Technics SA-828 Receiver from 70s- Is it worth trying to save this aging patient?

[73's de Edd]

Sir Iceman from McMurdo . . . . . . .

Well, well, well in looking at your good photos of the boards, I see no abnormality as far as any burnt resistors or cracked green and black power output transistors .
You may just find yourself having the rare abnormality of finding an open diode instead of a shorted unit contained within
the FULL WAVE BRIDGE unit with its small top finned heatsink.

In that situation, you can have one supply voltage level being up to its normal amount, while the other polarity will only be at only half that amount.
That would make the power output section angry in not being able to balance out and then it would create excess. current pull and pop a fuse.
Hopefully your probable only one or two power up attempts have not damaged the four large GREEN and BLACK
power output transistors .
Check out their two internal junctions just as you tested the diodes.
Then a replacement of the FULL WAVE BRIDGE unit, just might get your power supply operational again.


73's de Edd

 

[Iceman from McMurdo]

Hmmm, think you may be onto something there Mr 73's de Edd (care to explain the callsign - I'm curious)!

As I understand the testing procedure for diodes, I have 4 legs to test - 2 are AC input with 1 DC- and 1 DC+, correct? And I should get a reading only one way... However, for D701, I get a reading in BOTH directions!

Red probe on DC- going to AC gives correct reading of about 530 or so. BUT, Black probe on DC- to AC ALSO gives me similar reading. Same when checking DC+ to AC (red & black probes both give me similar readings). And, FYI, AC to AC reads about 002 regardless of probe direction.

So... that sounds like a bad rectifier circuit, doesn't it?? If so, how would I go about sourcing the bad part and trying to find a replacement (i.e., where to locate and what specs to order)?

 

[73's de Edd]

Sir Iceman from McMurdo . . . . . . .


You say . . .
" (care to explain the callsign - I'm curious)! "

Hmmmmm . . . . . I had the two aspects of yours, solved from the get go . . .so I guess that I am up one on you.

On the enigma . . . I will give you referencing to the first two, with the last being a given.
73's . . . . .

https://www.google.com/search?newwindow=1&q=ham+radio+term+73's&oq=ham+radio+term+73's&gs_l=serp.12..0i30.399901.404983.0.408863.3.3.0.0.0.0.121.344.0j3.3.0....0...1c.1.64.serp..0.3.344...0i13i30.3iED1gg7324

de . . . . .

http://www.ac6v.com/jargon.htm

de (Dah dit dit . . . dit) is much faster in high speed code than the much l. o. n. g. e. r . . .
from . . . ( Dit dit dah dit . . . dit dah dit . . . dah dah dah . . . dah dah )

Edd . . .is being your given

On the Full Wave Bridge testing:

The first time you came up with one diode junction being open ?, this time you are reading that diode junctions presence, so you must have erred that first test.

I always do a more than thorough test, in that I place one probe to either the + or - marked connections and then use the other probe lead to test to the two AC terminals and expect either two open circuit readings or two good diode junction readings.
And in your unit that seems to be in the order of approximately 565 millivolts across good diode junctions.
Then:
Move the first probe to either the untested + or - marked connection and use the other probe to do the same sequence of testing to the two AC terminals,expecting either two open circuit readings or two good diode junction readings.
Then as a final test . . .of ALL interrelated possibilities I test across + and - with the two possible probe polarities and expect one being an open circuit reading and the other being a double junction voltage reading.
The final test would be using the two possible probe polarities across the two AC terminals and expect
two open circuit readings.

Now in your situation you are reading a short across the FWB's AC connections, so that should not be the case, IF you have clipped the two grey wires loose, that connect to the power transformer.
That windings presence will test out as a short across the FWB's AC terminals.
Either you confirm this as being the case, or you pull the FWB totally off of the board to retest.

Mouser Electronics, a bit to the north of you, is a good place to get a new unit, cheeep if you find that you need one by the final retesting.

Thassssit


73's de Edd
(Best wishes ! from Edd)

 

[Iceman from McMurdo]

Aha, thanks for the callsign explanation (of the obvious, apparently)! Listened to a little shortwave in my youth, but never got into ham radio (looked like a VERY expensive hobby). Wasted most of my disposable income on motorcycles and hot rods (ha, an even MORE expensive hobby)!.

As for the FWB testing, I rechecked and this time got slightly different readings across the 2 AC connections (and yes, both grey wires & 1 green wire are clipped & no longer connect to transformer):
AC to AC = 1800 (same reading regardless of probe direction)

Also checked across the + and - connections:
DC- to DC+ = 005 (same reading regardless of probe direction)

FYI, did as "dorke" suggested and tried powering up receiver (2 grey & 1 green wire snipped) and 5amp fuse did NOT blow. In fact, other parts of the unit appeared to power up normally Tested output voltages across the grey & green wires and got exact values he said to expect, so I think the transformer is good.

