Snapped the lead on SM-150-01 biasing diode, ugh. Looking for replacement with proper voltage drop. How low?

saabracer23

saabracer23

Super Member
I’m working on a Marantz 2270 and as I was putting one of the channels back together and the lead of the biasing diode (H761) which is a SM-150-01 snapped off, right where it meets the body of course lol. I did some searching and found that the most commonly recommended replacement is the 1N4002 or higher. Then I found a couple of comments where people had tried the 1N400X and couldn’t get the amplifiers to bias properly, they found the voltage drop to be too high. They found some with lower forward voltage and then got it to work.

It was said that the SM-150 has a voltage drop of .557 and the 1N4001 has a voltage drop of .558 so should work (not sure if the 4001 is rated high enough, I would have to power it up to see the voltage across it). I removed the one good SM-150, let it cool completely and then measured it and oddly enough it measured exactly .558V on my Fluke meter. All of the 1N4004 and 1N4007 measure closer to .6V, most being around mid .590s.

I went through my diodes and found that the faster diodes like UF4004 and MUR120 measured in the high .490s, will go lower create an issue? Maybe that’s too low. I just want to avoid putting it all together and getting the new diodes mounted to the heat sink only to find I have to tear it all apart again because they don’t work.

I’m going to keep looking, I may be onto something that will work though, I have some 1N5393s, both in smaller cased Onsemi that measure too high, but some larger older ones made by Fairchild that measure pretty spot on, .557v. When increasing the current though using a Peak Atlas DCA75 the original SM-150 measured .632v at 5mV and the 1N5393 .643v at 5 mV. The 1N400X are up around .678v, so that much higher.

So basically I’m trying to find out if going with a lower forward voltage will work okay as I have hundreds that measure lower, but quite a bit lower. Or should I just use the Fairchilds, I only have a few of them, but have hundreds of the Onsemi where the forward voltage isn’t critical. Or am I overthinking this and the higher 1N400X will work absolutely fine in the circuit?

H760, the biasing transistor doesn’t need swapped as it’s already a TO126, a 2SC496.

Also, stupid question, using these new diodes, how critical is it to have it mounted to the heat sink? I know that as it heats the forward voltage drops. I may just use flexible multi strand wired to connect it to the board, that way this doesn’t happen again.

Dan
 
Last edited:
Register to hide this ad
Lower is safer, makes the base of the bias transistor more positive and cooler biased output stage. There's only a couple of volts there, if that, and little current. Did 1N414x come up in your search? Is the diode normally on the heatsink? Best to not change that.

Craig
 
llwhtt said:
Lower is safer, makes the base of the bias transistor more positive and cooler biased output stage. There's only a couple of volts there, if that, and little current. Did 1N414x come up in your search? Is the diode normally on the heatsink? Best to not change that.

Craig
Click to expand...

Yeah I figured the voltage and current had to be low, at least under 100v, the SM-150 looks to be a 100v 150mA device, so the 1N4148 did come up in conversation, but the measured forward voltage is about the same as the 1N400X, just shy of .6V. It is mounted to the heatsink, I won’t change it.

Seems like a shame to use a 1.5A device, but so far that’s the closest I’ve got. When you say lower is better, how much lower would you say is okay? Going from .558 to.495 work okay? Or stick with the 1N5393? I only have about 10 of the ones measuring lower, and several hundred of the UF/MUR devices in the .490v range. But no biggie if I need to use the one I have less of, gotta use them sometime, but saving them is always a plus as I’m sure you know.

Dan
 
Last edited:
What I like to do when checking out an amplifier is to set the idle current according to the Service Manual then get the heatsink nice and hot, 1/3 of rated power will get the heatsink the hottest in an AB Class amp. Then kill the signal and check the idle current again while the heat sink is hot, it will either be higher or lower than the initial setting. Ideally the current would be the same but rarely happens. If higher the bias circuit is undercompensating, if lower it's overcompensating. Overcompensating is safer but usually it's undercompensated. The heatsink mounted bias transistor and diode will conduct more when heated trying to keep the bias the same.

Check the stock channel to see how it reacts then compare to the channel with the replaced diode. Hopefully they're close. This should tell you how the new diode is working.
 
Be careful as most biasing type diodes are either 2, 3, or 4 small signal diodes in series, not a single diode. You can use an ald analog multimeter to see the difference. measure one of the ones intact, or the bad one if you can get to the broken piece of lead. I ould expect it to be a 2 or 3 diode in series.
Not sure if it is held to the deat sink via a screw like a STV-2H or STV-3H (Radial types), or is it an axial rectangular diode held by a clamp.
While the voltage reading you are getting appears to be a voltage drop across a single resistor, I have never seen bias diodes read as a single diode when using an analog multimeter. And bias diodes also seem to have temperature change properties that cant be duplicated with putting 2 or 3 diodes in series.
 
