Normal for diodes to be really hot?

[cpt_paranoia]

> What is crowbar protection, good sir? And what's an SCR?

Crowbar protection is used to protect against failures by killing the power supply 'putting a crowbar across the supply'. An SCR is a means to do this: it is a type of semiconductor switch that is turned on by a trigger signal, and turns off when the forward current falls below some threshold. Intended to be on long enough to blow the protection fuses.

The secret to debugging power supplies is to work from the transformer down the chain, making sure voltages are sensible. Start with the AC from the transformer, then the bridge rectifier output (DC, check for low AC ripple). Then be guided by the zener voltage ratings for what should be expected.

Randomly removing wires (like the red wrapped one), or replacing transistors, without understanding whether they are faulty is likely to either cause damage or waste time.

It's unlikely your new caps are faulty. But do check you've inserted them with correct polarity, and they're soldered correctly.

 

[birchoak]

> What is crowbar protection, good sir? And what's an SCR?

Crowbar protection is used to protect against failures by killing the power supply 'putting a crowbar across the supply'. An SCR is a means to do this: it is a type of semiconductor switch that is turned on by a trigger signal, and turns off when the forward current falls below some threshold. Intended to be on long enough to blow the protection fuses.

Very cool! I did not know that.


The secret to debugging power supplies is to work from the transformer down the chain, making sure voltages are sensible. Start with the AC from the transformer, then the bridge rectifier output (DC, check for low AC ripple). Then be guided by the zener voltage ratings for what should be expected.

Okay. I will do that.

Randomly removing wires (like the red wrapped one), or replacing transistors, without understanding whether they are faulty is likely to either cause damage or waste time.

My bad. I promise not to do that.

It's unlikely your new caps are faulty. But do check you've inserted them with correct polarity, and they're soldered correctly.

I always power up with a DBT, but I know that doesn't catch everything. I will check them again; I will feel very, very silly if it's something as simple as a backwards cap. Another AKer (amateur, like me, but very knowledgeable and way more patient) simply put big green Nichicon bipolars in his receiver (different one). That's one way to avoid polarity screw-ups.

 

[birchoak]

So start measuring from the transformer onward, using AC scale. When I get to the bridge rectifier (is this the thing that converts alternating current to direct current?), DC should be coming out the other side, but also check for AC ripple with the multimeter set to AC scale (mine doesn't have fine scale measurements for AC; I assume this is common on today's MMs and is okay?). Keep going, looking for any too high or too low values, especially on the power supply board. Since I recapped the amplifier board, too, I will check that once again. Thank you very much for your straightforward suggestions. Everyone here at AK is so very generous and knowledgeable.

 

[birchoak]

> What is crowbar protection, good sir? And what's an SCR?

Crowbar protection is used to protect against failures by killing the power supply 'putting a crowbar across the supply'. An SCR is a means to do this: it is a type of semiconductor switch that is turned on by a trigger signal, and turns off when the forward current falls below some threshold. Intended to be on long enough to blow the protection fuses.

The secret to debugging power supplies is to work from the transformer down the chain, making sure voltages are sensible. Start with the AC from the transformer, then the bridge rectifier output (DC, check for low AC ripple). Then be guided by the zener voltage ratings for what should be expected.

Randomly removing wires (like the red wrapped one), or replacing transistors, without understanding whether they are faulty is likely to either cause damage or waste time.

It's unlikely your new caps are faulty. But do check you've inserted them with correct polarity, and they're soldered correctly.

Ok, the update you've all been waiting for

I looked at the wires coming out of the transformer and, using the AC scale of my MM, got these voltages going into the power supply. I think maybe the rectifying also gets done in the PS because the trafo wires went straight to the PS board and nowhere else, near as I could tell. Measurements were made by grounding the negative probe of the multimeter to the PHONO ground on the back of the receiver and touching the + probe to the given component. Hope that is the right way or all these numbers are wrong! I can remember cathodes are negative because I dated a girl named Cathy and she was negative. The receiver in question is a 20 watt Concept 2.0.

Transformer to Power Supply PCB
Pin 5: 20.8 VAC
Pin 4: 20.8 VAC
Pin 3: 31.7 VAC Is this too much?
Pin 2: 7.1 VAC Power to lights?
Pin 1: 0.0 VAC I assume this is a ground.

Fuses immediately adjacent to PS
2.5A Fuse: 7.1VAC
1.5A Fuse: 31.6VAC

Power Supply PCB
Stereo Lamp Pin: 9.26 VDC
D506 (white, raised diode) -28.3 VDC @ anode+
D505 " 28.3 VDC @ cathode-
D504 " 28.2 VDC @ cathode-
D503 " -28.1 VDC @ anode+

Are the above diodes rectifying AC to DC? Are they the bridge rectifier thingy?