 

[Iceman from McMurdo]

OK, retested the transformer resistances between grey & green wires & got some fluctuating values, but did finally get some readings that approximated what you said to expect (grey-to-grey roughly double the value of grey-to-green).

Also, did as you suggested and tried powering up receiver (with 2 grey & 1 green wire snipped) and "voila!" - the 5 amp fuse did NOT blow! In fact, other parts of the unit appeared to power up normally Then proceeded to test output voltages across the grey & green wires and got exact values you said to expect, so I think the transformer is good.

As 73s de Edd suggested, I next tested the D701 Full Wave Bridge as the most likely culprit and here's what I get.
Red probe on DC- going to AC gives correct reading of about 530 or so. BUT, Black probe on DC- to AC ALSO gives me similar reading. Same when checking DC+ to AC (red & black probes both give me similar readings). As I understand it, his is bad, right? I should get a reading in only ONE direction - NOT both!

Also, FYI, AC to AC reads about 1800 regardless of probe direction, and DC to DC reads about 005 regardless of probe direction. And these readings are all taken w FWB still attached to board and both grey wires and the one green wire clipped & no longer connected to transformer.

So, think I need to replace the FWB as the next step (or at least remove it from board & test again in isolation)?

 

[Sunnysky]

In diode test mode forward bias may read 650 (mV) and blank for reverse give or take a wide margin.

in Ohm mode , it may read a few Ohms and > 1MOhm for reverse

Anode (like A in symbol) is +. and DMM's apply + on + terminal., wheras old analog meters were reverse polarity bias on Ohm scale)

Thus V- to V+ with diode test mode will read two diode drops or 1300 +/-20% and blank or open cct.


Any discrepancy means NFG and replace with same or better.. Digikey.com ship same day.

 

[Iceman from McMurdo]

Thanks for chiming in with your analysis SunnySky! Not 100% I'm completely understanding everything you say, but believe you're agreeing that the FWB is probably shorted if I'm get the same voltage drop readings (approximately 0.4 V) in BOTH directions when my DMM is testing in diode mode, correct?

BTW, closer examination of my FWB rectifier component reveals a marking of S10VB20 on the side, which appears to be hard to source based on an initial search query. Mouser Electronics tech assist suggested an alternate unit with equivalent specs might be GBPC5002, which I've successfully located multliple sources for... However, if I go with that unit it will take some serious creativity to swap-in & mount since its connection leads are completely different than the original S10VB20 mounted in the receiver. Hope to be able to re-use the original heat sink but, if not, I've located a new one that fits the GBPC5002.

Can anyone confirm I'm headed in the right direction with this approach??

 

[Sunnysky]

A 50A diode will easily replace a 10A diode but bigger but 200PIV may cause failure during power off transient. I would suggest MOV and ICL for OVP and OCP to reduce stress on caps and semi's.

try GBPC search for largest one that fits in 600PIV
http://www.comchiptech.com/products_search.php?banner=top2&partnumber1=gbpc&link=products_search.php

Digikey stocks them in -W wire leads

I am assuming you tested out of board.

 

[dorke]

O.K were lucky here, the transformer is o.k

I think the bridge-rectifier is probably fine as well.
I assume you are testing the bridge-rectifier "in-circuit",Is that so ?
The "DC-DC" reading of 005, is almost a dead short load on the output of the rectifier.

In that case that short on the output of the bridge needs to be found.
The most likely defective device(s) are the output transistor pairs.
Please check the resistance(in diode test mode) of the 4 output transistors Q635-Q638 (the 4 big ones in the "square")between C-E,B-E,B-C and the same with reverse polarity of the probes(a total of 6 measurements per transistor).
What do you get?
BTW,which meter do you use ,a digital one (DMM )or an analog one(with dial)?

 

[Sunnysky]

Any suspect part must be confirmed out of circuit in lowest ohm scale. dorke's suggestion is also good if bridge is confirmed good. I hope you have a solder sucker or fine braid for solder wick.

 

[Iceman from McMurdo]

OK, appreciate everyone's help so far, but please keep in mind I'm no electronics expert... I'm an enthusiastic learner & know how to identify a resistor from a capacitor, but when we start getting into FWBs and transistors, and how to test them & interpret the results, I'm definitely reaching my limits - at least until you all can educate me. So, please, if you can, walk me thru this at a little slower pace & don't assume I understand all the acronymns you all probably know by heart...

To answer some of your questions, yes, I'm testing everything (so far) "in circuit" - I've not removed any components from the board (although I understand any suspect components will likely have to be confirmed bad by testing again once removed).

Sure thought it seemed like the bridge-rectifier was the cause of my troubles, but now "dorke" is saying the bad readings I was getting on that component (while testing in circuit) might be caused by a shorted output transistor further down the line. Hmmm, ok, performed the suggested testing on those four components (Q635-Q638) and got the results below:

Q635: Q636: Q637: Q638:
B-C = 075 (red probe to base) B-C = 556 B-C = 297 B-C = 670
B-E = 075 B-E = 552 B-E = 297 B-E = 684
C-E = 001 C-E = 1082 C-E = 001 C-E = 1080
B-C = 076 (black probe to base) B-C = 670 B-C = 287 B-C = 528
B-E = 075 B-E = 688 B-E = 287 B-E = 538
C-E = 001 C-E = 1060 C-E = 001 C-E = 1063

Ugh... Lots of various readings there - what does it all mean?? Multiple failures maybe??

And, BTW, my test instrument is not very sophisticated - just a cheap little DMM with a digital readout, although it does have a diode test function and that's what I used to get the readings on the FWB and the power transistors above. Where do I go from here??

 

[73's de Edd]

Well since you didn't take the FWB out of circuit for its testing.
The erroneous reading from the AC to AC terminals of the FWB is actually being related to
your BAD Q635 and Q637 transistors. There also might be collateral damage to immediately
associated transistor / diodes.


73's de Edd

 

[Sunnysky]

Q635&Q637 CE are both obvious shorts. The others are ok.

http://www.ebay.com/itm/2SC4029-2SA1553-Power-Transistor-15a-230v-150w-NPN-PNP-BY-TOSHIBA/161173356439?_trksid=p2047675.c100623.m-1&_trkparms=aid=222007&algo=SIC.MBE&ao=1&asc=37428&meid=d02da78901c745c29368fe2bfa071051&pid=100623&rk=2&rkt=6&sd=160965253716
less expensive replacements

or more .. Note suffix letter determine current gain, may be important. O is higher (better)

http://www.ebay.ca/itm/5PAIR-2SB755...dd70dd5&pid=100005&rk=3&rkt=3&sd=120969037876

 

[dorke]

It looks like the Left Channel power amplifier of your receiver is blown.
You have 2 burned "bandits" Q635 and Q637 .
Their C-E junctions are shorted (1 ohm) as expected ,creating a short on the bridge rectifier.
There could be some more bad devices on the Left Channel.
We shall do some more tests later on.

The transistors should be removed ,in the proper way only,
so no damage is done to the PCB!
Therefor,as by your own words, you don't have good skills on de-soldering and soldering.
I would advise you to use solder-wick of good quality(chemtronics 0.1") for that.
Never use a "hand held pump device"!!!
But before you actually try that,take your time and learn from this soldering resource ,
it is well worth every single moment...

Please take some very good close-up photos of the blown transistors.
from top and bottom.
Is the white thermal grease(between the transistor and metal chassis dry?

Edit:
Don't run to buy any device yet.
These so called "Bat wings" Transistors are long obsolete.
If you buy on Ebay you may unwillingly enter the fitly world of Chinese counterfeiting...
Believe me, you don't want to go there

 

[Sunnysky]

best price http://www.ebay.ca/itm/1pair-Transi...057889?hash=item5d5ef4c461:g:4dUAAOSw1DtXEbUp

get two pair

 

[Iceman from McMurdo]

THANKS to all for the helpful troubleshooting advice thus far, especially the links to potential component replacements, & (of course) the training suggestions - I will avail myself of all. Additionally, I will definitely pick up the recommended soldering wick & invest in a more appropriate soldering station before attempting any repairs.

Please stay tuned to this thread and don't abandon me while I take a few days to get smarter & acquire the necessary skills & equipment to move forward. In the meantime, I'm attaching pics of the 2 suspect transistors:

Q635 (top & bottom views):





Q637 (top & bottom views)





PS - I checked the thermal paste between the transistors & the big side-mounted heatsinks and it was still gooey (NOT dried out)... Also, as you'll notice, Q637 (as well as Q638 for the right channel) has an additional smaller heatsink attached (which Q635 & Q636 both lack).

 

[dorke]

Please show us a photo of the soldering iron you already have(tip included)
,it may be good enough for this repair.
You will need a good 40w iron with "pointed" chisel type tip ,
and good thin soldering wire 0.025" diameter(like Kester 44 ) as well.
Soldering stations may be expensive and not always needed,like in this case.

