STV-4H questions

wd409

Super Member
Greetings all, this post is part of a larger overall problem im working on but I wanted to make this separate in the hopes of a quicker answer to this specific line of questions (i assume this is allowed, if it isnt, my sincere apologies.) the original thread will be linked at the bottom of this post.

so my situation is this, on a pioneer sx-950, I have one STV-4h that (in circuit) wont give me a reading in either direction on diode test, and the other one gives me a reading of 0.04 one way and 0.03 the other way..

so, my question is, what reading SHOULD these give me, because ive been attempting to find that information and cant, i'd rather not take them out of circuit to measure them if i dont absolutely have to, as ive read that the wires are rather fragile, so far i havnt broken them and if one of them IS good i'd rather not tempt fate if it can be avoided, but i dont know which one of these is the bad, or if both are bad.

I am aware of echowars post on replacements and will be referring to that as well if replacements are required







original thread: http://audiokarma.org/forums/index....rk-on-any-advice-is-appreciated.801437/page-2
 
You are able to remove them, but just be gentle. In circuit reading might tend to give you erroneous readings.
 
ill take them out of circuit and measure them..

but just so im clear, the stv4h is basically 4 silicon diodes in series right? so i should get a forward reading of 2.000 or higher and nothing in reverse right?
 
Affirmative.
With my Fluke 117 they read 2.4v (Froward voltage) and Open the other way.
I recently had a bad STV on a 950 amp I restored, where it only read .6v. With the bad STV, the idle not increase above 1mV.
 
Affirmative.
With my Fluke 117 they read 2.4v (Froward voltage) and Open the other way.
I recently had a bad STV on a 950 amp I restored, where it only read .6v. With the bad STV, the idle not increase above 1mV.

Just curious, where did you find a replacement STV-4H?
 
Here is a new STV4-H and it's reading.
 

Attachments

  • STV4-H 001.JPG
    STV4-H 001.JPG
    47 KB · Views: 146
  • STV4-H 002.JPG
    STV4-H 002.JPG
    67.3 KB · Views: 130
i took them out of circuit, the readings are unchanged. one is open both ways, other one is shorted.
what could cause that? the left side of the board appeared to survive, only the right side of the board had damage, but the left side is the one thats open. i think it kinda does explain tho why that side STARTED at 40mv fully ccw when i tried to adjust the bias, it'd go up when i turn the trimmer but it wouldnt go lower than 40mv. the right side wouldnt do anything at all i just got a straight 0.000 volts on that one.
 
A lot of meters don't generate enough voltage to check 4 diode junctions. (Mine doesn't) so the open both ways may be ok. Simplest would be 9 volt battery and enough resistance to let through a few mA (2k would be about 4 ma, 6.8K would be about 1 mA. Should get 2 to t2.5 volts one direction, and 9 volts (open circuit) the other direction (across the diodes). .
 
Having more history on the unit would be helpful, what the original problem was and what you've done so far. Pictures also help greatly in this case. The SX-950 is not a hard unit to troubleshoot.
 
A lot of meters don't generate enough voltage to check 4 diode junctions. (Mine doesn't) so the open both ways may be ok. Simplest would be 9 volt battery and enough resistance to let through a few mA (2k would be about 4 ma, 6.8K would be about 1 mA. Should get 2 to t2.5 volts one direction, and 9 volts (open circuit) the other direction (across the diodes). .

I never thought about that, will give it a try in the morning. I know my meter can test 2 in series but i dont know where it draws the line. just another reason to get a fluke i guess.

Having more history on the unit would be helpful, what the original problem was and what you've done so far. Pictures also help greatly in this case. The SX-950 is not a hard unit to troubleshoot.

Theres a link to the original thread on the receiver in the first post, it has everything i know about it up to this point.
 
i put a 4.67k ohm (all i could find, seem to have lost my giant box of resistors) in series with a 9v battery and the meter, measuring vdc i got 400mv one way and 9.3v the other way on the left sides 4h, i tried a diode test with this arrangement on some other diodes to ensure i understood the testing process on something safe, all i got was a minus sign, so i didnt bother trying a diode test this way with the 4h.
anyway, the numbers dont quite line up with what you said they'd do but it does seem to do something at least.
 
