I broke a wire on one of these in my XS 750. Does anyone know where can I find one of these diodes or possibly a substitute .

One thread about a year ago talked about it. You may have to search for stv-4h to find it. It either talked about some one was selling it or get one from a parts unit. It is used in many pineer amps. Some people glue/epoxy/pot a few diodes in series and rig up a mounting on the heat sink. If the potting have good thermal conductivity then it should work fine.

Not sure how reliable they are but here:

http://www.encompassparts.com/products/?STV4H
http://www.ceitron.com/semi/ceisanken.html
http://electronicsexpo.partsearch.com/GetModel.aspx?MfgName=Sony&BrandName=Aiwa&ModelNumber=AA8300&page=3

Thanks I will give it a try

hello,

I have a Pioneer Spec2 that has a broken STV-4H also. I've been searching for months without succes.
The last link seems to have them available but doesn't ship to Belgium. If it so they have the right part could you please let me know ?

thanks in advance

Most of the ones that show them in stock usually come back with on order.
Haven't been produced for many years. Very tough to find.

If you have a good one on the other chanel, measure the forward voltage of the good one. Find two general purpose diodes and series them to equal the voltage drop of what you have now.
You must figure a way to mount the new one to the heat sink. Sometimes you can use heat sink compound to attach.
They keep the bias stable and need to detect the heat from the sink.
If you do one you should probably do both side the same.

Sounds like a candidate for a couple of 1N4148's in series.

Stv-4h is 2.10v to 2.35v @ 7mA.

Three 1n4148 diodes at 7mA have a forward drop of 2.15v at (my) room temperature.

Is it 3 or 4 diodes? I have seen it both ways and should I do both sides or leave the good one in?

Stv-4h is 2.10v to 2.35v @ 7mA.

Three 1n4148 diodes at 7mA have a forward drop of 2.15v at (my) room temperature.

Is it 3 or 4 diodes? I have seen it both ways and should I do both sides or leave the good one in?

There is nothing that would withstand 4 diodes.

A stv-3h is 1.6 to 1.8 volts and two diodes would be 1.43v, so three diodes at 2.15 would be way too much voltage / bias / idle current for a stv-3h.
A stv-4h is is 2.10v to 2.35 volts and three diodes would be 2.15, so four diodes at 2.86v would be way too much voltage / bias / idle current for a stv-4h.

I don't know where you saw four discrete diodes replacing a stv-4h, but I'll almost guarantee that the picture was taken before it was turned on.

Leave the good one in.

You might be better off with a vbe multiplier circuit - BUT I have not actually compared the two yet as to temperature performance.

edit: I have seen Vbe multipliers used in Pioneer equipment, the sx-3400, sx-3500, sx-3600 and sx-3700 for starters.

Look in the schematic for the number. That is the best estimate unless you look into the device spec. Silicon diode is 0.6 to 0.7 volt each. Germanium diode is lower. The tem coeff is what maters.

Maybe one of these will work: http://www.partstore.com/PartSearch.aspx?PartNumber=stv-4

I got mine here a few months back

Enjoy,
Rich P

I just restored a Marantz 2226 which came to me with the leads broken off both of the bias diodes. I took my handy exacto razor saw and removed some of the plastic casing to expose enough lead to solder to. It was a little bit of work, but it turned out well. I would post pics, but I accidentally deleted them.:dammit:

I just restored a Marantz 2226 which came to me with the leads broken off both of the bias diodes. I took my handy exacto razor saw and removed some of the plastic casing to expose enough lead to solder to. It was a little bit of work, but it turned out well. I would post pics, but I accidentally deleted them.:dammit:

1. excavate to expose some lead material (1mm)
1.1. use a dremel tool with an abrasive blade, but don''t try to actually expose the metal of the wire, just come close, then with a needle nose pliers, crack away the remaining brittle material around the outside of the wire with some squeezing pressure.
2. get very thin gauge (28?30?) FLEXIBLE / STRANDED wire
3. solder the wire on (very very carefully)
4. make a mold around the back of the diode
5. fill the mold with slow curing epoxy for a permanent fix.


He now has my address, we'll see if he sends the busted one.

edit: he did... and I'm probably going to use a #60 pc board drill, to get close to, but not touch the leads, then pick the thin remainder away with an xacto knife. I think I'll try to keep the depth to 1/16 inch.

