi John,
The 0.22 ohm needs to be a non inductive (or low) type really...
C604, 606 and 608 need replacing along with C613. Are you planning replacing all the electroylitic caps in the PSU ?
BUT11 are pretty universal. I mentioned the BUT11AF but I see that the tabs may be by design in contact with the board below... I can't tell from your picture. If they are then get the BUT11. The BUT 11AF is a 100% isolated package and needs no isolating washer.
The 0.22 ohm needs to be a non inductive (or low) type really...
C604, 606 and 608 need replacing along with C613. Are you planning replacing all the electroylitic caps in the PSU ?
BUT11 are pretty universal. I mentioned the BUT11AF but I see that the tabs may be by design in contact with the board below... I can't tell from your picture. If they are then get the BUT11. The BUT 11AF is a 100% isolated package and needs no isolating washer.
Would something like this fit, at 22mm by 8 mm
WELWYN|W22 0R22 JI|RESISTOR, WW 7W 5% 0R22 | CPC
WELWYN|W22 0R22 JI|RESISTOR, WW 7W 5% 0R22 | CPC
Amazed that CPC don't list the BUT11 as it was such a common device.
Looking at,
ON SEMICONDUCTOR|MJE13007|SI-N 400V 8A 80W | CPC
which on first examination looks a reasonable choice.
Looking at,
ON SEMICONDUCTOR|MJE13007|SI-N 400V 8A 80W | CPC
which on first examination looks a reasonable choice.
This was the one I was looking at as it's the same type as the original:
WELWYN|SQM5-0R22JI|RESISTOR, 5W, 5%, 0R22 | Farnell United Kingdom
Doesn't say anything about low inductance though, though as it's a ceramic wirewound I would have thought it's low enough already?
Anything I order has to be from either Farnell or Ebay at present also. Farnell have both the BUT11A and the TIP150s in stock.
In answer to your other question all the following transistors are metal backed:
Q601, Q613, Q610-612, Q602. Q604 has a metal tab but is free standing. Q603 is a circular metal can'd Sony A911.
I was looking at this article about testing transistors about halfway down: http://www.kpsec.freeuk.com/components/tran.htm
I'd rather only replace those that need replacing if poss.
- J
WELWYN|SQM5-0R22JI|RESISTOR, 5W, 5%, 0R22 | Farnell United Kingdom
Doesn't say anything about low inductance though, though as it's a ceramic wirewound I would have thought it's low enough already?
Anything I order has to be from either Farnell or Ebay at present also. Farnell have both the BUT11A and the TIP150s in stock.
In answer to your other question all the following transistors are metal backed:
Q601, Q613, Q610-612, Q602. Q604 has a metal tab but is free standing. Q603 is a circular metal can'd Sony A911.
I was looking at this article about testing transistors about halfway down: http://www.kpsec.freeuk.com/components/tran.htm
I'd rather only replace those that need replacing if poss.
- J
Last edited:
Ah good find! Do I need to worry about the hFE with these modern transistors or are they all pretty well matched these days compared with those old Sankens? The Sanken datasheet for the 2SC2023 says vce=4V, Ic=0.5A (which means nothing to me I'm afraid!).
That resistor should be fine.
Testing transistors isn't always foolproof... experience comes into play as well. If a device reads obviously duff then fair enough.
I think I would go for BUT11's as I see they are available from Farnell but given the cost it might be worth getting 4 MJE13007's as well to use in the switching section. Buying from Farnell and there's a good chance they are from the same batch. I wouldn't use ebay for parts. CPC is part of Farnell by the way. Free web delivery if you order before midnight Monday and order over £10... not sure if you need to be an account holder to take that up.
The hfe shouldn't really be an issue with modern devices... famous last words.
I would be surprised if Q604 were faulty.
Q602 and 603 and anything is possible as they are directly connected and failure of the pass transistors Q601 and 613 could see these two drivers off.
I think the first step is too replace and rebuild the first section, the linear regulator around Q601 and 613 and to omit fitting the four switching transistors. Doing this and the operation of the reg can be confirmed by altering the preset pot. Next step would be to connect a small load (60 watt bulb) across C613 and confirm the reg maintains the same voltage.
Then replace the switching transistors and disconnect L607 and 608. Fit a 100 watt bulb in series with the mains at this point and see if the thing seems happy. Confirm the DC output on the secondary side.
I was thinking more on this and some PSU's are not happy without a load. The 100 watt bulb while a good safety feature limits what can be drawn from the PSU of course, so maybe use a small bulb (fridge type, low wattage) might be sensible in conjunction with the 100 watt.
Testing transistors isn't always foolproof... experience comes into play as well. If a device reads obviously duff then fair enough.
I think I would go for BUT11's as I see they are available from Farnell but given the cost it might be worth getting 4 MJE13007's as well to use in the switching section. Buying from Farnell and there's a good chance they are from the same batch. I wouldn't use ebay for parts. CPC is part of Farnell by the way. Free web delivery if you order before midnight Monday and order over £10... not sure if you need to be an account holder to take that up.
The hfe shouldn't really be an issue with modern devices... famous last words.
I would be surprised if Q604 were faulty.
Q602 and 603 and anything is possible as they are directly connected and failure of the pass transistors Q601 and 613 could see these two drivers off.
I think the first step is too replace and rebuild the first section, the linear regulator around Q601 and 613 and to omit fitting the four switching transistors. Doing this and the operation of the reg can be confirmed by altering the preset pot. Next step would be to connect a small load (60 watt bulb) across C613 and confirm the reg maintains the same voltage.
Then replace the switching transistors and disconnect L607 and 608. Fit a 100 watt bulb in series with the mains at this point and see if the thing seems happy. Confirm the DC output on the secondary side.
I was thinking more on this and some PSU's are not happy without a load. The 100 watt bulb while a good safety feature limits what can be drawn from the PSU of course, so maybe use a small bulb (fridge type, low wattage) might be sensible in conjunction with the 100 watt.
Thanks for all that Mooly - I wish I had your knowledge and trouble-shooting skills.
At the risk of sounding very dense, would you mind explaining to me which transistors (and Q numbers) are part of the switching section you mention? And what is the linear regulator around Q601/613 please?
So have I got this right, order 4 MJE13007s for Q609-612, and two BUT11As for Q601/Q613?
Have you any suggestions for replacing Q602 (2SC1775F) & Q603 (2SA911) please? Sorry for the myriad questions Mooly and I'm hoping getting this up and running isn't beyond me. Really appreciate your help.
- John
At the risk of sounding very dense, would you mind explaining to me which transistors (and Q numbers) are part of the switching section you mention? And what is the linear regulator around Q601/613 please?
So have I got this right, order 4 MJE13007s for Q609-612, and two BUT11As for Q601/Q613?
Have you any suggestions for replacing Q602 (2SC1775F) & Q603 (2SA911) please? Sorry for the myriad questions Mooly and I'm hoping getting this up and running isn't beyond me. Really appreciate your help.
- John
Q609-612 are the switching devices. These are used as switches, they are either fully on or fully off, and thus drive the transformer T603 at high frequency.
To regulate and adjust the final output voltage in this design is accomplished by the simple method of adjusting the overall voltage the switching part of the circuit operates at. That second part is performed by the "linear" stage which is Q601 and Q613. These two transistors don't switch on and off rapidly but just vary how hard they conduct in relation to the current required. If the voltage at the output falls, these conduct a little harder to maintain it. The zener diode D602 gives a "fixed" refernece on one side of the error amplifier formed by Q605 and 606. By comparing that reference against the output (the feed via the preset pot) the output can be maintained within close limits over a wide range of output current.
This might help,
http://www.national.com/appinfo/power/files/f4.pdf
Q602 a 2SC1775. That doesn't sound right at all. The 2SC1775 is a small signal low current device. Has someone put that in previously I wonder ? It's only rated at 50 millamps.
Getting something in a T05 package could be a problem if you want it from Farnell.
I suspect this would be OK but it's a different package.
MAGNATEC|BD232|TRANSISTOR, NPN, TO-126 | Farnell United Kingdom
I'll have a think on those... there's always the old favourites of MJE340 (npn) and MJE350 (pnp). I'll see if I can come up with anything else.
To regulate and adjust the final output voltage in this design is accomplished by the simple method of adjusting the overall voltage the switching part of the circuit operates at. That second part is performed by the "linear" stage which is Q601 and Q613. These two transistors don't switch on and off rapidly but just vary how hard they conduct in relation to the current required. If the voltage at the output falls, these conduct a little harder to maintain it. The zener diode D602 gives a "fixed" refernece on one side of the error amplifier formed by Q605 and 606. By comparing that reference against the output (the feed via the preset pot) the output can be maintained within close limits over a wide range of output current.
This might help,
http://www.national.com/appinfo/power/files/f4.pdf
Q602 a 2SC1775. That doesn't sound right at all. The 2SC1775 is a small signal low current device. Has someone put that in previously I wonder ? It's only rated at 50 millamps.
Getting something in a T05 package could be a problem if you want it from Farnell.
I suspect this would be OK but it's a different package.
MAGNATEC|BD232|TRANSISTOR, NPN, TO-126 | Farnell United Kingdom
I'll have a think on those... there's always the old favourites of MJE340 (npn) and MJE350 (pnp). I'll see if I can come up with anything else.
