Looking for transistor data: ETD41-035

[torch]

Short story: I'm looking for a datasheet for an ETD41-035, which I believe was manufactured by Fujitsu back in the 70's and 80's. I think it might be a Darlington and know it is an NPN in a TO-220 case.

Long story: These were used in a number of Yamaha motorcycle Transistor Controlled Igniters.   The ignition system supplies a constant 12v to the coils and the TCI provides a ground path through these transistors to fire the coil based on a signal from the respective magnetic pickup. There are 2 coils and therefore 2 transistors per TCI, firing on opposite strokes.

The TCIs are generally pretty reliable (biggest headache with these TCI units is cold solder joints). However, I am trying to repair one with a considerable mis-match between the two transistors. I cannot find any data about these on-line.

According to this Chinese vendor (http://www.aliexpress.com/store/product/ETD41-035-TRANSISTOR/107262_1098158355.html) the manufacturer is "YXS Technologies Inc.":


But while the photo is identical to the Yamaha TCI  component the vendor is calling it a voltage regulator! So I can't say that I trust any info from that vendor, and YXS seems to make cables, not semiconductors, so I suspect that this is one of those infamous Chinese re-labelled forgeries.

Some people report that they have successfully used a MJE5740 as a drop-in replacement in a similar TCI. So I tried a pair of those but with less than satisfactory results. I get a very weak spark compared to a known good TCI.

I back-probed the coil grounds on a running bike. Here is the signal from the known good TCI:



Note the wide pulse after the negative spike.

Now here is the "repaired" TCI swapped into the same bike:



The negative spike is gone and the positive pulse is virtually a spike, so the coil isn't fully discharging. 

Now I just need to figure out if it's the transistors, or something else in the circuit. FWIW, both halves of each TCI behave identically. I already replaced all the electrolytic caps, measured the resistances of all the discrete resistors and tested the 8.2v zener diode. There is a mystery IC with a daughter board full of laser-cut resistors on one side and SMD devices on the underside, but I think that just controls the timing advance curve, which is working properly.

[JimRemington]

The logo on the package is pretty distinctive and it may be quite helpful if you can find a match. There are a number of IC logo image collections (with respective manufacturers) on line, (e.g. http://www.electronicspoint.com/resources/logos-for-electronic-component-manufacturers.47/ ) -- looks like Fujitsu.

[torch]

Yeah, thanks for the confirmation.

Anyone have a book of old Fujitsu datasheets lying around gathering dust in the basement?

[Zarhi]

See if that helps: http://www.jetav8r.com/Vision/TCIRebuild/TCI_Rebuild.html

[wraper]

But while the photo is identical to the Yamaha TCI  component the vendor is calling it a voltage regulator! So I can't say that I trust any info from that vendor, and YXS seems to make cables, not semiconductors, so I suspect that this is one of those infamous Chinese re-labelled forgeries.

YXS is their store name. It's not uncommon on aliexpress that instead of brand, sellers just put their store name. This have nothing to do with the part being genuine or not.

[torch]

Yeah, I'm quite familiar with the Rider's of Vision article (Vision was a Yamaha motorcycle model). And the inside of that TCI is very, similar to the one I'm working on. Down to the value and number of most capacitors. That's why I tried their choice for a replacement. However, note that he has no specs on the original and mentions it was made by Fuji, not Fujitsu. Mine "works", sort of, in that I have successfully started the engine. But it was struggle getting it going with such a weak spark.

YXS is their store name. It's not uncommon on aliexpress that instead of brand, sellers just put their store name. This have nothing to do with the part being genuine or not.

I think you are quite possibly right about the store name being substituted for the brand name. That said, I wonder what the odds are that genuine NOS Fujitsu transistors that nobody has been able to source for years happened to end up in China? And without a data sheet, how would I go about verifying that they are the real deal?

[wraper]

I think you are quite possibly right about the store name being substituted for the brand name. That said, I wonder what the odds are that genuine NOS Fujitsu transistors that nobody has been able to source for years happened to end up in China? And without a data sheet, how would I go about verifying that they are the real deal?

Like this (if they are original fuji ones ):

[MagicSmoker]

Some people report that they have successfully used a MJE5740 as a drop-in replacement in a similar TCI. So I tried a pair of those but with less than satisfactory results. I get a very weak spark compared to a known good TCI.

Based on the difference in the scope shots it looks like the original transistor did not have an anti-parallel diode across the C-E junction and was designed to operate in reverse avalanche mode. The latter would also explain why the circuit seems to be very finicky about what kind of transistor is used.

Note that very few transistors are characterized for intentional avalanche mode operation (forward or reverse) but there are loads of Darlington NPN transistors without an anti-parallel diode and in a TO-220 case so I wouldn't abandon all hope; you do have some tinkering ahead of you, though.

[mikerj]

The MJE5740 is most definitely not a suitable part for a Kettering type ignition, the diode across the collector and emitter is dumping most of the coil energy.

There are numerous NPN darlingtons that are designed for just this application however, e.g. the ST901T is pretty cheap and available through Mouser/Digikey/Farnell etc.

[torch]

Oh, you can't always get what you want,
But if you try sometimes,
you just might find
You get what you need

Gotta love it. I ask for some info that may not even exist, so instead of ignoring my ignorance, people jump in here with really useful info appropriate to my problem Not only an explanation (that even I can understand) of what is going on, but a pointer to a part that probably will work instead.

Thanks guys, I appreciate it! 

[torch]

Comparing the datasheets, the ST901T seems a little less robust than the MJE5740 in terms of maximum voltages and currents (although with a higher power dissipation). However, since I don't have a datasheet for the original ETD41-035, it may well be just fine. I guess the only way to find out is to try them, so I ordered some.

And just for $!+$-and-giggles, I ordered some of the supposed original from that Chinese vendor. Maybe they really are surplus NOS genuine Fujitsu transistors.

If one or the other seems to work well in the TCI I'm repairing, then maybe I'll risk desoldering one of the originals from the known-good TCI and try comparing the three in some sort of test circuit. (In which case, I'll probably be back asking for advise in designing the test circuit  ).

[torch]

UPDATE:

The ST901T "works a treat" as Dave would say. In fact, it works better than the original. I forgot to save a screenshot from the scope, but the pulse is much broader and the spark produced is much stronger. The known-good unit craps out at about 1/4" (this is a waste-spark system, so that is in addition to the spark under compression in the opposite cylinder). With the MJE5740 installed, the spark barely jumped a 1/8" gap (in free air, so that would be less under compression. No wonder ran so poorly.) But equipped with the ST901T I had it up to 1/2"! (It was still strong, but I did not want to open the gap further for fear of stressing the coils and wires with the excessive voltage).

Thanks to all who chimed in, especially mikerj and MagicSmoker.

(If anyone ever does come across a datasheet for the ETD41-035, I'd still be curious to see it and compare to the ST901T one.)

[MagicSmoker]

Heh, I'm glad I was of some help here, but I demur to mikerj for providing a specific solution!

[torch]

Don't sell yourself short -- you told me why the other one didn't work and helped me learn something.

There are many forums on the net where people asking questions are met with derision. I appreciate that people here took the time to help me out. This forum is great that way.

[Seekonk]

Thanks for pointing to the ST901T, a good data sheet to save.  I had a Toyota Supra a long while back and the ignition module failed.  Went to the dealer and without even looking it up quoted like $200.  I thought to myself....... Hey, who won the war.  Tore it apart and everything was buried under some sticky silicone goop.  Noted my three points of interest and brought wires out.  That was touchy soldering under that goop.
Had a bunch of 700V small power NPNs.  I think the first effort used four in a darlington.  That ran for about 2 minutes before smoke.  Next used six and that ran for over 40K miles till I sold the car.  I know the next guy ran it another two years.

[unitedatoms]

There is archived 1987 article on Fuji site about ignition transistors: http://www.fujielectric.com/company/tech_archives/pdf/33-01/FER-33-01-28-1987.pdf

May be the closest device in table 1 is:

ETE41 in TO-3P package 450V 8A 80W 500HFE Darlington, speed unknown, may be in order of 0.5uS

After reading article, it looks like a precursor of IGBTs. The snubber circuit is named "snapper". May be secret of this ignition transistor is an embedded snubber R-C inside the plastic package ?

[torch]

There is archived 1987 article on Fuji site about ignition transistors: http://www.fujielectric.com/company/tech_archives/pdf/33-01/FER-33-01-28-1987.pdf

May be the closest device in table 1 is:

ETE41 in TO-3P package 450V 8A 80W 500HFE Darlington, speed unknown, may be in order of 0.5uS

After reading article, it looks like a precursor of IGBTs. The snubber circuit is named "snapper". May be secret of this ignition transistor is an embedded snubber R-C inside the plastic package ?