Tom B said:
Be careful as most biasing type diodes are either 2, 3, or 4 small signal diodes in series, not a single diode. You can use an ald analog multimeter to see the difference. measure one of the ones intact, or the bad one if you can get to the broken piece of lead. I ould expect it to be a 2 or 3 diode in series.
Not sure if it is held to the deat sink via a screw like a STV-2H or STV-3H (Radial types), or is it an axial rectangular diode held by a clamp.
While the voltage reading you are getting appears to be a voltage drop across a single resistor, I have never seen bias diodes read as a single diode when using an analog multimeter. And bias diodes also seem to have temperature change properties that cant be duplicated with putting 2 or 3 diodes in series.
Click to expand...

It’s a rectangular type that is mounted using a clamp. When I saw it I immediately figured that I would need to wired 2, 3, or possibly 4 diodes in series. It’s marked S-05.
IMG_7403.jpeg

In the manual they have it as a SM-150-01 which I found to be a 100v 50mA device so most anything should be able to go in. The voltage drop I am measuring is the diode on its own out of circuit.
IMG_7402.jpeg

I figured if it needed multiples in series it would measure over a volt forward voltage. The one in the photo is clearly the good one pulled from the other channel.

I went through more diodes and found a couple more that look like they’d work well unless I’m told otherwise, but I have some 1N4150TAP (50v 150mA) and some others labeled at NDP-590. If anyone has any info on those diode that would be incredible. I have a thread over at DIYAudio just hoping to at lease get a voltage and current rating. The 1N4150 measure about .565 and the NDP-590 measure between .550 and .562 from the 20 I measured. So pretty much perfect. Sounds like that position has at most a couple of volts and current is very little.

Dan
 
Dang, it took me awhile to figure it out as it doesn’t pop up on a google search, but was able to decipher down to the manufacturer. Then finding a list of parts that manufacturer has made I was at least able to find acknowledgment of the part being made. Couldn’t find a spec sheet, but with minimal info provided I can see that the NDP590 is basically a 70v 200mA diode with a peak forward current of 400mA. So basically a slightly lower voltage 1N4148 that has a slightly higher peak current rating. Since they measure in the .556 range it should make for a perfect replacement, as long as I can get it reliably mounted to the heatsink.

IMG_7405.jpeg

Dan
 
That "5961" federal stock class number is a US military number. 5961 is solid state and 5960 is vacuum tube. Probably need a Top Secret clearance and/or a 2nd mortgage to buy one.:)
 
you can check the curve before installing with some heat to see if close to an original then test in circuit for real life values .
 
saabracer23 said:
I’m working on a Marantz 2270 and as I was putting one of the channels back together and the lead of the biasing diode (H761) which is a SM-150-01 snapped off, right where it meets the body of course lol. I did some searching and found that the most commonly recommended replacement is the 1N4002 or higher. Then I found a couple of comments where people had tried the 1N400X and couldn’t get the amplifiers to bias properly, they found the voltage drop to be too high. They found some with lower forward voltage and then got it to work.

It was said that the SM-150 has a voltage drop of .557 and the 1N4001 has a voltage drop of .558 so should work (not sure if the 4001 is rated high enough, I would have to power it up to see the voltage across it). I removed the one good SM-150, let it cool completely and then measured it and oddly enough it measured exactly .558V on my Fluke meter. All of the 1N4004 and 1N4007 measure closer to .6V, most being around mid .590s.

I went through my diodes and found that the faster diodes like UF4004 and MUR120 measured in the high .490s, will go lower create an issue? Maybe that’s too low. I just want to avoid putting it all together and getting the new diodes mounted to the heat sink only to find I have to tear it all apart again because they don’t work.

I’m going to keep looking, I may be onto something that will work though, I have some 1N5393s, both in smaller cased Onsemi that measure too high, but some larger older ones made by Fairchild that measure pretty spot on, .557v. When increasing the current though using a Peak Atlas DCA75 the original SM-150 measured .632v at 5mV and the 1N5393 .643v at 5 mV. The 1N400X are up around .678v, so that much higher.

So basically I’m trying to find out if going with a lower forward voltage will work okay as I have hundreds that measure lower, but quite a bit lower. Or should I just use the Fairchilds, I only have a few of them, but have hundreds of the Onsemi where the forward voltage isn’t critical. Or am I overthinking this and the higher 1N400X will work absolutely fine in the circuit?

H760, the biasing transistor doesn’t need swapped as it’s already a TO126, a 2SC496.

Also, stupid question, using these new diodes, how critical is it to have it mounted to the heat sink? I know that as it heats the forward voltage drops. I may just use flexible multi strand wired to connect it to the board, that way this doesn’t happen again.

Dan
Click to expand...
I have the reverse problem. Idle current is too low on my rebuild - 0.4-1.0mV. I wonder if changing the bias diode to 1N4007 would help?
 
You must log in or register to reply here.
Back
Top Bottom