Hybrid Op build questions

Support for JLM Audio Kits

Moderator: Joe Malone

Post Reply
jonte-72
Posts: 11
Joined: Fri Jul 10, 2009 3:39 am

Hybrid Op build questions

Post by jonte-72 »

Hi Joe or anyone more experienced than me!

I just bought a Hybrid op and an output transformer from JLM (very happy with my BA).

I somehow blew the BD140 transistor and have to buy a new. Is there any recommendation on brand? What about the -10 and -16 or G mark after BD140, which should I choose? Is it important that I exchange the BD139 as well to the same brand? Should they be manually matched for best performance?

Also if not too many questions, could the reason for blowing the transistor be that I forgot to strap the IC-socket on BA and not using output transformer, or could I've have killed it using an XLR to unbalanced 1/4"? I could of course accidentaly have shortened something under the Hybrid PCB, or maybe not all pins had contact. The 10R's also got some nice black colour and smoke, but there are still working :-). The LED started pulsating and then turned off just before the smoke. It's still working though. The BA card seems also unharmed.

Before something happend it sounded beautiful anyway! Great preamp. And very useful to switch flavors.

Best regards,
Jon

chrisp
Posts: 184
Joined: Sun Mar 04, 2007 11:46 am
Location: Sydney, Australia

Re: Hybrid Op build questions

Post by chrisp »

Hi Jonte

So far as I know, the BD140 is a stock standard PNP transistor, and any brand that's up to spec should be OK. Do not use the lesser power BD136 or 138 even though they are PNP.

As for trouble shooting, are you running in class A or class A/B (or switching between the two?) DC servo?

It sounds like you might have too much voltage/current running around the Hybrid, so look carefully for shorts, including on the BA board. You have no transformer out, to what are you connecting your three XLR output pins?
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.

jonte-72
Posts: 11
Joined: Fri Jul 10, 2009 3:39 am

Re: Hybrid Op build questions

Post by jonte-72 »

Hi Chris! Thanks a lot for your answer.

I tried both modes, and got both A and AB to work before the crash. DC servo was off.

One little strange thing was that the first LED never worked, even though it was correctly mounted. The second led during the first tests only lit for half a second, every time I switched on power. It also flashed turning off the power. Despite this the amp worked beutifully. I switched to some other OPs, and then back to Hybrid. Now the LED lit constantly in A-mode and was off in AB as expected. Switched OPs some more and then the problems arised. I also noticed a little flash near the 10R on the BA board every time I turned on the power, perhaps after the breakdown of the BD140. The 10R is also still in one piece though, somewhat black (it was a little black earlier).

The XLR was going to a balanced line in on Mbox1, though using XLR to unbalanced 1/4" cable. This in combination with no strap on the IC-socket beneath the Hybrid.

I haven't found any shorts yet. But maybe a temporary short have occured that took the transistor out, giving the other problems.

/Cheers Jon

chrisp
Posts: 184
Joined: Sun Mar 04, 2007 11:46 am
Location: Sydney, Australia

Re: Hybrid Op build questions

Post by chrisp »

Okay - if other opamps work in the BA amp slot just fine, then you've identified the problem lies in the Hybrid build and not with the BA build or its onward connections.

"little flashes" are A Bad Thing(tm). Could be something like a lifted pad (which can be very hard to see) where the pad solder has come away from the circuit board trace, something like that. Also have a good look at the underside solder pads to make sure that the servo link and the class A links are correctly oriented and not shorted. That's one place where voltages can go to wrong places.

DC servo is off? If so, are you aware that you should NOT have the 2M2 resistors mounted? (You can leave off the 47K and both 470nf as well, but you must leave of the 2M2 or you will create a path to ground you don't want.

Failing all this, it would be time to send Joe some nice photos of both sides of your build and let him get his genius onto it.
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.

jonte-72
Posts: 11
Joined: Fri Jul 10, 2009 3:39 am

Re: Hybrid Op build questions

Post by jonte-72 »

Hi again,

I had no idea that the 2M2 resistors should be left out. I will take those out, exchange the transistors and just to be on the safe side exchange the 10Rs as well. This will probably get me going again. By the way, the flash was only seen when the Hybrid was mounted, not otherwise (And as said earlier, probably after the BD140 was gone).

Thanks for all help! I will post again after the changes.

