PDA

View Full Version : OT:electronics cross reference resource?



jacampb2
07-12-2008, 11:04 PM
Sorry for all the OT BS out of me lately, but I know we have people on here into everything, and this is just the best place to ask some of this stuff!

Are there any good component cross reference resources available online, preferably something w/o a membership fee. Or, are there book for references I can purchase? I don't necessarily want the information for free, but I would rather not have to pay $10 a month for something I wont use more than a few times a year. A catalog guide would be optimal IMO.

FWIW, I have never had too much trouble tracking components, or replacement components down, but I have some complimentary transistor pairs out of an audio amplifier that need replacement, and I am striking out everywhere. They were initially made by Rohm, part number 2sc5248 (actually labeled c5248) and part number 2sa1964 (labeled a1964). They are BJT pairs in TO-220FP case. So far, I have only been able to find one data sheet for them, and it is more of a catalog description than a data sheet. It includes only major specs, and no SOA or any other chart data. That is making it difficult to match them up with newer components. Rohm does not list them on their internet resources.

I can find similar rated transistors out there, but these have a Dc gain bandwidth of 150Mhz and the closest stuff I can find in similar case style and specifications typically has a much lower bandwidth.

I found a few of the original transistors on ebay for a premium price, but I would rather cross them with a currently available pair, thus the need for a good cross reference. I have tried the cross reference resources at many popular semiconductor manufacturers websites, but they all act as if the devices don't exist... This is frustrating. I have been searching for 3 days. Is there a easier way?

Thanks,
Jason

Evan
07-12-2008, 11:09 PM
I suspect you don't need 150 mhz bandwidth in an audio amplifier.

I'll look it up in some of my old data books after supper.

Bmyers
07-12-2008, 11:12 PM
try here:
http://nte01.nteinc.com/nte/NTExRefSemiProd.nsf/$$Search?OpenForm

jacampb2
07-12-2008, 11:20 PM
Yes, I suspect so as well, but I am not comfortable enough yet with the design to understand why the manufacturer went with these particular components. These are the "pre-driver" pairs in the OPS of a triple Darlington output configuration. The actual output devices are 30Mhz bandwidth, which still seems unnecessarily high for a audio amp. I am reading a few good books on amplifier design theory right now, and from what I understand, significant deviations in bandwidth from the original design can result if instability problems, so I am not very comfortable making a seat of the pants guess on what is going to be acceptable here.

Thanks,
Jason

jacampb2
07-12-2008, 11:22 PM
try here:
http://nte01.nteinc.com/nte/NTExRefSemiProd.nsf/$$Search?OpenForm

Thanks, but I ran across that site before, and no dice, unfortunately.

Thanks,
Jason

Evan
07-12-2008, 11:48 PM
I have three different transistor databooks but none of them list either of those part numbers. Sorry.

BTW, substituting a device with a lower gain/bandwidth product doesn't usually pose a problem. It when you sub a part with a higher gain product that spurious oscillation may occur.

jacampb2
07-13-2008, 12:15 AM
Well, thanks for looking. Maybe it is just a very rare bird. I really wish I could find a full data sheet for it, it would make me a lot more comfortable finding a replacement.

Oddly enough, I have I think 8 of this manufacturers amps in 5 different models spanning about 11 years of their product, from 1989 to about 2000. All of the models use this same pair in the pre-drivers. I am beginning to think maybe they bought a (the only???) truck load of them, and just used them. These are from a fairly high end manufacturer, so I would hope that the components were chosen for the design, and not the other way around, but who knows, almost any corporation will do whatever they can to save a buck.

Thanks,
Jason

J Tiers
07-13-2008, 01:30 AM
It is possible that they DID have the only ones. We did for a different part.

We used a paricular pair for pre-drivers on an audio amp at my former work..... then we got a notice that they had been discontinued. it turns out that we were, literally, the only people to order quantity of them.

TWO manufacturers, Motorola (now "ON SEMI") and Texas Instruments BOTH discontinued the parts on almost the exact same day. It was very annoying, since those parts were the best for the design, and when I had to re-design for a different part, it wasn't nearly so satisfactory.

While is SEEMS silly to have 150 mHz or 30 mHz parts in an audio amp, it may NOT be.

For one thing, there are different frequency limits for different usages. The straight F(t) is for a particular test. It is generally the "transition" frequency (often predicted, not measured) where the beta is one (no current gain).

The direct relevance of that to a particular circuit depends on how the part is used. If it is an emitter follower, it MAY be directly relevant. If the circuit is a common emitter circuit, the F(t) may be less relevant except for how that affects the various parameters that control frequency response in a CE circuit. You won't get response remotely close to F(t) in most circuits.

A 30 mHz part may (generally is) capable of better response to transients and can be used to reduce such effects as "transient intermodulation distortion", or slew-induced distortion.

Since the actual frequency response of bipolar transistors tends to be affected by the current through the part, a power transistor, which has a very large variation of current, will have less intermodulation distortion when its gain is essentially independent of load current over the full range of frequencies in the passband.

A transistor with a high F(t) will tend to have less effect of current on response at frequencies much less than the F(t).

As far as stability, a transistor with much lower F(t) ABSOLUTELY can reduce stability. It produces a much lower frequency pole in the gain, and may give enough phase change to upset a frequency compensation that "expects" a much higher pole.

Of course, it MAY not have any effect. it all depends on whether the parts in question were chosen for that parameter, AND whether it is necessary, AND whether, necessary or not, the higher F(t) is "used" or depended upon in the design.

I have not investigated the following suggested crosses.....