The additional small "heat-sink" is actuality a "thermal-coupler",
so that Q637 and Q625(Q623,Q621) are working at the same temperature.

Please do a resistance test for Q631 and Q633(TO-220 package ,mounted on separate heat-sinks) ,the same test as you did for Q635 and Q637.
Their pin-out are the same.

 

[73's de Edd]

.


Sir . . . . . . the Iceman cometh . . . . .from McMurdo . . . . . . .(Sound)




Relevant to the two small / minor heatsinks:

Notice that you have 2 Series arranged power transistors in the output stages, of each channel, so they are just taking heat sampling of a power transistor from each channel.
Figuring, so goes one, so goes its companion.
Typically a tab of that small heat sink is transferring its averaged heat to either a small diode or transistor case.
That sensing action could then enact parameter shifts in the amp circuitry to compensate for stone cold through hot sensed temperature ranges.
And most importantly, at some ELEVATED temperature, an exceeded threshold level could enact a shut down of the amp to protect it.
From your " TOP " view provided, I can't see the part(s) the tab touches, can you give feedback on that aspect, with your closer examination ?

OBSERVATION:

Those power transistors will be very easy to remove in a non-destructive foil pad operation.
Because . . . . . they are using slotted tabs, just heat the tab until it's lead can be swung out to the side.

DE SOLDERING:

I always use liquid rosin flux on BOTH my worked joint PLUS the soldering wick.
ALSO let GRAVITY help you :
Wick absorption vertically can be bested by placing the board offset by 90 degrees, so you can
work from the top of a joint downward and get even MORE solder extracted / carried away with your wick.

ASIDE:

Rosin flux . . . . .ask me why I haven't bought any within the last ~ 45 yrs.
Ever heard of making maple syrup ?
In my case I have two majestic 55 ft pines in the front of the homestead. About 4 times a year I will get a large galvanized spike . . . . normally used for retaining water guttering . . . . and drive it in and pull it out to make 6 or so holes around the base of each tree. It is driven in about an inch at a 45 degree upward angle.
Then two toothpicks just fit within the hole and are inserted in each hole.
Sap flows down and creates amber drops . . . . initially soft . . .or rock hard, if forgotten about too long.
Then the toothpick clusters are "harvested" and dropped down into denatured alcohol in a 2 oz storage bottle that is capped.
Then every time it is used, the shaking of the mix, eventually dissolves all, and leaves only bare toothpicks.
A reminder to collect some more pine tar.

And that's . . . . . . .The Rest of the Story . . . . .


73's de Edd

 

[Iceman from McMurdo]

Aha, thanks for the additional insight into the mechanics of temp-sensing protection, as well as the purpose of the thermal-coupler heatsink protecting the nearby components Q631 & Q633. As suggested, took readings on these 2 units (diode mode) and got following values:

Q631:
B-C = 1612 (red probe to Base)
B-E = 1905
C-E = 292
B-C = 624 (black probe to Base)
B-E = 632
C-E = 301

Q633:
B-C = 638 (red probe to Base)
B-E = 642
C-E = 083
B-C = 1587 (black probe to Base)
B-E = 1653
C-E = 083

As for soldering equipment, sad to say I only have a choice of two extremes... Either a small Weller SP-12 soldering pencil (with pointed conical tip, not chisel), or my trusty old Weller 8200PK (which is probably too hot for board work, as it only has two selectable heat outputs of 100 or 140 watts.) I've done some limited electronics work with both in the past, but the patients were generally pretty hardy (and inexpensive) automotive type components. Think if I'm gonna have to desolder & solder multiple components, especially on anything pricey, it's probably time to invest in something more in the middle of those two extremes. Think I've found a nice variable output 60-watt station for about $50 that will be a valuable tool for this & other projects down the road.

Also, great observation on the slotted tab mounting of Q635 & Q637 - hadn't noticed that before & it should certainly make them easier to remove, regardless of what equipment I'm using!

What's the consensus verdict on Q631 & Q633??

 

[dorke]

Q631 & Q633 are probably fine(no dead shorts on them),
the readings show "bad values", but that is due to the faulty Q635 & Q637
e.g. Q633(C-E)=75[Q635(B-C)] + 2.2 + 5.6 =83 .
We should re-measure them after Q635 & Q637 are pulled out .

About the soldering iron,
Agree you don't have the correct tool.
If you go with Weller you can't go wrong,
I'm using using a Weller WECP-20 station(variable temp.) and very pleased with it.
with LR-21(50W) and MLR-20(25W).

Which station are you looking at?