/Cheers Jon

chrisp
Posts: 184
Joined: Sun Mar 04, 2007 11:46 am
Location: Sydney, Australia

Re: Hybrid Op build questions

Post by chrisp »

Jonte

For information, the Hybrid's actual amp circuit uses just one half of the IC opamp in an open loop (no feedback) configuration as an input buffer, ahead of the transistor push/pull amplifier. Joe's idea is to use the second (unused) half of the opamp as a DC servo - a mechanism that takes the output signal and uses a feedback loop to correct any DC offset that has snuck into the signal. You can use an electro cap in series with the amp's output for this purpose as well, and the BA board includes such a DC offset correcting cap in its circuit - so its not necessary to have the Hybrid's DC servo in place when used in a BA build, or alternatively you could have the DC servo on the Hybrid and bridge the capacitor pads on the BA board. Having both does no great harm either.

If you look at the Hybrid's schematic (on the old JLM site) you'll see that the servo path starts from the point where the two 10R resistors meet at the output pin. The two 2M2 resistors and the two 470uf capacitors are used in the inverting and non-inverting input chains and feedback loop, the 47k resistor lies in series with the servo opamp output. Strapping the two servo solder strap pads on the rear of the board then connect the output to the buffer inverting input. But there's a catch if you do not run the DC servo function - even if you do not strap those pads together, the servo circuit still has a path from the Hybrid's output to ground, through the 2M2 resistor and the 470uf capacitor. Removing the 2M2 resistor breaks this path. It does not explain the problems you experienced, but it is another place where things may go wrong.

From my understanding, if you want to build the Hybrid with no servo, its actually very simple. Mount the opamp, the LED, the two transistors, the two 10R and the 8K2. That's it for class A. 7 parts only, compared to 14 with the DC servo switched on. Add in the two additional diodes for class A/B operation if desired.

Joe, please chime in if any of this is wrong.
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.

User avatar
Joe Malone
Site Admin
Posts: 1644
Joined: Wed Jan 24, 2007 11:35 pm
Location: Brisbane, Australia
Contact:

Re: Hybrid Op build questions

Post by Joe Malone »

Sorry I had missed this thread. I just skimmed the posts quickly so let me know if I missed anything.

1. Use any BD140 type for replacement and there is no need to match the BD139.
2. Not sure what is the problem with the led but if it was off and the opamp was working I presume that would have been a short on the led so the hybrid would have been working in class B. (Not Class A or A/B).
(If the led goes open circuit when in Class A mode the hybrid will draw heaps of of current and kill one of the output transistors.)
3. Only need to remove 47uF next to BD139 if running the Hybrid in the BA or BAD or BAN PCB's. (This improves the noise floor when used in the BA BAD BAN)
4. To disable DC servo just do NOT solder the servo pads under the hybrid PCB together. Do NOT remove any parts as this may make the servo opamp oscillate and degrade the hybrid audio opamp section.
Joe :-)
JLM Audio
Capturing Audio without Injury

chrisp
Posts: 184
Joined: Sun Mar 04, 2007 11:46 am
Location: Sydney, Australia

Re: Hybrid Op build questions

Post by chrisp »

Joe wrote:3. Only need to remove 47uF next to BD139 if running the Hybrid in the BA or BAD or BAN PCB's.
4. To disable DC servo just do NOT solder the servo pads under the hybrid PCB together. Do NOT remove any parts as this may make the servo opamp oscillate and degrade the hybrid audio opamp section.
There you go. Its the cap to ground that you remove, not the resistor. And the two opamps on a dual opamp IC aren't actually as independent as they look on paper. That's why Joe's the boss!
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.

jonte-72
Posts: 11
Joined: Fri Jul 10, 2009 3:39 am

Re: Hybrid Op build questions

Post by jonte-72 »

Hi again!

Thanks for the interesting info about the Hybrid!
And thanks for the new tips!

Unfortunately I've already trashed my 2M2 resistors. Better luck next time :-). Anyway, I'm learning a lot on the road.

I got it to work, but are still experiencing some problems. And I have some clues that I want to ask about.

First a confession. In the early days after finishing my BA, I mounted the OP2604 in the wrong direction (tricked a little bit of the shape of the socket). It was then my 10R on the BA got a little black, still working though. When I mounted it correctly, everything worked fine, and the preamp sounded great. So I didn't think anything got harmed.

With the Hybrid, I still have similar problems as described earlier. The LED is a bit unstable. After 8 minutes is suddenly went out. I toggled the power, and the LED lit for less than half of a second. I took out the Hybrid and blew on the board to get rid of heat. Mounted it again. Now the LED lit again. The transistors get very hot. You could get burned even on the legs of the LED. And the caps next to the transistors also get heated.