Cross Reference Tool :: 2SC5248
NOTE: Our cross reference tool is for suggestion purposes only.
Manufacturer Part Similar Part / Replacement
2SC5248 2SC3298/A find 2SC3298/A datasheet
2SC5248 2SC3364 find 2SC3364 datasheet
2SC5248 2SC4159 find 2SC4159 datasheet
2SC5248 2SC4370 find 2SC4370 datasheet

OTE: Our cross reference tool is for suggestion purposes only.
Manufacturer Part Similar Part / Replacement
2SA1964 2SA1306/A find 2SA1306/A datasheet
2SA1964 2SA1332 find 2SA1332 datasheet
2SA1964 2SA1606 find 2SA1606 datasheet


http://ezinearticles.com/?Understanding-Transistor-Data-and-Replacement&id=38000

About 80 cents apiece:
http://www.newark.com/jsp/displayProduct.jsp?sku=85C5251&CMP=KNC-G10000679&HBX_OU=50&HBX_PK=2SC5248

jacampb2
07-13-2008, 01:59 AM
Thanks. That link you provided seems interesting. I am going to do some reading in a minute.

It does look like the OPS is a EF design. This amp has an identical output stage to the one that you (Jerry) looked over for me, just fewer parallel output devices and it is a stereo amplifier with lower rail voltages. Everything from the pre-drivers on is identical.

The other exception is, this one is actually broken. It still works, but with some pretty bad distortion. DC offset is about 60mVdc per channel, it has a fairly bad "turn off thump", and likes to break out into a self sustaining 3Hz oscillation. I have verified that it's power supply, and input stage are working correctly, and now I am trying to replace the pre-drivers, mostly based on my research of high DC offset problems. Most advice says to start there.

Hopefully I will turn up a suitable replacement part. I am fairly certain that I can still get them from the OM of the amp, and it is likely to be a fair price (they have been good to me), but I would just as soon find a current replacement, as I have a lot of these amps and would rather have a long term solution to problems that may arise than purchasing discontinued parts. The good news is that the Toshiba output pairs are still in full production :)

Later,
Jason

jacampb2
07-13-2008, 02:03 AM
I have not investigated the following suggested crosses.....

Cross Reference Tool :: 2SC5248
NOTE: Our cross reference tool is for suggestion purposes only.
Manufacturer Part Similar Part / Replacement
2SC5248 2SC3298/A find 2SC3298/A datasheet
2SC5248 2SC3364 find 2SC3364 datasheet
2SC5248 2SC4159 find 2SC4159 datasheet
2SC5248 2SC4370 find 2SC4370 datasheet

OTE: Our cross reference tool is for suggestion purposes only.
Manufacturer Part Similar Part / Replacement
2SA1964 2SA1306/A find 2SA1306/A datasheet
2SA1964 2SA1332 find 2SA1332 datasheet
2SA1964 2SA1606 find 2SA1606 datasheet


http://ezinearticles.com/?Understanding-Transistor-Data-and-Replacement&id=38000

About 80 cents apiece:
http://www.newark.com/jsp/displayProduct.jsp?sku=85C5251&CMP=KNC-G10000679&HBX_OU=50&HBX_PK=2SC5248


Awesome! You always get me with your edits man! I will read up on them tonight. Thank you so much. That is more than I ever hoped for. What kind of service do you use? Is it available to people outside "the industry".

Thanks again!
Jason

J Tiers
07-13-2008, 11:31 AM
I used Google, and searched thru a couple pages of hits each time. I may have used the advanced search option with some "modifiers", (with XXX, without YYY, etc)

No special services.

I agree, the only datasheet for both parts I was able to locate was a very bare one, suggesting that the parts were even then old and "not being pushed".....

jacampb2
07-13-2008, 08:44 PM
Well, the most I managed to get out of google was about 3 pages of hits. Most of them links to chi-com pieces of unknown quality.

I had already found that MCM electronics (the supplier that Newark is showing has them) has the 2sc5248 in stock, but the compliment is not available. I found another place that list's the parts, Audio Lab of Georgia, but they show 0 in stock. I have a email in to them asking if they are still available.

As far as I could tell, the cross reference part numbers didn't pan out either. It appears that most of those are discoed as well. I found one that shows some promise, but the fT is 100Mhz, with an Ic of about 1.2a instead of 1.5a. I will likely just buy the original parts from the manufacturer and hope they hold together for another 10 years or so.

Later,
Jason

J Tiers
07-13-2008, 11:13 PM
BTW, with DC offset problems, while drivers etc can be the issue, it is well to look around.

Feedback SHOULD drive offset to near zero, and 60 mV might qualify as near zero, unless the service info says different.

If feedback fails to zero the output sensibly, then there is usually a GROSS problem, and you can make measurements to find out.

Look at voltages on bases starting at the first drivers. BEFORE the problem, you will see the feedback 'trying to correct" it, pulling the bases in the opposite direction. IOW the problem will be shown as OPPOSITE. If the output offset is positive, then BEFORE the bad parts, the offset will usually be NEGATIVE.

AFTER the problem, you will see the residual un-canceled offset, i.e. the basic offset problem.

That may also work on the last voltage amp stages, you will see the "direction" of the offset (if any) at that point. it is either MAKING the offset, or "trying to fix it".

The 3 hz issue sounds much like a filter issue, or possibly a problem with the SMPS portion....... a;though you DO say that the PS is OK...

Usually the feedback has a capacitor to ground at the "minus input". A problem with that may cause offset, and also may cause instability.

Leaky capacitors at that point can cause offset, and ones with faulty values can cause instability.