Transistor Cross Reference help - Yamaha DX7S switching power supply

[grantb5]

I'm pretty far out of my element here, but got a deal on a dead Yamaha DX7S synthesizer (USA version) and would like to at least try and fix it. I'm in the US, so I'm not converting, just fixing hopefully.

It has a switching power supply from about the late 80's. I've located some dead components and there is one transistor here that I can't find. It's a 2SC3570. The original is in a plastic TO-220 package.  On another synth that uses the same power supply (DX7II) the service manual says 2SC3570 or 2SC2929. They are similar, but to my untrained eyes, definite differences. I'll start by showing the schematic and parts list:

The part in question is Q1 to the left of J4.


Layout (US and Canadian versions, very similar):




Schematic:
http://yates.ca/dx7/Schematics%20&%20PCBs/DX7IIFD%20Schematic.jpg

I can't find the original anywhere (2SC3570), at least on this side of the pond. The NTE cross for this and the alternate come up with completely difference numbers, so I'm not 100% confident that the alternate (2SC2929) is a good idea.

I'll post the transistor info in a followup message. Let me know what you think. Or what parameters I should focus on.

[jmelson]

Well, this doesn't look very exotic.  I think just about any 250 V 5A transistor likely would work fine.  Seems the spec is a 5A, 400 V transistor, so that gives plenty of margin.  You ought to be able to use any number of 2N type switching power transistors with similar ratings to substitute.

Jon

[grantb5]

Here is a link to info on the 2SC3570:

https://alltransistors.com/transistor.php?transistor=14906

And the 2SC2929:

https://alltransistors.com/transistor.php?transistor=13897

Comparing:

Code: [Select]

Type Designator:                              2SC3570        2SC2929
Material of Transistor:                       Si             Si
Polarity:                                     NPN            NPN
Maximum Collector Power Dissipation (Pc):     25 W           40 W
Maximum Collector-Base Voltage |Vcb|:         500 V          450 V
Maximum Collector-Emitter Voltage |Vce|:      400 V
Maximum Emitter-Base Voltage |Veb|:           6 V            7 V
Maximum Collector Current |Ic max|:           5 A            3 A
Max. Operating Junction Temperature (Tj):     125 °C         150 °C
Forward Current Transfer Ratio (hFE), MIN:    40             20
Package:                                      TO220            TO220

I can find the 2SC2929 but the case might be metal tab instead of plastic ... eBay sellers are a bit vague in the descriptions. Do you think this cross is acceptable? Or should I try and find something closer to the original 2SC3570?

Thanks.

[grantb5]

You're going by the fuse and the maximum collector current? Is it true that since this is a switching application, the different hfe isn't a concern?

Here is what NTE comes up with ( http://www.nteinc.com/search.php ). I know they are not universally loved. Curiously that they are not the same NTE cross:

2SC3570    NTE2337  http://www.nteinc.com/specs/2300to2399/pdf/nte2337.pdf
2SC2929    NTE51      http://www.nteinc.com/specs/10to99/pdf/nte51.pdf

I can break out the specs for these too. People are saying that NTE isn't really great at adhering to their specs though. I'm certainly not one to say.

Actually I can't break out the specs. The two NTE pages have different abbreviations for almost everything and I have no confidence of transposing the correct numbers.

[Jr460]

Just wondering did you find any of R2,  R3, R9 open?

[grantb5]

R2 and R3 both open, I'll have to recheck R9 since you've asked. No visible signs of failure though. Q1 and Q2 both toast.   

I just did a Yamaha SPX90 power supply fix (I got suckered twice by the "but it's so cheap! I have to buy it!" bug) and that one is somewhat similar, but the failure was all electrolytics that spilled their guts and the glue on the PCB, it ate one film cap and one resistor. I just had to replace passives in that one and built up the confidence for this task. Kind of.

[floobydust]

I would consider FJPF13007 H2 $1.39 as a replacement.
NTE parts can be confusing because they will use a 500V transistor to cover all the 300-400V parts as well. Note the hFE is lower for NTE2337 compared to 2SC3570.
Non-isolated TO-220 parts will generate more EMI as the heatsink will have high-voltage switching pulses present, and you need to ensure HV spacings to nearby parts is enough.
I would stay with the fullpack, and use fusible resistors where they are spec'd.

[grantb5]

Thanks. I never really gave a second thought to "fullpack". I knew the original TO-220 was "isolated" (insulated) but didn't connect the fullpack dots. (It didn't help that I was sourcing a PCB from Japan hours earlier and the guy there referred to it as "loose pack" which I guess is his translation for PCB-only.) But interesting to know about the EMI. The pads for the heatsink are unconnected so any regular TO-220 I would have just screwed right in there without a second thought.  I will check out that part for sure. 

The service manual calls R2 and R3 metal oxide film and those are obtainable. As far as I know only those two resistors, plus Q1 and Q2 are fried. I'll probably recap the electrolytics, which isn't my thing, but after my SPX90 fix, I think it's warranted for <$20. All of them measure OK on a simple meter. I don't think this thing died a natural death. The power switch was missing and a small toggle switch was duct-taped in instead. Plus another PCB was crushed. Those have been replaced/restored.

[vk6zgo]

Thanks. I never really gave a second thought to "fullpack". I knew the original TO-220 was "isolated" (insulated) but didn't connect the fullpack dots. (It didn't help that I was sourcing a PCB from Japan hours earlier and the guy there referred to it as "loose pack" which I guess is his translation for PCB-only.) But interesting to know about the EMI. The pads for the heatsink are unconnected so any regular TO-220 I would have just screwed right in there without a second thought.  I will check out that part for sure. 

The service manual calls R2 and R3 metal oxide film and those are obtainable. As far as I know only those two resistors, plus Q1 and Q2 are fried. I'll probably recap the electrolytics, which isn't my thing, but after my SPX90 fix, I think it's warranted for <$20. All of them measure OK on a simple meter. I don't think this thing died a natural death. The power switch was missing and a small toggle switch was duct-taped in instead. Plus another PCB was crushed. Those have been replaced/restored.

The switching transistors in most BJT SMPS are fairly generic.
I routinely replaced the "2S" series devices in the PSU of Sony KX27 "Profeel"picture monitors with BU208D devices.
It helped that I could use them to replace the horizontal output transistors as well.

If the switching transistor has failed, I would be quite suspicious of transformer T1.
In one of the Profeels, the transistor initially failed short circuit, but as the thing was slung up above head height, & had no input, nobody noticed.

It sat there for weeks till somebody tried to use it, giving the output of the rectifier bridge plenty of time to "cook" the transformer.
It "looked" OK, but had shorted turns.
After rebuilding the supply, it went "bang" upon initial turn on!

[grantb5]

Thanks for the tip. Not sure I can asses that part and certainly cannot find the original. If it  dies again I may try and wedge some other power supply in.