Mark, that is almost exactly what I did, except I didn't use a Dremel out of fear of slipping and causing an even bigger headache. Also, I was going to epoxy over the solder joint, but I ended up just using shrink wrap and calling it good.
I gave that receiver to my daughter for her B-day. :music:

hello,

I tried to repair my STV-4h as suggested. I soldered 2 flexible/stranded wires and used " Pattex Repair Express " to hold it in place. It's an epoxy based "knead" sort of glue. You can see it in the picture, it seems stable but would it last long ?

The amp (Pioneer Spec2) seems fine, I can bring both channels to 30mV after 30minutes but there is a problem. Bias slowly rises about 0,5mV every hour.
When it's turned on for a day without playing music bias is 44mV on both channels and the amp is hot.

Seems like it's not the STV's since both sides (repaired and original) behave the same.

http://www.audiokarma.org/forums/attachment.php?attachmentid=124381&stc=1&d=1229777744

hello,

I tried to repair my STV-4h as suggested. I soldered 2 flexible/stranded wires and used " Pattex Repair Express " to hold it in place. It's an epoxy based "knead" sort of glue. You can see it in the picture, it seems stable but would it last long ?

The amp (Pioneer Spec2) seems fine, I can bring both channels to 30mV after 30minutes but there is a problem. Bias slowly rises about 0,5mV every hour.
When it's turned on for a day without playing music bias is 44mV on both channels and the amp is hot.

Seems like it's not the STV's since both sides (repaired and original) behave the same.

http://www.audiokarma.org/forums/attachment.php?attachmentid=124381&stc=1&d=1229777744

First, the key to longevity is to use the very small gauge stranded wire, and if possible to find "flexible" stranded wire - which means when they make the stranded wire they use a lot of smaller wires rather than a few larger ones to make up the "bundle" of the wire. It must be more flexible than the soldered and secured junction.

There is an EchoWars thread about replacing a relay that is part of the Spec-2 bias circuit that is a rather non-reliable part. That's where I would start. It is a very good thread.

http://www.audiokarma.org/forums/showthread.php?p=1180448

I would also replace (carefully) the stock bias and offset trimpots with Bourns 3386H single-turn pots.

There are other spec-2 threads that may shed light on this, or the Master, EchoWars may see this and have answers.

While this may seem counter-intuitive, there is a specific current that flows through the bias diode, generating a specific forward voltage. This forward voltage then varies by the temperature of the heatsink. But if the current through the diode changes from bad relay contacts - all bets are off.

Also important is the rest of the spec-2, what has been done to it, since it left the factory?

ALSO, have you checked the current balance between the three transistors in each output group under correct and extreme bias?
(Do you know what I'm talking about?)

This does deserve it's own thread.

I put 3 in series and attached to the heat sink and it works for about 3 to 5 minutes then shuts down. Must have done something wrong. It cools down and works again. I won't try it again.

I put 3 in series and attached to the heat sink and it works for about 3 to 5 minutes then shuts down. Must have done something wrong. It cools down and works again. I won't try it again.

I wouldn't risk that, and I haven't forgotten about your diode, it is continuously in my line of sight, and is a high priority as it's a small job.

edit : further experience has shown there is an additional, as yet unknown, problem with that unit in the amp circuits. (3/3/09)

That's ok I just thank you for even trying to repair it!

Success.

Both the diode's leads were broken off just about flush with the top of the epoxy fill that holds the diode in the horseshoe case.

With a stationary dremel tool (edit: SPEED CONTROLLED, regular rpms would have torn it out of my fingers, if I pressed too hard, the wheel would stop - just right for speed....) with thin cutting wheel acting like a bench grinder I excavated down about 1/32 inch around the leads, and then used an exacto knife to actually expose the metal of the leads. The lead spacing coming out of the package was about 1/16 inch - REAL close... (I used my bare fingers so I would be REALLY careful, and think through everything before EVERY cut.)

Some very small stranded wires were pretinned and whittled down to three strands from 8, and soldered on.

An electrical test mold was made of a pair of the same variety wires with bare metal exposed about 1/4 inch on each, with the metal parallel to each other about 1/16 inch apart and filled with JB Weld epoxy. This material takes 6 to 12 hours to start hardening - thus the mold.

After it hardened, electrical resistance / conductance readings were taken, and were beyond the ability of any of my meters to detect a resistance. Thus assured there would be no electrical problems with using that material, the back of the diode was encased.

After about 24 hours an electrical test was performed, flowing 7 milliamps of current through the diode, and reading a forward drop of 2.28 volts. (it's cold on the bench right now)

The forward voltage drop of diodes when tested depends upon the the test current injected by the meter, and the Japanese specs (and most others) specify the forward voltage drop at a specific current, seven milliamps. Thus when reading these diode's forward voltage drop with your meter's diode test the voltage reading WILL most assuredly vary - YMMV (Your Mileage May Vary).

Most Pioneers that use these diodes also set the current flowing through them at 7 milliamps.

This diode's epoxy is a little larger than the case, so from now on it will need a thin washer between itself and the heatsink to avoid putting pressure on the epoxy - but it was a necessary trade off for me, allowing for a more stable grip for the epoxy on 5 sides of the diode rather than just the solder joints and back face of the original.

http://www.audiokarma.org/forums/attachment.php?attachmentid=129387&stc=1&d=1231891540

Nice work, Mark!

Tom

Yes. A good save.:thmbsp:

Just one more way to keep these old units up and running.
Thanks Mark.

Looks like a winner. Nice job!
Thanks Mark, you make this site the best on the web. Can't wait to get it installed.

Do the leads break anywhere, or just up against the body typically?

I wonder if it would be worth the risk of putting some reinforcing epoxy around the base for support if they are off the heatsink for any reason?

I think that those diodes are the only thing I am nervous about when working on my recievers. I think pioneer should be ashamed for choosing an obviously sub-par part that frequently breaks when moved after 30 years of use. /sarcasm

Basically they'll break at the point where it flexes most, which basically means up against the body is pretty much where it'll break most of the time. Thats where it'll bend more often than not if you move it because of the epoxy originally there.

Do the leads break anywhere, or just up against the body typically?

I wonder if it would be worth the risk of putting some reinforcing epoxy around the base for support if they are off the heatsink for any reason?

I think that those diodes are the only thing I am nervous about when working on my recievers. I think pioneer should be ashamed for choosing an obviously sub-par part that frequently breaks when moved after 30 years of use. /sarcasm

Basically they'll break at the point where it flexes most, which basically means up against the body is pretty much where it'll break most of the time. Thats where it'll bend more often than not if you move it because of the epoxy originally there.

The diodes were the only ones really available, and the problem is that the leads were solid material instead of flexible, and that solid material is somewhat brittle.

Which I rectified by using flexible small gauge wire, and epoxying it all together - it will be a lot harder to hurt now.

And there's NOTHING preventing folks from changing them over to flexible small gauge wire BEFORE they break off short, and AT LEAST heat shrinking the leads together (but be SURE they're insulated from each other by the FIRST layers!!) to make sure the stranded wire flexes first. Epoxy IS a better choice, but if the last layer of shrink grips the diode body as well as the leads, that'll do.

would there be any benefit to putting epoxy on before you begin messing with it as a precaution?

would there be any benefit to putting epoxy on before you begin messing with it as a precaution?

Not really, if the wire's messed up enough to be ready to crack - I think it's better that it cracks off before the problem area is buried.

On a good diode, I would trim them off to about .150 to .250 inch and solder to them, what is hidden by the epoxy I put on is that even though the stranded wires come straight out, they were soldered to the nubs at a right angle, then the wire bends around in a 270 degree loop (like a question mark(?)), finally coming out of the epoxy straight (the dot end of the question mark(?)) as a strain isolation technique.

Update: I installed the newly repaired diode and had problems "smoke" . No fault of the repair (nice job Mark!) just MY ignorance I guess.

double post??????????

see below:

Changed your mind since the PM?

Gonna try to find what smoked?

Just have to cool down before I dive back in, I don't take failure to well. I don't give up easy just have to rethink......... I have not even taken the cover back off maybe this weekend

No rush, in fact don't rush.... look for burned areas, and melted wire insulation, especially around the output transistor connections.

Maybe a post picture of the board, showing the thermal diode connections.

I will take some pics and post later

These are notes to myself, after doing research:
This is a sx-750 power amp, with 4 to-3 power transistors on one heatsink.
The to-3's are socketed with connector-ized cables to the amp board.
There appears to be a DC feedback path with the outputs removed, as long as once removed, the output transistor's collectors are de-energized.
The amp board may not need to be removed or moved far, for access for troubleshooting, it is mounted to the heatsink..
There are 10 ohm, 1\2 watt resistors between the output's collectors and the drivers collectors that may have smoked, perhaps before the 150 ohm emitter resistors on the drivers popped in an open bias loop situation.
Do voltage readings for output's ecb's.
Topology simple, no current sources/mirrors, just a dual for the diff, a VAS, and a pair of drivers.