Okay, same test conditions as above.
MJL1302A average 11n22 out of 20 pcs, deviation less than 0n2. Gains were all over the map.
MJL3281A average 9n03 out of 20 pcs, deviation less than 0n5. Gains spreads were better, but still much more variable than the Toshibas.
The second decimal is probably not justified, even though the readings were all stable. Leads were attached by hand and so there is variability there.
This lot was 95, so 11 years old now.
-Chris
MJL1302A average 11n22 out of 20 pcs, deviation less than 0n2. Gains were all over the map.
MJL3281A average 9n03 out of 20 pcs, deviation less than 0n5. Gains spreads were better, but still much more variable than the Toshibas.
The second decimal is probably not justified, even though the readings were all stable. Leads were attached by hand and so there is variability there.
This lot was 95, so 11 years old now.
-Chris
Okay, same test conditions as above. Batch from 2005, 5 pcs each.
MJL1302A average 11n9, deviation less than 0n1.
MJL3281A average 8n5, deviation less than 0n2.
These figures are still in the ballpark, their process appears to be getting better (less spread in gain and base - emitter capacitance).
-Chris
MJL1302A average 11n9, deviation less than 0n1.
MJL3281A average 8n5, deviation less than 0n2.
These figures are still in the ballpark, their process appears to be getting better (less spread in gain and base - emitter capacitance).
-Chris
Interesting devices these MJL0281A and MJL0302A.
Average capacitance over a 10 pc sample of each were 7n4 (MJL0302A) and 7n3 (MJL0281A). DC beta's were all pretty close.
I'm liking these. I think MikeB's Symasym will be built with these little guys. I have a few other ideas up my sleeve for these too. I'll have to check another set later to see if these are really as consistant as they seem to be.
-Chris
Average capacitance over a 10 pc sample of each were 7n4 (MJL0302A) and 7n3 (MJL0281A). DC beta's were all pretty close.
I'm liking these. I think MikeB's Symasym will be built with these little guys. I have a few other ideas up my sleeve for these too. I'll have to check another set later to see if these are really as consistant as they seem to be.
-Chris
Just measured a whack of MJ21196G's Variations are small there as well compared to earlier measurements. Capacitance averages 5n5. Waiting for the MJ21195's to arrive.
Tired of testing transistors now......
From what I am seeing Clem, it should be enough tp measure E-B capacitance with an idea of where the value should be. As you could see from my 4 fakes, they are getting really good at copying the case!
-Chris
Tired of testing transistors now......
From what I am seeing Clem, it should be enough tp measure E-B capacitance with an idea of where the value should be. As you could see from my 4 fakes, they are getting really good at copying the case!
-Chris
Yes, i really can recommend the MJL0302/0281, they sound very precise, powerful and non fatigueing in the symasym, not as soft as the Toshibas (sc5200/sa1943), but this could be a matter of taste... I was excited hearing them for the first time.
Right now, they should be too new for fakes too exist ?
Chris, thanks for the measurings ! I would have expected the MJ21196 to have much higher capacitance regarding their high SOA... (2.5A at 80v)
Mike
Right now, they should be too new for fakes too exist ?
Chris, thanks for the measurings ! I would have expected the MJ21196 to have much higher capacitance regarding their high SOA... (2.5A at 80v)
Mike
Hi Chris,
How is the gain spread on the 21196 compared to the 0281/0302? I remember you were considering using these for the Symasym as well.
Hi MikeB,
Have you tried the 21195/21196 on Symasym? How do they compare sonically to the other outputs you have auditioned?
I might favor using the 21195/21196 instead for my amp, since I currently do not have the 0281/0302.
Cheers!!
Clem
How is the gain spread on the 21196 compared to the 0281/0302? I remember you were considering using these for the Symasym as well.
Hi MikeB,
Have you tried the 21195/21196 on Symasym? How do they compare sonically to the other outputs you have auditioned?
I might favor using the 21195/21196 instead for my amp, since I currently do not have the 0281/0302.
Cheers!!
Clem
Hi Clem, i haven't tried the 21196/5 for symasym yet, in my other amps these always sounded warmer and a bit less powerful than the modern devices. The 21194/3 are sometimes used in commercial high end amps, they can't be bad...
The 21194 i've measured all had hfe at ~100. One out of ten had "only" 80. I haven't measured the 21196/5.
Mike
The 21194 i've measured all had hfe at ~100. One out of ten had "only" 80. I haven't measured the 21196/5.
Mike
Hi Mike, Clem,
The earlier MJL2119X and MJ2119X devices were all over the map for DC beta, I was unhappy with that as the matched yields were low. Compared to MJ1502(2,3,4,5) they were inferior in that respect. Also, the P to N groups were widely off. Reminded me of those horrible MJ15015 / 15016 devices.
So the batch of MJ21196G were a very nice surprise as I didn't expect that much of an improvement. I want to see how the compliment behaves and how close it is. Also the drivers. Solderability is going to be another thing again. I wonder if the joints will crack after a while?
