Featured
Latest
BC148, BF194 and th...
 
Share:
Notifications
Clear all

Forum 141

BC148, BF194 and their little Lockfit Friends...

19 Posts
11 Users
0 Likes
5,303 Views
Alastair
(@alastair)
Posts: 294
Reputable Member Registered
Topic starter
 

I'm planning on another session with my Philips N1500 video in the next couple of weeks.

Last time I played around with it ended up with me chasing various faults and replacing 11 faulty Lockfit transistors of various types--mainly on the Servo board, with a couple on the video/chroma mainboard.
--At the time, I used BC337/BC327 types or BC108, but I'm sure there's a more appropriate type for some I replaced....

Like--for the Video chroma stages the machine seems full of the low-noise signal amp type BF194.......

What would you guys use in place of BC148 and BF194 and any other Lockfit types you can think of....? :thumb

Any idea Why these types seem to fail, while other case/package types survive OK?

Cheers for any guidance....

 
Posted : 05/12/2014 4:40 pm
Tazman1966
(@tazman1966)
Posts: 382
Reputable Member Registered
 

Hi there.

For BC148 I used the ever popular BC108 - same transistor, different package. As far as the BF194 is concerned, I managed to source some brand new old stock ones which so far have been okay...

Tas

 
Posted : 05/12/2014 5:00 pm
Refugee
(@refugee)
Posts: 4055
Noble Member Registered
 

I have had a look in an old Mullard book and they have TO92 versions.
BF241 and BF324.
http://alltransistors.com/transistor.ph ... stor=26707
http://alltransistors.com/transistor.ph ... stor=26851

 
Posted : 05/12/2014 5:13 pm
Terrykc
(@terrykc)
Posts: 4005
Member Rest in Peace
 

When Lockfit transistors first appeared, the BC147/8/9 were definitely based on the BC107/8/9 series - in fact, I think some of the early Mullard application circuits had previously been released using the BC107/8/9 range ...

I think the BF194 is derived from the BF184, so I suppose that would apply to all the BF19x series ...

However, I'm sure that there is a wide variety of suitable alternatives out there.

For the BF184, http://alltransistors.com/transistor.ph ... stor=26631 points to BF179 , BF179A , BF179B , BF179C , BF180 , BF181 , BF182 , BF183 , TIP122 , BF185 , BF186 , BF187 , BF188 , BF189 , BF194 , BF194A , BF194B, whilst for the BF194, the list is BF182 , BF183 , BF184 , BF185 , BF186 , BF187 , BF188 , BF189 , A733 , BF194A , BF194B , BF195 , BF195C , BF195D , BF196 , BF196-01 , BF196P , so all very much of a muchness ...

When all else fails, read the instructions

 
Posted : 05/12/2014 7:03 pm
Kalee20
(@kalee20)
Posts: 256
Reputable Member Registered
 

Any idea Why these types seem to fail, while other case/package types survive OK?

I'd be interested in this, too. On the face of it, there is no difference between the Lockfit package and the TO92 package, apart from formed leadouts and a more generous helping of epoxy encapsulant.

So why do Lockfits give up?

 
Posted : 05/12/2014 8:17 pm
crustytv
(@crustytv)
Posts: 11869
Vrat Founder Admin
 

I've not the years of experience you guys have, I would ask do they fail as badly as is made out?

All my colour hybrid and all solid state sets are running with a huge compliment of these lockfits in place and functioning fine. Some of those are 47 years old, as you know I've also worked on some right pig sets that have been stored in less than ideal conditions, I've not had to blanket replace them, most if not all work just fine. The ones that were replaced were due to rusted through lead-outs.

That's not to say I've not had to replace one or two but its been just that, one or two. I've certainly not found them to be as bad as is suggested, I've probably replaced as many T092 packages as I have lockfit.

I have plenty of NOS lockfit and have no hesitation to fit them as replacements and will continue to do so.

Can only speak as I find, maybe I've just been lucky, if so its a huge coincidence that 20+ sets all seem to have reasonable lockfits installed and a failure rate comparable to T092.

CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
Crustys Youtube Channel: My stuff
Crusty's 70s Lounge: Take a peek

 
Posted : 05/12/2014 8:27 pm
Mark Hennessy
(@markh)
Posts: 410
Reputable Member Registered
 

Since I started actively restoring vintage radios, I have changed a *lot* more Lockfit types than anything else. There's no doubt in my mind that they are less reliable than the equivalent TO18 or TO92 types.

I think it's useful to consider the failure modes, and how it affects the circuit in question. Because basically, you could have a whole load of degraded Lockfit transistors, but not know about them.