D501 (black, raised) 42.4 VDC @ cathode- Is this too high?
D502 " 9.0 VDC @ cathode-
D509 (the hot one!) 24.7 VDC @ cathode- "Help me; I'm in diode hell! I'm innocent!"
D507 (small, glass) 12.7 VDC @ cathode-
D508 " 31.9 VDC @ cathode-

Okay, out of Diode Land and on to Resistor World:

R506 (flat on pcb, red) 24.7 VDC
R508 (stripey thing) -28.1 VDC
R503 (big, red, raised) 12.8 VDC
R507 (red, raised) 24.6 VDC
R505 (big, red, raised) 31.2 VDC
R502 42.1 VDC Too much? (Scottie: Captain, she can't take that kind of power! You'll fry the dilithium crystals!)
R504 (stripey thing) 40.1 VDC (Scottie: She'll tear herself apart! Kirk: Damnit, Scottie, I need warp seven, now!
R501 (stripey thing) 9.0 VDC For lights?

SCR "crowbar" G Leg: -28.3 VDC A Leg: 24.6 VDC K Leg: -28.3 VDC (sorry about all the VDCs and VACs; this is to try to drill it into my brain)

TR501: Emitter: 31.0 VDC Collector: 40 VDC Base: 31.5 VDC

Holy crap there's a lot of stuff on this board! Here is the underside of this evil thing:

 

[cpt_paranoia]

Well I think most of that looks right.

Essentially, you have three AC supplies from the transformer
21/0/21
7
32

The first is the main power, and, when rectified by the bridge, will give you +/- (21-0.7)*sqrt(2) = 28.7VDC, so your measured 28.3V looks fine.

The second is the lamp drive: AC for the illumination, and 9VDC for the stereo lamp
(7-0.7)*sqrt(2) = 8.9, so, again, your 9.1V looks fine.

The last is used by the PSU board to generate a number of voltages, probably 32V and 13V
(31.7-0.7)*sqrt(2) = 43.8V, so your 42.4V is fine. This tallies with the 50V working voltage of C504 on the schematic.

TR501/D503/R504 on the scm is a regulator, dropping the 42.4V to 32V nominal, in your case 31V. D503 is marked as a 32V zener.

D505 on the scm is marked as a 13V zener, with R503 as a dropper R, and C509 as smoothing. Note with 16V working volts of C509. You're measuring 12.7V, which is fine.

The SCR gate is the same as the cathode, so it isn't being triggered (needs a pulse above K to rigger). The fact that scm D504 and R506 are the same voltage as SCR anode also shows the SCR is off; no drop across R506, so no current. The SCR cathode is connected to the main -ve supply.

Now, the diode numbering between the scm and your PCB are different. It looks like the correspondence is

scm pcb
501 501
502 502
503 508
504 509
505 507

So, the mystery is the purpose of D504/D509, and why it is getting hot.

It's not marked as a zener (and it would make no sense to have a 32V series regulator followed by a 24V zener).
It's marked as being reverse biased.
It's connected to the 32V series regulator via R505, a 3W 680R resistor.
Why has it only got 24V7 across it?

Since this node disappears off the scm, via pin 5, I don't know what it is meant to power. R505 is intended to be a current limit for when the SCR fires, I think.

Ah. If the SCR fires, and pulls its anode down to its cathode, that would pull the 32V supply down to -28V. D504/9 would then be forward biased, and hold the 32V supply at 0V. Ah2. R505 and R506, both being 680R, would hold the 32V rail at the mid-point between 32V and -28V, just about 2V if the SCR fires.

I'd suggest that the diode is faulty, and is pulling down the 32V series regulator voltage. If that is the case, the current must be (31-24.7)/680 = 9.3mA. So the power in the diode would be 24.7*9.2mW = 229mW. Not sure I'd expect it to get that hot, but...

 

[birchoak]

Well I think most of that looks right.

Essentially, you have three AC supplies from the transformer
21/0/21
7
32

The first is the main power, and, when rectified by the bridge, will give you +/- (21-0.7)*sqrt(2) = 28.7VDC, so your measured 28.3V looks fine.

The second is the lamp drive: AC for the illumination, and 9VDC for the stereo lamp
(7-0.7)*sqrt(2) = 8.9, so, again, your 9.1V looks fine.