I really think they got an analog amplifier designer who understands linearity at On Semi now. Good things to come? Previous to this, the Japanese transistors were superior to the US parts in every respect. Especially match yields, to the point that the Japanese parts were actually less expensive than the US parts.
For what it's worth, that's my experiences with outputs over the last 30 years.
-Chris
The earlier MJL2119X and MJ2119X devices were all over the map for DC beta, I was unhappy with that as the matched yields were low. Compared to MJ1502(2,3,4,5) they were inferior in that respect. Also, the P to N groups were widely off. Reminded me of those horrible MJ15015 / 15016 devices.
So the batch of MJ21196G were a very nice surprise as I didn't expect that much of an improvement. I want to see how the compliment behaves and how close it is. Also the drivers. Solderability is going to be another thing again. I wonder if the joints will crack after a while?
I really think they got an analog amplifier designer who understands linearity at On Semi now. Good things to come? Previous to this, the Japanese transistors were superior to the US parts in every respect. Especially match yields, to the point that the Japanese parts were actually less expensive than the US parts.
For what it's worth, that's my experiences with outputs over the last 30 years.
-Chris
Hi Mike,
I have to admit to being a little excited myself to hear those new MJL0281A / MJL0302A. They have the tightest grouping of parameters I've seen yet (I wonder if they were pre-sorted?). I want to wait and try another production run to really see. I'll bet you start seeing copies once commercial amplifiers turn on to these. At least we are now aware of this and have some ballpark measurements to work with.
I want to believe On Semi has done as good a job with the newer MJ2119X series. And yes, the capacitance was lower than I expected. I can guarantee those were from On Semi too.
Clem,
The gain spread on the MJL2119X is a little wider, but nothing like earlier batches I've seen. I expect good things there. I must say it was freaky how close MJL0281A and MJL0302A were. Just like advertised on the spec sheet. Now there is a process that is under control!
-Chris
I have to admit to being a little excited myself to hear those new MJL0281A / MJL0302A. They have the tightest grouping of parameters I've seen yet (I wonder if they were pre-sorted?). I want to wait and try another production run to really see. I'll bet you start seeing copies once commercial amplifiers turn on to these. At least we are now aware of this and have some ballpark measurements to work with.
I want to believe On Semi has done as good a job with the newer MJ2119X series. And yes, the capacitance was lower than I expected. I can guarantee those were from On Semi too.
Clem,
The gain spread on the MJL2119X is a little wider, but nothing like earlier batches I've seen. I expect good things there. I must say it was freaky how close MJL0281A and MJL0302A were. Just like advertised on the spec sheet. Now there is a process that is under control!
-Chris
Hi Eva,
That's a theory I've had for a long time. I have had comments from customers (positive, unrequested) when I take the time to match all the parts for DC beta in an amplifier. If you look at a Marantz 500 service manual, they go part way in doing this. That's actually where I got the idea to begin with. The one model 500 I did where I tightly matched all the parts, everyone who had an audition was very impressed. The guy who owns it has cloth ears!
But even he noticed. Even the Nakamichi Sales rep was very positive about it (he now reps Bryston).
The THD is a little lower but there is something else going on that isn't being measured. I'll bet it's dynamic in nature.
-Chris
That's a theory I've had for a long time. I have had comments from customers (positive, unrequested) when I take the time to match all the parts for DC beta in an amplifier. If you look at a Marantz 500 service manual, they go part way in doing this. That's actually where I got the idea to begin with. The one model 500 I did where I tightly matched all the parts, everyone who had an audition was very impressed. The guy who owns it has cloth ears!
But even he noticed. Even the Nakamichi Sales rep was very positive about it (he now reps Bryston).The THD is a little lower but there is something else going on that isn't being measured. I'll bet it's dynamic in nature.
-Chris
anatech said:
Hi Chris, the dataheet confirms your measuring, the MJL21196 is rated with 4.3nF at 2v (cib). The MJL21195 with 5.4nF...
My experience with MJE15030/1 is that DC-hfe measured 130 to 120 (20v), just taken out of the bag.
About the MJL0281A / MJL0302A, i guess they are from the same batch, ordered that early as samples ? At least mine all show the same batch numbers, pnp and npn.
Mike
Hi Mike,
My measurements were unbiased(DC ), didn't even look at the datasheet. I'm glad they are within experimental error.
The new parts seem to be much more consistant than ever before reading your gain statements. That is very good news for DIY'ers. No longer do you have to buy a ton of transistors to get matches.
Speaking of which, I am currently matching a ton of 2N5551 and 2N5401's for diff pairs. Two sets of about 40 pcs. I should just buy some new ones it sounds like.
Yes, they were samples. Same batch (different #s for NPN and PNP). I have a bunch of signal transistors purchased over ten years ago. As I mentioned I should really check out the new parts.