I've only seen a few 4H or 3H diodes damaged and the DIY had installed them backwards or broke them. I have a EXTREAMLY small inventory of new 4H diodes in stock for my customers. If your not seeing a forward drop of 1.7~1.8 and reverse @infinite you diode is damaged, end of story. My concern is what happen to these diodes. One damaged I can see both something has gone catastrophic. Unless you've isolated the parts that have failed history will just repeat but you already know that. I would start constructing a series of UF4004 to replace your 4H parts and work toward finding why they failed.
 
aye, I dont understand why the left one would be damaged either. ive replaced 22 parts on each side of the board, so 44 in total for the board, and all 8 outputs, but i missed this and that means i coulda missed something else too.

ill link a before and after shot in this thread for clarity but it was my understanding that the left side survived this event, none of what i tested on that side was shorted, i replaced the parts anyway so they matched the other side, it does power up, and comes out of protection (i dare not try that any more until this is resolved) but the problem that led me to where we're at right now is the left channel (or at least the left side of the board) wouldnt go below 40mv bias current, which had me concerned as it was my understanding that fully CCW should have started at around 3mv bias, and then i work my way up to 20, but it starts at 40, and the right side wouldnt do anything, 0.000v on that one which i now know is because the 4h was shorted. all of that said tho.. while im going to double check everything, i THINK at this point the rest of it is ok.

but yeah, i was actually thinking about the 1n3595 in echowars post he showed 3 of those in series matching up pretty nicely with the original.
what ive still got to figure out how to do is seal it in epoxy in a way that still mounts to and works as intended with the heatsink.

as far as, putting 4h in backwards. nah, i never desoldered them from the board until yesterday and even then, just one leg. whatever happened to them, (probably) wasnt me. but then again, hell it coulda been me, my dumb ass did put some transistors in backwards. ive since corrected them, they seem to be ok. i might play it safe and order all new ones when i order the diodes.
 
oh and uh, heres the pictures..
pay no mind to those disconnected output wires in the final shot, i soldered those back on
 

Attachments

  • 011.JPG
    011.JPG
    122.8 KB · Views: 80
  • 012.JPG
    012.JPG
    126.5 KB · Views: 80
  • 005.JPG
    005.JPG
    119.3 KB · Views: 77
ive been reading more of that thread and his various updates along the way, looks like uf1004 or 4004 is the better way to go in that case. the idea of mounting it to a ring terminal with jbweld is pretty clever as well. that certainly make getting it on the heatsink easier. and getting it off again if it doesnt work :p
 
Please check over All the BJT's you replaced. I understand you've identified and corrected some but check again there orientation. Check R71/72 for resistance change, seen these fry a lot. Your on the right track with the 4H replacement, the uf4004 have gained some reliability. I'm not able to get a good view of your board, getting old and need close up's. How about the solder side of that board, any lifted angular rings or broken traces?
 
Please check over All the BJT's you replaced. I understand you've identified and corrected some but check again there orientation. Check R71/72 for resistance change, seen these fry a lot. Your on the right track with the 4H replacement, the uf4004 have gained some reliability. I'm not able to get a good view of your board, getting old and need close up's. How about the solder side of that board, any lifted angular rings or broken traces?
is r71/72 directly under Q8 and Q7?
I see them in the parts list as 1 ohm, but i cant find them on the schematic by their R number, but i do see what appears to be a 1 ohm resistor under that transistor with an R number that i cant make out.

edit: i just measured them, if thats the right ones, in circuit i get 1.4 ohms on each one.
 
Last edited:
on the closeups, i think i know what yer wantin to see, this is the best my camera can do on this as, any closer and it wont focus. as far as a picture of the back of the board goes, ill have to take the outputs out and remove the heatsink to even get close to that. in the meantime though, i didnt damage the traces or the rings. i use a solder sucker to remove the solder and i never force the new parts through, they go in smoothly.



here are some full sized shots:
https://i.imgur.com/IQmnhyQ.jpg
https://i.imgur.com/1iOqc5H.jpg
 

Attachments

  • 022.JPG
    022.JPG
    105.4 KB · Views: 43
  • 023.JPG
    023.JPG
    103.7 KB · Views: 40
With my Fluke 117 they read 2.4v (Froward voltage) and Open the other way.
I recently had a bad STV on a 950 amp I restored, where it only read .6v. With the bad STV, the idle not increase above 1mV.
I mucked that up. Got to come clean.
Had to correct it:
I have a set of new STV3H-Y - Stabistor Diode - SANKEN STV3H STV3HY from the German eBay supplier and they read 1.3 forward and open the other. So yea 1.8 would be correct for the -4's.
Thanks "embrown57" for keeping this thread on track.
 
No problems, I'm sure a lot of us purchased the STv3H from that same seller. I have stocked up on them for now. The 4H on the other hand I have only nine NOS just for customers restoration. I believe the OP is on track to get this unit up and running very soon. It can be a PIA working with "technically induced errors", no shame here. You just have to make a mistake once too set off a cascade of events. Electronics for the most part is not forgiving. I've seen much worse than this happen, the OP is getting off easy and paying his dues. Learning from just something as simple as a BJT polarity is something you have take in stride. No mater at what level of skill there's always something to learn.
 
Last edited:
Back
Top Bottom