The last is used by the PSU board to generate a number of voltages, probably 32V and 13V
(31.7-0.7)*sqrt(2) = 43.8V, so your 42.4V is fine. This tallies with the 50V working voltage of C504 on the schematic.

TR501/D503/R504 on the scm is a regulator, dropping the 42.4V to 32V nominal, in your case 31V. D503 is marked as a 32V zener.

D505 on the scm is marked as a 13V zener, with R503 as a dropper R, and C509 as smoothing. Note with 16V working volts of C509. You're measuring 12.7V, which is fine.

The SCR gate is the same as the cathode, so it isn't being triggered (needs a pulse above K to rigger). The fact that scm D504 and R506 are the same voltage as SCR anode also shows the SCR is off; no drop across R506, so no current. The SCR cathode is connected to the main -ve supply.

Now, the diode numbering between the scm and your PCB are different. It looks like the correspondence is

scm pcb
501 501
502 502
503 508
504 509
505 507

So, the mystery is the purpose of D504/D509, and why it is getting hot.

It's not marked as a zener (and it would make no sense to have a 32V series regulator followed by a 24V zener).
It's marked as being reverse biased.
It's connected to the 32V series regulator via R505, a 3W 680R resistor.
Why has it only got 24V7 across it?

Since this node disappears off the scm, via pin 5, I don't know what it is meant to power. R505 is intended to be a current limit for when the SCR fires, I think.

Ah. If the SCR fires, and pulls its anode down to its cathode, that would pull the 32V supply down to -28V. D504/9 would then be forward biased, and hold the 32V supply at 0V. Ah2. R505 and R506, both being 680R, would hold the 32V rail at the mid-point between 32V and -28V, just about 2V if the SCR fires.

I'd suggest that the diode is faulty, and is pulling down the 32V series regulator voltage. If that is the case, the current must be (31-24.7)/680 = 9.3mA. So the power in the diode would be 24.7*9.2mW = 229mW. Not sure I'd expect it to get that hot, but...

Cpt., you are the man. You know so much! Yes, I got 680 resistance for those stripey resistors, too. I just got a starter diode kit from Amazon https://smile.amazon.com/gp/product/B01M7S2ROI/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1 and maybe the diode I need is in there! If not I will track it down. Aside from the heat on the ps, this receiver plays well and nothing else seems to get hot. I can't wait to swap out that diode and see if that helps things. Thank you, thank you, thank you. I feel like I asked for help with a bully and got The Road Warrior.

 

[birchoak]

Well I think most of that looks right.

Essentially, you have three AC supplies from the transformer
21/0/21
7
32

The first is the main power, and, when rectified by the bridge, will give you +/- (21-0.7)*sqrt(2) = 28.7VDC, so your measured 28.3V looks fine.

The second is the lamp drive: AC for the illumination, and 9VDC for the stereo lamp
(7-0.7)*sqrt(2) = 8.9, so, again, your 9.1V looks fine.

The last is used by the PSU board to generate a number of voltages, probably 32V and 13V
(31.7-0.7)*sqrt(2) = 43.8V, so your 42.4V is fine. This tallies with the 50V working voltage of C504 on the schematic.

TR501/D503/R504 on the scm is a regulator, dropping the 42.4V to 32V nominal, in your case 31V. D503 is marked as a 32V zener.

D505 on the scm is marked as a 13V zener, with R503 as a dropper R, and C509 as smoothing. Note with 16V working volts of C509. You're measuring 12.7V, which is fine.

The SCR gate is the same as the cathode, so it isn't being triggered (needs a pulse above K to rigger). The fact that scm D504 and R506 are the same voltage as SCR anode also shows the SCR is off; no drop across R506, so no current. The SCR cathode is connected to the main -ve supply.

Now, the diode numbering between the scm and your PCB are different. It looks like the correspondence is

scm pcb
501 501
502 502
503 508
504 509
505 507

So, the mystery is the purpose of D504/D509, and why it is getting hot.

It's not marked as a zener (and it would make no sense to have a 32V series regulator followed by a 24V zener).
It's marked as being reverse biased.
It's connected to the 32V series regulator via R505, a 3W 680R resistor.
Why has it only got 24V7 across it?

Since this node disappears off the scm, via pin 5, I don't know what it is meant to power. R505 is intended to be a current limit for when the SCR fires, I think.

Ah. If the SCR fires, and pulls its anode down to its cathode, that would pull the 32V supply down to -28V. D504/9 would then be forward biased, and hold the 32V supply at 0V. Ah2. R505 and R506, both being 680R, would hold the 32V rail at the mid-point between 32V and -28V, just about 2V if the SCR fires.

I'd suggest that the diode is faulty, and is pulling down the 32V series regulator voltage. If that is the case, the current must be (31-24.7)/680 = 9.3mA. So the power in the diode would be 24.7*9.2mW = 229mW. Not sure I'd expect it to get that hot, but...

Ok, I pulled D509 (or D504 on scm) and measured it as best I could:
Diode test on MM: 588
Ohms test on MM: 173.6 ohms (200K setting)
Cheap-ass* Kuman component tester: Uf 653mV, C=30pF

*Probably NOT very accurate!

You say D509/504 is reverse biased; do you buy the diode this way, or does that refer to the installation orientation? The scm labels D504 as 10D-1; when I googled "10D-1 diode" I got a hit at Mouser, which classified it as a zener diode, other sites call it a rectifier diode for power supplies. Would one of my shiny, new IN4007 rectifier diodes work in place of the misbehaving D509/504?

Edit: board silkscreening does not have that weird zener diode symbol, so I don't think it's a zener. Plus, you said it wasn't! Can you tell I'm trying to get a handle on this?

 

[audiodummy]

Googling 10D-1 seems to pull up a 1.5A 100PRV diode, I don't think a single 1N4007 (1A 1000PRV) will do normally. May need to plop two in parallel or two 1N4002 in parallel to be totally "equal or better". But if it was meant to always be reverse biased, it never conducts 1.5A!

However need to ensure it really is bad... if the device is working, don't muck with it if unsure and wait till it breaks?

BTW, I don't see the picture to the bottom of the board with the tracks...

 

[birchoak]

Scroll up on this page for pic of underside of this cursed ps board.

Testing would seem to indicate the suffering diode is working properly.
I thought we were close to solving this!

 

[audiodummy]

I see "[ I M G ]" where a picture normally would show, not sure why it's not working for me.

The wonder if this diode is leaking or not...also what are the markings on this diode?

 

[cpt_paranoia]

Yes, I got 680 resistance for those stripey resistors, too

The 680R I'm talking about are the big 3W devices, not the stripey ones.

You say D509/504 is reverse biased; do you buy the diode this way, or does that refer to the installation orientation?

It refers to the way it's used/installed. In normal operation, it won't be conducting; essentially, it's out of circuit. It's only forward biased if the SCR is conducting.

Googling 10D-1 seems to pull up a 1.5A 100PRV diode, I don't think a single 1N4007 (1A 1000PRV) will do normally

In the circuit configuration, the diode will only conduct via the 680R in series with the SCR, so, at most, it should only carry 28V/680 = 41mA.

 

[birchoak]

I see "[ I M G ]" where a picture normally would show, not sure why it's not working for me.

The wonder if this diode is leaking or not...also what are the markings on this diode?

I will post again; diode is black, with a thick white ring, then a thin white ring next to it at one end. Gotta go eat; we have company over

 

[birchoak]

Update to come after guests leave! Strange and stranger; stay tuned to this Bat channel!

 

[birchoak]

Okay, here is the underside of the power supply board, once again:




The company is still here but I'm on a bathroom break (allegedly; don't blow my cover).

Sooooo, last night I unsoldered every single cap on the ps, checked it for capacitance, checked the board location it came out of for polarity, then re-installed each cap. One at a time, in and out. Then, I saw a lifted pad at one of the cap's leads, so I scraped the conformal coating off the trace, tried to solder to that but it wasn't taking, so I soldered a jumper in. That may have been part of the problem; I don't know.

I replaced the MV-13 diodes last night with the suggested replacement diodes.


Today, I pulled out that diode, checked it (seems fine according to MM diode checker), put it back in. I checked the bias again; it had wandered a few millivolts up so I corrected each side back to 50mV.

THEN, on a whim, I hooked the temperature probe up to my Triplett MM (it has a temp. function) and checked the components again, after firing up the Concept 2.0 with several loud Tom Petty songs.
New temp readings with Triplett MM:
C512: 111˚F, measured at the top
C513: 86˚F "
D509: 86˚F at anode; 87˚ at cathode!
R507: 97˚F at top
R503 is the only other relatively hot thing on the PS now, at 161˚F.

What the what?

I still may rebuild the PS (beyond the full recap I did) at a further date, but for now I'm calling it good. The output transistors and new diodes (the old MV13's were glued to the heat sink) are all running cool to the touch. DC offsets are still the same, in the negative mV values, but that doesn't seem to be a big deal. Correct me on that. Music is clean and strong in both channels and FM stereo is now spot on, sounding fantastic. It is a very well-built receiver and sounds much bigger than 20 watts.

Thank you to all for your help; I really appreciate it. I think the fancy new infrared thermometer might be giving some false readings, although when it read a resistor as super hot and I touched it, I did get a nasty burn, so it's not WAY off. Maybe all the concentrated audio karma from you folks made it to my basement bench and put the cure on the hurt. I honestly don't know. Perhaps that diode looked properly soldered in but wasn't making full electrical connection, and simply taking it out and installing it with fresh solder did the trick. Maybe the PS will start running hot again tomorrow and I'll pull the rest of my hair out!

 

[birchoak]

The 680R I'm talking about are the big 3W devices, not the stripey ones.



It refers to the way it's used/installed. In normal operation, it won't be conducting; essentially, it's out of circuit. It's only forward biased if the SCR is conducting.



In the circuit configuration, the diode will only conduct via the 680R in series with the SCR, so, at most, it should only carry 28V/680 = 41mA.

I'm going to market a machine, invented by you, that you insert an entire amp or receiver into. You close a door, push one button, and after you hear a chime, you open the door and pull out a fully functioning, back-to-factory-spec machine. We will laugh all the way to the bank.

 

[audiodummy]

Yeah this doesn't make sense right now other than if that diode is simply leaky. Funny that it looks like some logo on it but no other identification. Thanks for reposting the copper layer, not sure why it didn't show up at first. Definitely looks very much like the schematic but wit different component designations.

That diode appears to be "insurance" that if the SCR needs to crowbar and cut power to the amp (speaker short/overload), the SCR will not need to do so with the full differential of the power supplies and only to GND hence "normally" if it needed to constantly pass full or derated current at 100% duty cycle.

Since the anode is 0V and cathode was measured positive (+28.3V) with respect to anode, the diode should be off and not dissipate power. The only possible explanation is the diode is a zener or leaky, but leakiness should show up in the component tester (if you have a cheap component tester, it will not detect the voltage of a 28V zener.)

 

[birchoak]

Wow, it felt weird to put the bottom cover back on this thing. I almost don't recognize it. My new AK friend Ron says that these were built by Tandy (although Tandy contracted out to Hitachi and others) and it definitely reminds me of an STA-2000 I had years ago, but better sounding and even better build quality. Tom Petty sang again and all the measurements are holding, and the output transistors are cool as clams at about 78˚F each. This machine was worth the trouble and I thank you all again for helping me. I'm sure it can be frustrating watching someone flounder and you aren't there to see for yourself what the problem is. It's like when your kid explains a fight at school to you--"OK, hmm--I wondered what really happened?" you think.

 

[birchoak]

Yeah this doesn't make sense right now other than if that diode is simply leaky. Funny that it looks like some logo on it but no other identification. Thanks for reposting the copper layer, not sure why it didn't show up at first. Definitely looks very much like the schematic but wit different component designations.

That diode appears to be "insurance" that if the SCR needs to crowbar and cut power to the amp (speaker short/overload), the SCR will not need to do so with the full differential of the power supplies and only to GND hence "normally" if it needed to constantly pass full or derated current at 100% duty cycle.

Since the anode is 0V and cathode was measured positive (+28.3V) with respect to anode, the diode should be off and not dissipate power. The only possible explanation is the diode is a zener or leaky, but leakiness should show up in the component tester (if you have a cheap component tester, it will not detect the voltage of a 28V zener.)

For a few bucks and an hour or two of my time, I think it would be wise to at least replace the diodes, the transistor, and the SCR on the PS. I've got all this time invested in it and it is truly a well-crafted instrument. I haven't even hooked it up to a good pair of speakers yet--all testing is on Lafayette Criterions I rescued from a dumpster about a year ago. For now, I can live with 161˚ on one resistor.

 

[cpt_paranoia]

Hot capacitors is very unusual, unless they are being heated by the devices around them.

I would replace that diode; it's possible that it is the cause of all your overheating.

 

[birchoak]

Hot capacitors is very unusual, unless they are being heated by the devices around them.

I would replace that diode; it's possible that it is the cause of all your overheating.

cpt, I will do this. What diode should I replace it with? IN4001? Two IN4007 in parallel as suggested by audiodummy? Whatever is best for this silver-haired gentleman (I refer to the receiver here). Aforementioned, I have no problems with rebuilding the entire PS for future reliability if that's what it takes.

Your handle = Captain Paranoia? Care to elaborate? Pure curiosity on my part. If you tell me, I will explain my bizarre avatar (although you're probably old like me and already get it).