For the BC147/8/9 types, the failure mode is much-increased noise. It's a very distinctive type of noise - not the hissing "white noise" that we are used to from amplifiers in general. Rather, it's an impulsive, low-frequency type of noise. It's almost certainly a type of noise called "shot noise".

But, would this be noticeable? It depends entirely on the circuit. For illustration, let's take the audio amplifier circuit used by Hacker in many of their sets - the A205 PCB as used in the RP25, RP35 and RP37. The same basic circuit was used in the RP38A, and a version without tone controls in the RP72/74/75/76 sets. And in others besides. I'm personally very familiar with these, and if anyone wants to follow along, I'm assuming (as I don't have access) that at least one of these can be found in the Library...

So, failures of the first transistor (T1) are very common indeed. I've changed dozens of these - I replace them with the BC108 (the very earliest versions used the BC108, so this is a reasonably sympathetic repair IMHO). But what about other transistors on the PCB?

The important point is that T1 is followed by a lot of gain, which amplifies the noise produced by T1. If the next transistor in the amplifier (T2, another BC148) was to suffer the same problem, it wouldn't be picked up until it was really bad - it's only followed by one more stage of voltage amplification (T4, an AC128).

Now, these Lockfit transistors are used for other purposes as well. In the Sovereign III, a BC148 is used as a simple voltage regulator for the tuners. It's used as a voltage follower, and so the output isn't followed by lots of gain - in fact it's followed by generous de-coupling. So if this one was to go noisy, you'd probably never know.

Likewise, in the Super-Sovereign, a pair of BC148s are used to drive the tuning/battery meter. Not a critical circuit, and again, you're unlikely to notice a problem until things were seriously faulty.

I could go on, but as you can hopefully see, the noise problem in these transistors is only a problem if the transistor is followed with lots of gain. If not, you'd probably be none the wiser...

Regarding the BF194/5 types, the failure mode for these is a loss of sensitivity in the receiver. Whether it's actually the same problem (rising noise giving the impression of less sensitivity), or some other effect, I haven't tried to determine. But, when faced with low sensitivity, it's normally a Lockfit transistor at fault. I really wish I had a bit of spare time to build up some sort of test jig that can characterise these transistors outside of a radio, as I'd like to be able to sort through my spares (including some NOS). Especially as changing transistors in the FM front end is a PITA - even on sets as nice as Hackers, where everything else is usually a pleasure. It's worth saying that like with the audio amp, it's the "earlier" transistors - earlier in terms of where they appear in the chain - that give the most trouble, which does tend to reinforce the idea that excess noise might be the problem.

Occasionally, they can fail completely, giving no results. In that case, DC voltage measurements give the game away, so diagnosis is easy. I've had a fair few go this way - whether they were noisy beforehand I can't say.

Why they fail is a good question. It's obviously something related to the packaging, and I wonder if the package is somehow allowing moisture in? There could be a mechanical problem, as the short and very stout legs - that go into the PCB with a distinct clicking action - could be stressing the package in some way. Compare the legs of a Lockfit to a TO92 - it's certainly conceivable that this could be part of the problem. And those legs will certainly conduct the heat of soldering towards the die much more effectively than the skinny legs of a TO92. Anecdotally, I sense that they might have become more reliable as time went on (just thinking about the sets I repair, and noticing that it's more of a problem for the earlier models), so perhaps Mullard improved their processes over the years...

There were some quite "beefy" Lockfit transistors - the BC465/464 pair, for example, were used as audio output transistors on many radios, and often with no heat sinking beyond what the copper on the PCB afforded. These fail pretty regularly, but that's probably down to overheating.

Anyway, sorry to have rambled, but it is a relatively complex problem. Hope it makes some sense 😉

Mark

 
Posted : 05/12/2014 9:30 pm
valvekits
(@valvekits)
Posts: 780
Honorable Member Registered
 

Some good points Mark. Back in the day not all BC14X s were the same as they had a suffix letter. Occasionally the difference between a BC148 and a BC148C did matter.

Eddie

 
Posted : 05/12/2014 9:42 pm
Mark Hennessy
(@markh)
Posts: 410
Reputable Member Registered
 

Some good points Mark. Back in the day not all BC14X s were the same as they had a suffix letter. Occasionally the difference between a BC148 and a BC148C did matter.

Hi Eddie,

In common with many similar devices, the letter indicated the minimum Hfe.

As a designer, one tries to avoid circuits that rely on Hfe, but that's not to say that all circuits are immune - as you suggest.