Funny you mention Fuji. The ETD41-035 arrived today. Different from the photo in the ad. This one bears a modern "Fe" Fuji mark instead of the mark identified as Fujitsu by several websites. That is now in question as, while exploring the Fuji website to see if they could have made this, I found Fuji claiming ownership of that mark on their website! They say they used that mark from 1923 to 1978.

However, I still couldn't find the datasheet 

Anyway, back to the transistor, the Chinese-sourced ones test similar on my $20 transistor tester (early Karl-Heinz type) compared to the 2 de-soldered from a known-good TCI. Hfe around 8. Forward voltage drop is a bit lower -- about 0.9 for these ones, about 1v for the originals. But looking very close at the face under the faint printing, I see fine striations. I wonder if it was sanded and reprinted. If so, the question becomes "by whom?"

[unitedatoms]

I think there is no direct replacement from contemporary parts, so it is impossible to get negative spike as on oscillogramm. The negative spike is possibly coming from diode + capacitor in "snapper" circuit. To make an equivalent module it should be high voltage darlington with 2 resistors + avalanche diode + protection slow diode + fast diode + capacitor + resistor. As on Fig 2 and Fig 9 in the paper.

[torch]

Actually, that ST901T comes pretty close:



For comparison, here is the original known-good one again:



(The bias is because I forgot to set the coupling the same when looking at the replacement. But the overall waveform is very similar).

[mikerj]

Sorry I missed your reply a few days back, it's great to know the replacement worked out ok 

One thing that's interesting is that the dwell time appears to be much shorter with the ST901T, about 2ms compared to ~10ms with the original.  Presumably this system has an active dwell control based on peak coil current, have you tried looking at coil current?

I'm just wondering if the driver stage before the darlington maybe wasn't providing enough base current to saturate the original transistor, which could both explain it's failure (increased power dissipation) and the longer dwell.  The much newer ST901T probably has a higher current gain than the original and requires less base drive.

The RPM is a fair bit higher on the ST901T trace though, so maybe the system (12v) voltage was higher leading to lower dwell time?

[sanchel]

Здраствуйте!Несмотря на то, что тема старая, хотел попросить у Вас совет по ремонту TID14-53 для мотоцикла YAMAXA.У меня такая же проблема - слабая искра на низких оборотах и байк очень трудно завести.Покупал эти блоки зажигания через интернет , но к сожалению (жулики) прислали нерабочие.Теперь у меня уже три нерабочих и у всех одинаковая неисправность - очень слабая искра.Отдавал в ремонт разным мастерам но пытливого как Вы среди них не нашлось.Все разводят руками ссылаясь на неисправность микропроцессора в блоке (который невозможно достать).Я считаю что дело не в нем так как после запуска байк работает нормально.Лишь на малых оборотах искра слабая.Так же при работе на холостом ходу двигатель сильно греется - думаю из-за нестабильной искры.Все (датчик,катушки,провода и даже двигатель )кроме блока зажигания менял.Так как мне деваться некуда и помочь некому, байк простаивает,пытаюсь починить блок сам.Познания в электронике -минимальные.Пытался найти у нас ST901T -  бесполезно,про оригинал ETD41-035 вообще молчу.Так как Вы проблему "плохой искры" в аналогичном блоке решали -прошу помощи!Может подскажете какую из распространённых деталей вместо ST901T подойдёт,или может какая то ещё деталь в этом блоке может влиять на силу искры!?Заранее благодарен!

[wraper]

Ну так заказывайте у китайцев на алике или где нибудь ещё ST901T, да его можно и в ЧИП и ДИП заказать например.
https://www.chipdip.ru/product1/8497424629

[torch]

Здраствуйте!Несмотря на то, что тема старая, хотел попросить у Вас совет по ремонту TID14-53 для мотоцикла YAMAXA.У меня такая же проблема

Google translate tells me you want to know more about fixing the Yamaha TCI. I recommend you see http://xjcd.org/TCI_repair for more detailed information, pictures and part numbers.

I have repaired several now and they are all still working well, after years of use.

One other thing: make sure you have good clean connections throughout the system: fuse terminals, ignition and kill switches, ignition relay and all grounds. Make sure the battery is strong. This Yamaha ignition system is sensitive to low voltage conditions and will have weak spark if corroded connections are reducing the primary voltage to the coils.