Thanks the above explanation - I can't pretend to understand it on first reading, but I'll print that up along with that link and read that, referencing the actual board in front of me and hopefully someone will pull the light chord on inside my excuse for a brain
Re. Q602 - yep that's what it says in the service manual - 2SC1775F for the UK model (US/Candian version has the 2SC1962). Written on the actual transistor case for Q602 is C1810. Same type 2SC1775F mentioned in service manual for C604 too (and again in actuality C1810).
Last edited:
Thinking about it.... I'm starting to wonder if someone serviced this amp with the SMPS board from a US model, and tried to change the components that needed changing to UK spec, but didn't entirely succeed. Will check the circuit board number and verify that with the service manual and hope that isn't the case......!
EDIT: Just checked and no reference to circuit board numbers so hopefully that PCB is universal, with just parts differences between US and UK spec.
Here's the 'things to buy' list so far then:
Parts replacements for Sony TA-F6B SMPS
Q601 2SC2023 change to BUT11A
Q602 2SC1775F change to ???
Q603 2SA911 change to ???
Q609-Q611 2SC2023 change to MJE13007 (or BUT11As?)
Q613 2SC2023 change to BUT11A
R622 0R22 5W change to Welwyn W22 series, 7W.
EDIT: Just checked and no reference to circuit board numbers so hopefully that PCB is universal, with just parts differences between US and UK spec.
Here's the 'things to buy' list so far then:
Parts replacements for Sony TA-F6B SMPS
Q601 2SC2023 change to BUT11A
Q602 2SC1775F change to ???
Q603 2SA911 change to ???
Q609-Q611 2SC2023 change to MJE13007 (or BUT11As?)
Q613 2SC2023 change to BUT11A
R622 0R22 5W change to Welwyn W22 series, 7W.
Last edited:
Just noticed that Q605 and Q606 you mentioned as being the 'error amplifier' are listed as being 2SA678 types, and the ones actually fitted are A733s. Mistake or are they an equivalent?
EDIT: I think at this point I'll remove ALL transistors and start again from scratch. Will also check the resistors on the board with what's in the service manual - I really don't trust what this 'tech' did to the amp...
EDIT: I think at this point I'll remove ALL transistors and start again from scratch. Will also check the resistors on the board with what's in the service manual - I really don't trust what this 'tech' did to the amp...
Last edited:
We need to be logical with this.
A possible problem could be Q602 as I have had a closer look at the way it's mounted in your first photos. Now the manual as you pointed out
says a 2SC1775 for this... so something amiss here. The 2SC1775 is a tiny small signal device albeit of highish voltage rating... but it's the wrong package anyway. Q602 is bolted to the heatsink. It sounds like the parts list is incorrect tbh.
Looking at the photo Q604 soldering seems original. I would swap this for Q602 which keeps Q602 as original type and package.
That means we have to find something for Q604 location. I think we have to go for an MJE340 here. Is Q604 free standing or does it have any clip on heatsink ?
In the UK 240volt mains will give around 340 volts DC across C601 which is the input to the regulator.
If Q604 is the correct type
So we have,
Q601 = BUT11
Q613 = BUT11
Q602 = 2SC1810 ? is it.
Q603 = MJE350 It's the wrong pakage but that's no problem looking at the photo.
Q604 = MJE340
Q605 and Q606 I think will be OK. MPSA92 would be OK here though... again different pin outs but small TO92 package.
Q609/10/11/12 as MJE13007 initially with BUT11's available.
Remember to take note of any insulating arrangements on transistors and remember that the metal tab is always the collector so is live. I can't make out from the photo whether the four transistors are mounted on insulators or not.
I see the manual shows the two versions of the amp US and UK. Some of the resistors are different between the two, particularly around Q603 and 604.
Solder braid works great on this type of board for instant desoldering and you may need heatsink compound too.
Did you check the audio output transistors. If faulty they would probably read short from collector to emitter.
A possible problem could be Q602 as I have had a closer look at the way it's mounted in your first photos. Now the manual as you pointed out
says a 2SC1775 for this... so something amiss here. The 2SC1775 is a tiny small signal device albeit of highish voltage rating... but it's the wrong package anyway. Q602 is bolted to the heatsink. It sounds like the parts list is incorrect tbh.
Looking at the photo Q604 soldering seems original. I would swap this for Q602 which keeps Q602 as original type and package.
That means we have to find something for Q604 location. I think we have to go for an MJE340 here. Is Q604 free standing or does it have any clip on heatsink ?
In the UK 240volt mains will give around 340 volts DC across C601 which is the input to the regulator.
If Q604 is the correct type
So we have,
Q601 = BUT11
Q613 = BUT11
Q602 = 2SC1810 ? is it.