However, this is quite separate from the high failure rate of Lockfits 🙂

Cheers,

Mark

 
Posted : 05/12/2014 10:09 pm
Terrykc
(@terrykc)
Posts: 4005
Member Rest in Peace
 

Back in the day not all BC14X s were the same as they had a suffix letter. Occasionally the difference between a BC148 and a BC148C did matter ...

Again, Eddie, a hangover from the BC10x series ...

These transistors had such a wide spread in the HFE - for a BC108, typically from 110 to 800 - that these were also available in three groups selected for low, middle or high gain.

When all else fails, read the instructions

 
Posted : 05/12/2014 10:20 pm
Marconi_MPT4
(@marconi_mpt4)
Posts: 375
Reputable Member Registered
 

Like--for the Video chroma stages the machine seems full of the low-noise signal amp type BF194.......

What would you guys use in place of BC148 and BF194 and any other Lockfit types you can think of....? :thumb

Hi, for many years I used the more readily available Philips/NXP BF494 and BF495 (hFE rank B) to replace BF194 and BF195 respectively. Specification wise they appeared to be electrically similar but in a different encapsulation. Although the BF49x tended to be more reliable, failure mode of HF transistors in general often amounted to open circuit base-emitter junctions. It was a very common problem with Telefunken (TFK) BF199 used in B&O sets.
Cheers
Rich

 
Posted : 06/12/2014 1:16 am
Cathovisor
(@cathovisor)
Posts: 6390
Famed Member Registered
 

Although the BF49x tended to be more reliable, failure mode of HF transistors in general often amounted to open circuit base-emitter junctions. It was a very common problem with Telefunken (TFK) BF199 used in B&O sets.

I've also encountered that exact same fault, with the very same transistor, in Grundig Satellits. In one case it rendered a set dead on the turret-tuned SW3-10 bands, when the set becomes a dual superhet. It was an o/c transistor in the fixed-frequency oscillator that converted the IF output of the turret section to the 455kHz IF of the main part of the set.

 
Posted : 06/12/2014 9:33 am
valvekits
(@valvekits)
Posts: 780
Honorable Member Registered
 

However, this is quite separate from the high failure rate of Lockfits 🙂
Mark

Whilst it is plausible that the lockfit encapsulation process may have contributed to their demise, there may be other contributory factors. The semiconductor industry wasn’t mature back then and rather than chuck half of the product away, it made sense to select the transistors into different gain groups. These days we would call it a yield problem but back then it was something that came with the territory. I can think of a number of problems in implant, diffusion, photo or etch that alone or together would impact the wafer yield and I wouldn’t be at all surprised if a few years down the road such devices then failed - regardless of how they were packaged. We simply don't have all of the evidence to speculate.

Eddie

 
Posted : 06/12/2014 12:14 pm
Mark Hennessy
(@markh)
Posts: 410
Reputable Member Registered
 

However, this is quite separate from the high failure rate of Lockfits 🙂
Mark

Whilst it is plausible that the lockfit encapsulation process may have contributed to their demise, there may be other contributory factors. The semiconductor industry wasn’t mature back then and rather than chuck half of the product away, it made sense to select the transistors into different gain groups. These days we would call it a yield problem but back then it was something that came with the territory. I can think of a number of problems in implant, diffusion, photo or etch that alone or together would impact the wafer yield and I wouldn’t be at all surprised if a few years down the road such devices then failed - regardless of how they were packaged. We simply don't have all of the evidence to speculate.

Hi Eddie,

Let's not conflate Hfe and reliability, as they are separate factors. Large Hfe spreads have always been the norm, and remain so today.

Regarding reliability, one simply has to compare the BC108 to the BC148 (for example). As noted earlier, they are the same transistor, but packaged differently. So why is the BC108 so much more reliable?

I agree that we might not have "hard" data about this, but there is plenty of pretty convincing anecdotal evidence out there.

Additionally, it's worth remembering that many transistors either work completely or fail entirely. The "partial failure" mode that Lockfits exhibit is relatively unusual, and means that in practice, the actual "casualty rate" is probably much higher (as I explained in detail earlier, whether you notice a "faulty" Lockfit depends on how it's being used in the circuit).

Of course, Mullard stopped using the Lockfit package relatively quickly - no-doubt there were many reasons for this, but I wonder if they'd spotted a reliability problem by then?

All the best,

Mark

 
Posted : 06/12/2014 12:57 pm
valvekits
(@valvekits)
Posts: 780
Honorable Member Registered
 

Hi Mark
I pretty much agree with the anecdotal evidence, I was just trying to consider other factors. A couple of points though

Let's not conflate Hfe and reliability, as they are separate factors. Large Hfe spreads have always been the norm, and remain so today.