Clue 1, my first and maybe definite clue to the problem:
• When switches are to the left side and components are facing upwards, the topmost resistor above the OP socket measures only 16 ohms, mounted on the board. The corresponding resistor at the bottom measures 8,1k. This seems suspicious? Could the resistor have been damaged when I mounted the OP2604 in the wrong direction? Or could a short give this value? If this is not supposed to be, it will be easy for me to just exchange it, and you don't have to read the rest of my clues.

Clue2:
• With electronic balancing, the level drops about 4dB using JLM99v or Hybrid (A or A/B) compared to OP2604 at the same gain setting. With the trafo all levels are the same. Is this correct?

Clue3:
Even the Op2604 gets quite hot. I thought this was normal, but after 10-12 seconds putting your index finger on top of it, you feel an strong urge to withdraw it. Is this normal temperature?


Cheers!
/Jon

PS. I'm now using a balanced cable to the soundcard. PS2, I also have more professional soundcards than mentioned in the earlier mail, but those are at another location. PS3, Thanks for a beautiful sounding preamp. Quieter than a mouse (the noise I mean :-)). DS.

jonte-72
Posts: 11
Joined: Fri Jul 10, 2009 3:39 am

Re: Hybrid Op build questions

Post by jonte-72 »

Hi!

Just wanted to tell something glad and interesting. First of all I understand now why I got 16 ohms over the 10k resistor on the + side. It was the output trafo of course. Sorry I didn't see that. Now to the glad news. I watched the Hybrid some more. I measured the current over the 10R resistor in the power circuit (don't know if this is a good place to measure though). The Hybrid gave 15mA whereas the other OPs gave 0. I don't know what part the measuring instrument plays here. Anyway, the Hybrid got as usual very very hot as described earlier. But after about 8 minutes the temperature suddenly dropped. The cap and LED went perfectly normal and could be touched. Also the measuring over the 10R was now only 3mA. Another difference is that the LED know fades out at power off instead of going off instantly. The sound is great! And all OPs know works perfect! Actually I think my Hybrid got so hot that some solder points melted, and that a poor solderpoint thus got fixed making the Hybrid work as expected. I have never experienced something similar where the electronics have repaired itself. Very funny!

One last question, just to be sure everything is as it should be. Is my measure over the 10R expected? I'm also a little curious about what the fault could be.

Now I have a lot of sonic colours to play with. And all sound great in their different ways. Thanks! And I'm sorry if I due to my limited knowledge have put some amateurish questions on the board.

Cheers!
Jon

User avatar
Joe Malone
Site Admin
Posts: 1644
Joined: Wed Jan 24, 2007 11:35 pm
Location: Brisbane, Australia
Contact:

Re: Hybrid Op build questions

Post by Joe Malone »

Clue2:
• With electronic balancing, the level drops about 4dB using JLM99v or Hybrid (A or A/B) compared to OP2604 at the same gain setting. With the trafo all levels are the same. Is this correct?
The electronic balancing due to its inverting side will be 6dB louder. The 99v and hybrid will be the 6dB down unless a 1:2 output transformer is fitted which will make it the same level as the electronic balanced version.
Clue3:
Even the Op2604 gets quite hot. I thought this was normal, but after 10-12 seconds putting your index finger on top of it, you feel an strong urge to withdraw it. Is this normal temperature?
The OPA2604A runs hot as it has 48v across a DIP8 package which is normal.
PS. I'm now using a balanced cable to the soundcard. PS2, I also have more professional soundcards than mentioned in the earlier mail, but those are at another location. PS3, Thanks for a beautiful sounding preamp. Quieter than a mouse (the noise I mean :-)). DS.
Cool :-)
One last question, just to be sure everything is as it should be. Is my measure over the 10R expected? I'm also a little curious about what the fault could be.
The 10R are 5% tolerance so it could be tolerance or the resistors have been stressed current wise so have changed to a higher value. Measuring also depends a lot on multimeters and if there is any residual voltage across the resistor when measuring.
Now I have a lot of sonic colours to play with. And all sound great in their different ways. Thanks! And I'm sorry if I due to my limited knowledge have put some amateurish questions on the board.
Cool :-)
Joe :-)
JLM Audio
Capturing Audio without Injury

Post Reply