-Chris
My measurements were unbiased(DC
), didn't even look at the datasheet. I'm glad they are within experimental error.The new parts seem to be much more consistant than ever before reading your gain statements. That is very good news for DIY'ers. No longer do you have to buy a ton of transistors to get matches.
Speaking of which, I am currently matching a ton of 2N5551 and 2N5401's for diff pairs. Two sets of about 40 pcs. I should just buy some new ones it sounds like.
Yes, they were samples. Same batch (different #s for NPN and PNP). I have a bunch of signal transistors purchased over ten years ago. As I mentioned I should really check out the new parts.
-Chris
I don't know if 2SD424 are on the list of fakes but if they aren't they are now.
I just cracked open 2 Toshiba 2SD424 TO3 transistors that had the lable printed the wrong way. Besides that the top came off fairly easy with a knife (unlike the real ones that you end up destroying the case). Inside the case has a very tinny sliver of slicon thats maybe 3mm x 3mm, it's covered in white silicon and has very small leads for the base and emitter (emitter has 2 wire).
I had put these in a customers unit last week (before I knew of this thread) and the unit came back the same day. The guy didn't even have it past 3/4 volume and they shorted 
Just wanted you to know.
P.S if you want pictures just ask and I'll get some.
I just cracked open 2 Toshiba 2SD424 TO3 transistors that had the lable printed the wrong way. Besides that the top came off fairly easy with a knife (unlike the real ones that you end up destroying the case). Inside the case has a very tinny sliver of slicon thats maybe 3mm x 3mm, it's covered in white silicon and has very small leads for the base and emitter (emitter has 2 wire).
I had put these in a customers unit last week (before I knew of this thread) and the unit came back the same day. The guy didn't even have it past 3/4 volume and they shorted
Just wanted you to know.
P.S if you want pictures just ask and I'll get some.
Hi ifrythings,
(I cook things slowly)
love the handle!
They were close to the first ones to be faked. I used 2SD424 and 2SB554 all the time. They were excellent outputs. They early fakes were truly obvious. Very amusing really. I couldn't believe they would be able to pass any. They were even marked on early Motorola cases (you know, early aluminum 2N3055 cases!!)
Then they started to turn up in repairs that had been to other shops. I mean, you had to laugh. Later fakes became quite good. I assume you got nailed with good fakes.
Anyway, unless you pull them from an old stereo or luck into old factory spares, they be fake man! These have not been available for around 15 years or so.
Sorry to hear about your bad luck. It's the labour that costs (and the reputation).
-Chris
(I cook things slowly)
love the handle!They were close to the first ones to be faked. I used 2SD424 and 2SB554 all the time. They were excellent outputs. They early fakes were truly obvious. Very amusing really. I couldn't believe they would be able to pass any. They were even marked on early Motorola cases (you know, early aluminum 2N3055 cases!!)
Then they started to turn up in repairs that had been to other shops. I mean, you had to laugh. Later fakes became quite good. I assume you got nailed with good fakes.
Anyway, unless you pull them from an old stereo or luck into old factory spares, they be fake man! These have not been available for around 15 years or so.
Sorry to hear about your bad luck. It's the labour that costs (and the reputation).
-Chris
clem_o said:
Hi Clem, i did now and was surprised by the results. These BJTs produce a very lovely/warm sound, but need a lot higher bias or you will get "scratchy" trebles. Of course they do not deliver the ultrasharp detailed trebles like with fast OTs, but you will get a nice sounding amp with them. The minimum bias is >100ma, with the increased bias you can use the MJL21196/5 without problems.
Mike
And now for the measurements Clem was waiting for ...........
Measurements were taken with an HP 4263A and 16089B Kelvin lead set. The leads were compensated just prior to the measurements. Settings were 10KHz and 100mV.
[repeated for reference]
The On Semi MJ21195G averaged 6.6 nF with a max deviation of 0.4 nF. The DC beta was about twice the value for the NPN devices on average . A bit looser spread and more "flyers". I hate that.
Right now, barring reliability issues, the best audio devices may be the new MJW0281A and MJW0302A. The MJ2119X series can handle more voltage and current. If you don't need this the newer parts may yield a better sounding amp.
Mike, you know where I'm going on past experience. Which ones do you like the sound of in your amp the best? I realise it's your personal opinion.
-Chris
Measurements were taken with an HP 4263A and 16089B Kelvin lead set. The leads were compensated just prior to the measurements. Settings were 10KHz and 100mV.
[repeated for reference]
The On Semi MJ21195G averaged 6.6 nF with a max deviation of 0.4 nF. The DC beta was about twice the value for the NPN devices on average
. A bit looser spread and more "flyers". I hate that.Right now, barring reliability issues, the best audio devices may be the new MJW0281A and MJW0302A. The MJ2119X series can handle more voltage and current. If you don't need this the newer parts may yield a better sounding amp.
Mike, you know where I'm going on past experience. Which ones do you like the sound of in your amp the best? I realise it's your personal opinion.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.