If you have astigmatism in the photo process you will produce contact holes that instead of being round will be elliptical. Ironically, the junction capacitance will be decreased and hf response will be better and everything will look hunky dory during probe and test. That little time bomb will be a superb C grade transistor if it slips through the net.

Regarding reliability, one simply has to compare the BC108 to the BC148 (for example). As noted earlier, they are the same transistor, but packaged differently. So why is the BC108 so much more reliable?

Are they the same though? On a production line the wafers will most likely be coded with some other number so it is unlikely that anyone could tell you with certainty. It could even be that the 108 was made in Nijmegen and the 148 in Southampton or vice versa . FAIK the lockfits could have been made in more than one place and that would be something to factor in as well. Where they get packaged after that though is anybody’s guess and personally this is where I favour the packaging argument. A hot humid country with less red tape than we have sounds like a good recipe for things to go wrong.
Cheers

Eddie

 
Posted : 06/12/2014 2:15 pm
Refugee
(@refugee)
Posts: 4055
Noble Member Registered
 

The lockfit package was an early machine assembly part and was a bit of a compromise.
The technology did improve to the point of TO92 parts being machine placed but the real improvement came with SMD.

The lead outs on lockfits have a much greater metal to plastic joint due to them being flat pins and this is most likely the reason for failure but on the other hand the bent pin failures on those early assembly machines would have left the overall reliability much the same.
As soon as the machines improved the package was discontinued and the next change had to wait for SMD to be introduced.

 
Posted : 06/12/2014 2:56 pm
Alastair
(@alastair)
Posts: 294
Reputable Member Registered
Topic starter
 

I'm thinking that the failures are sorta batch/age related...

11 failures--and counting--of both 148 and 194 types all from the same machine that initially did work but over the space of a week started its antics suggest something odd is happening.

The machine in question is an early Philips N1500, recognised as 'early' by the omission in the machine of the cat-gut traces from the lacing springs, no brackets that were fitted to the lower chassis or the wire formed guide from the lacing motor bracket. None of the parts needed/associated with the extra cat-gut, its springs or anything. I don't think the chassis has the holes for it either.....

What date they added the additional protection, Ive no idea, but a wild guess I put this machine at 1972 ish....

Thanks for the heads-up, Ive got on order some BF494, BC108, BC547 and BC558, I went for 'C' suffix where I could, as the schematic seems to have mostly 'C' types noted.
As an aside, Ive a 'sync separator' board for a N1500 from a spares box still in its little grey Philips box wrapping. All the transistors are BC558C, BC547C and that TO92 range, no Lockfits anywhere.....

I'm thinking as the last time I ran the machine, Playback became very noisy, while still having good FM envelope from tape, I'll blanket-bomb the thing with new transistors--much as it goes against the grain to do so!
I seem to recall, E-E getting flaky at the same time too....

 
Posted : 06/12/2014 4:56 pm
Anonymous
(@anonymous)
Posts: 16832
Group Deactivated Account
 

It's a mystery to me. Complete failure I would put down to bond wire failure, it's a bad package design and wasn't long needed, if at all. Though I wasn't standing beside an auto-insertion machine till 1984.

I can't understand the partial failure devices. Is it possible there are two unrelated issues? A process issue (perhaps localised in time and place) and general stress issue, perhaps due to poor dimension of PCB holes causing excess insertion stress?

 
Posted : 06/12/2014 10:13 pm
Alastair
(@alastair)
Posts: 294
Reputable Member Registered
Topic starter
 

Thanks for all the advice and recommendations as to equivalents...

Ive mostly cleared the N1500 of these little time-bombs, I now have a little ceramic dish full having changed 102 of them, The Luma, Chroma PSU/Control-board and the Servo board are now done, and machine basically tests OK. I still have the IF Strip/signals block behind the clock to do, and a couple of BF197 in the chroma cans, as well as the couple of BF194 hiding under the head-drum....

For reference I used the following as replacements--

BC148A-C --BC108C and BC548B/C
BC149B/C --BC109C
BC147 --BC337C
BC157 --BC327C
BC158B/C --BC558C
BF194 --BF494

I tested most of these old ones, and what has become apparent is most of the BF194 all seem rather low HFe, at around 60 odd, and a couple were well below that. The new replacements (BF494) were all well above 100, average was 150--Maybe this is an age thing--the new ones being made with better materials or tolerances than they were 40 years ago.......

So,--If anyone wants some slightly dodgy, well-used and rather old Lockfits--I'm your man! :thumb

 
Posted : 18/12/2014 6:20 pm
Share: