Transistor replacements

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hope you don't mind me tacking onto your old thread calmart. I'm chasing my tail on this one. I've since discovered some errors in the schematic from the 5M21 service manual. The picture attached is out of the service manual with the measurements I'm getting. First, pin 14 on the hybrid IC shows it connected to the -67V supply when in fact after removing the ic, the pin was folded over and not connected at all. Second, missing from the schematic is a 120K resistor connecting from the -67V supply to pin 5. Lastly, pin 12 is connected to no where. You can see Luxman originally had a trace leading to pin 5 but was later unconnected. The actual makeup of the IC has been posted earlier thanks to jacco. I'm since confinced my problem does not lie with the hybrid IC. Can someone help me make sense of this thing?

 
jacco vermeulen said:
I didn't even know the DML-02 existed, found this one at a Lux page : DML-02
Now we all have something new(old) :clown:


Yes, that's it. It belongs in the preamp and it's used in the MM and line preamp.

Do you know that 2SK60 part? What I found was a single, not dual, MOSFET, or dual FET. I doubt they would be using a MOSFET for that.

The different DML-01 version on the left looks to be drawn wrong, BJT symbol insted of JFET, the matched pair NEC should read uPA63H (or alternative uPA68H)

Maybe there was a bipolar version.
 
The 2SK60 number is wrong as well, but my head hasn't yet dropped a simular looking JFET number.

2SK60 is a power VFET, complementary of the 2SJ18, TO3 devices with a power handling of a little over 60W.
2SK50//2SK65/2SK66 are close lookalike numbers, those are JFETs, but neither of them would do.
There's one that would work, the 2SK63, but that's a really really exotic small signal VFET device and only used in very few Sony amps from the 70s.
 

Attachments

  • 2sj18-2sk60.jpg
    2sj18-2sk60.jpg
    19.2 KB · Views: 914
Easy to find, but would become costly to obtain well matched duos.
Duh, why replace 250mW single JFETs by a very expensive dual 600mW ?
(a 2SK146 has 10 times higher Gm than the NEC JFET)

Much cheaper would be matched 2SK170, glued together and resin encapsulated in a DIY DML box would ensure pretty good thermal tracking.
Looks more authentic too.
 
SUCCESS..... Kind of

After sorting out the errors in the service manual and schematic, I've finally found my problem with the pre driver stage. There were a couple of problems and unfortunately one of them is in fact the DML-01 FET hybrid IC. The odds of finding an IC are next to nil so it looks like the only other way to save this amp is to build the IC with discrete components.

If I'm not mistaken from the two different schematic versions jacco provided I could use either of the dual FETs uPA63H/uPA68H providing I can get them or I could use matched 2SK150A's. Obviously the dual FETs would save me the daunting task of matching single FETs and would be a better choice but are there any other options/suggestions?

jacco vermeulen said:
DML-02
Now we all have something new(old) :clown:

The different DML-01 version on the left looks to be drawn wrong, BJT symbol insted of JFET, the matched pair NEC should read uPA63H (or alternative uPA68H)

I believe your correct and assuming the bottom two transitors are suppose to be drawn as FETs, it looks like the gates (bases) and the sources (emitters) are labelled wrong also. The left FET has the gate as pin 7 and the source pin 3. The right the gate is pin 2 and the source pin 6. Should these not be left FET gate as pin 2 and the source pin 3 while the right gate pin 7 and the source pin 6?

I can't believe all the trouble encountered over an unobtainable IC and the work that is still instore to get this thing back up and running. I'm very close to using the PS to build another amp. 60 volt rails should yield some decent DIY amp options.
 
Providing I can obtain one or two of the dual FETs, I have decided to build my own DML IC. I'm having a hard time sourcing either of the 2SK150A, uPA63H or uPA68H FETs naturally. I guess another option might be getting a bunch of 2SK246 and match them. What about a 2SK389? Anyone have some options or suggestons?
 
Building discrete DML-01s

Hoping to revive this old thread since I am now myself in the position of having to build a discrete replacement for a DML-01 in my 5M21. I have many questions, probably mostly dumb ones:

1. Which of the attached is the correct schematic? Or are both correct?
2. Did anyone ever successfully build such a board?
3. If so, was a PCB fabricated or is there some method of breadboarding to an IC socket I don't know about (I don't know much, so there probably is...)

If I can get this figured out, I will build some extras for other Lux fans in the same boat (and as best I can tell, this is a common fault).

Looking forward to your responses.
regards,
Fred
 

Attachments

  • diffamp1.jpg
    diffamp1.jpg
    16.2 KB · Views: 413
  • DML-01.jpg
    DML-01.jpg
    32.2 KB · Views: 417
Last edited:
I feel your pain. I have a C-12 that needs the DML-2 IC. Interestingly, this pre-amp has already been repaired once and it has a small board labeled DML-2 built with discrete components. I have no idea who built that board but it does appear to be functional.

From a quick check with a meter, it does not appear to be the same circuit as described by the drawings referenced within this thread. I imagine I will have little choice but to desolder it and draw it. This drawing is the only DML-2 drawing I've found and it is not complete. It is missing the 5 pin which appears to be the signal out.

I probably won't get to this for a few days. If anyone has crossed this bridge since this post went dormant I'd sure appreciate hearing what you learned.
 
Can any of you guys post an actual picture of the hybrid with dimensions, pin spacing, etc.?

I can source most of the parts, design the PCB, match single JFETs if needed and build them.

The original part is 30mm by 20mm with 12 pins (6 to a side of course). The pin spacing is 5mm and the rows are separated by 15 mm. So the pitch is 5mm by 15mm.

The photo is of the work around someone built with discrete components. It is a slightly different circuit than the other drawings. You will note two TO-92 packaged components on top and there are four below. Only one pair are marked (C1345). It's difficult for me to say if the circuit design is any good.

Thanks for the interest.
 

Attachments

  • photo (11).JPG
    photo (11).JPG
    112.3 KB · Views: 362
I have not used Utsource. I'm really a vacuum tube guy so I've not that much experience sourcing solid state stuff. I did go to their site last night at diy didi's recommendation but was unable to locate the required devices from their site. Perhaps you are suggesting that I should contact them in the hope that they will have components not listed in their data base?

Like I hinted at above, I'm butt stupid about solid state design, so I'm coming at this at something of a disadvantage. I does occur to me that I, or perhaps someone more knowledgable than I, should be able to come to some understanding of the circuit and select current production devices to substitute for the obsolete devices. Now that may be easier said than done; I'm too much of a noob to know.

Perhaps you or someone else would be so generous as to give me the "devices for dummys" thumbnail sketch of the bumps in the road to such a solution. After all, this hobby is about learning and I'm interested in getting a little sand under my fingernails.
 
Dave,

The UTsource question is meant for didi not for you. I hate it when people post one liners with no explanation, useless information in my opinion.

Anyway, please confirm if the pitch is indeed 5mm and not 5.08mm. All the hybrids I've designed have 0.1"/2.54m pitch or multiples of this. Use a 20-24pin IC or socket to confirm the pitch. Line up the pins and eyeball them.

Yours is a DML-02 with 6 active devices. The DML-02 here (Luxman audio products) has 8. More confusion.

So this project is for 2 hybrid versions. I have to look at Luxman's schematics first and figure out something.

Cheers!
 
You are correct about the confusion. The version I have does have six active devices. I must point out that the DML-02 "luxman circuit" that you reference has been questioned as it does not appear to use the #5 pin as the Luxman chip does. The signal-in is pin #4 and the signal-out is pin #11. I scoped the signals to get an idea of gain and I'll post that when I get back to my bench and my notes.

I looked over the work-around circuit and think it is very similar to the on-line sketch on that Luxman site, using the top six active devices and ignoring the lower two and the 3 resistors. Pin 12 is connected to pin 2 via a 22K resistor rather than the circuit as drawn. Pin 5 is connected opposite pin 9. The 64 K resistor was 47K and that agrees with the original I have desoldered. The 58K resistor measures 56.7K. I'd guess 56K as a standard value. The odd thing is that pin 1 (the 0V pin) does not appear to be in use on the work-around. I'll desolder the thing tomorrow and post a drawing and sort out the pin 1 thing.

On the original, the resistance between pin 1 and pin 12 shows 17.1K, which is a pretty close correspondence with the "luxman drawing".

My guess, and it would be just a guess, is that the work around circuit was simplified to just 6 active and 2 passive devices so that the circuit would be manageable to build on such a small card. A second guess would be that it is an inadvertent mistake that the drawing omits the 5 pin opposite the 9 pin. Both those are just intuition and we know what that is worth.

Of course you are correct about the pitch: it fits a 0.1 inch pitch predrilled card using every other hole on the length and six holes to the width.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.