Q603 = MJE350 It's the wrong pakage but that's no problem looking at the photo.
Q604 = MJE340
Q605 and Q606 I think will be OK. MPSA92 would be OK here though... again different pin outs but small TO92 package.
Q609/10/11/12 as MJE13007 initially with BUT11's available.
Remember to take note of any insulating arrangements on transistors and remember that the metal tab is always the collector so is live. I can't make out from the photo whether the four transistors are mounted on insulators or not.
I see the manual shows the two versions of the amp US and UK. Some of the resistors are different between the two, particularly around Q603 and 604.
Solder braid works great on this type of board for instant desoldering and you may need heatsink compound too.
Did you check the audio output transistors. If faulty they would probably read short from collector to emitter.
Thanks for all that Mooly - all sounds fine to me and all those parts are easily available
Regarding the enigma that is Q602, perhaps we could instead look at the US part number listed in the service manual, which is 2SC1962. It would be rather unlikely (fingers crossed!) they'd get that part wrong twice, so perhaps by looking at the specs for that US part number (taking into account the voltage difference betwixt US and UK) it might shed light on what should be fitted here for the UK model?
http://www.datasheetarchive.com/pdf/getfile.php?dir=Datasheet-040&file=DSA00100702.pdf&scan=n
Then again if you say the C1810 from Q604 may be soldered into the C602 position I'll defer to your knowledge, and I'll order a MJE340 to occupy Q604s empty space.
I'll test the output transistors once I have that board out also - they too have been replaced at some point from the original Hitachi (2SB655 & 2SD675) devices to On Semi MJ21193 & MJ21194s. I happen to have some more of those On Semi spares to hand also which might be useful.
I'm wondering what originally went wrong for the tech to replace many parts on the SMPS as well as those outputs. From reading other threads on this amp I know those 85C caps are the cause of many problems, and a couple of TA-F6B owners have mentioned replacing transistors in the SMPS several times before realising the caps were to blame and causing all the faults. I think that happened here and the tech replaced everything with 'near enough' components, but many of the original (and long dried up by now with the SMPS heat I'd imagine) caps are still there!!! Doh! Those inadequate caps plus the fact he forgot to refit the four switching transistors heatsink clamp in place obviously resulted in the amp going wrong again.
Regarding the enigma that is Q602, perhaps we could instead look at the US part number listed in the service manual, which is 2SC1962. It would be rather unlikely (fingers crossed!) they'd get that part wrong twice, so perhaps by looking at the specs for that US part number (taking into account the voltage difference betwixt US and UK) it might shed light on what should be fitted here for the UK model?
http://www.datasheetarchive.com/pdf/getfile.php?dir=Datasheet-040&file=DSA00100702.pdf&scan=n
Then again if you say the C1810 from Q604 may be soldered into the C602 position I'll defer to your knowledge, and I'll order a MJE340 to occupy Q604s empty space.
I'll test the output transistors once I have that board out also - they too have been replaced at some point from the original Hitachi (2SB655 & 2SD675) devices to On Semi MJ21193 & MJ21194s. I happen to have some more of those On Semi spares to hand also which might be useful.
I'm wondering what originally went wrong for the tech to replace many parts on the SMPS as well as those outputs. From reading other threads on this amp I know those 85C caps are the cause of many problems, and a couple of TA-F6B owners have mentioned replacing transistors in the SMPS several times before realising the caps were to blame and causing all the faults. I think that happened here and the tech replaced everything with 'near enough' components, but many of the original (and long dried up by now with the SMPS heat I'd imagine) caps are still there!!! Doh! Those inadequate caps plus the fact he forgot to refit the four switching transistors heatsink clamp in place obviously resulted in the amp going wrong again.
Last edited:
Hi John,
The 2SC1962 seems to be 500ma rated, which sounds about right. The 2SC1775 is only 50ma. You could fit a new MJE340 here, it's just the physical difference that's the issue.
Can't really say what and why stuff was changed... the key to repairing SMPS is understanding how they work and what the failure modes are. The problem here is the devices are so old that direct equivalents are hard to come by. The regulator part is all standard stuff and should be non component critical to a large extent. The switching circuit is a little more problematic and we have to watch carefully what we fit and test along the way.
The 2SC1962 seems to be 500ma rated, which sounds about right. The 2SC1775 is only 50ma. You could fit a new MJE340 here, it's just the physical difference that's the issue.
Can't really say what and why stuff was changed... the key to repairing SMPS is understanding how they work and what the failure modes are. The problem here is the devices are so old that direct equivalents are hard to come by. The regulator part is all standard stuff and should be non component critical to a large extent. The switching circuit is a little more problematic and we have to watch carefully what we fit and test along the way.