The No.1 Website for Pro Audio
Roland JC120-H Output Transistors
Old 21st December 2016
  #31
Gear Addict
It's weird because those measurements are dead on consistent with Q23's base not being properly biased through R105 (56k), but you redid those solder joints, so I can think of two possibilities: either the resistor itself is dead, or the trace going from it to ground is broken at some point. I would remove it, measure it, put it back and then check continuity between its leg going to ground and point 39. If you don't get 0 ohms, follow the PCB trace to see if it's broken at some point. Note that Q23 losing its biasing is also consistent with the output transistors going poof, so I think we could be onto something here.

The poking test is indeed interesting, there may be some cold solder joints around those components, especially dangerous with Q15 and Q16, if something goes wrong there the healthy channel may lose its health, so I would re-do those solder joints too.

If you got replacements for those leaky caps (C6 and C7, right?), replace them, install also the replacements for C59 and C86 and repeat the previous test on the faulty channel, again with the output shorted to ground. To give you an idea this is the ballpark of what you should get if all is finally in order:

R113: ~650mV
R114: ~33V
R115: ~850mV
R144: ~23V
C68: ~33V

Note that R115 and R144 can vary a lot depending on the matching of the input transistors, but the others shouldn't be too far off those values.

As for the healthy channel, if you redo R55, R53, Q15 and Q16, you may want to check the voltage across the resistors again in case something has changed, although the values you've measured are fine.
Old 22nd December 2016
  #32
Gear Head
 

Redid solder joints on Q15 and Q16, replaced caps 6,7, 59, 86. Took out R105, tests exactly on spec.. Cleaning traces with alcohol everywhere and getting any flux or anything I see off of anything. Also cleaning the inside of the chassis. Before I do tests again, I found something interesting.. C66, which is right there in the section that we are dealing with.. it is a 10 uf cap and I cannot, for the life of me, find it anywhere on the schematic. I went back and looked at my pictures of the board from a while back and yeah its been there the whole time. Should this even be there??? There are other C(insert number here) on the board but no physical element, I look on the schematic and they are nowhere to be found. This is the only case where there is an element there and isn't there at all on the schematic itself. Look at the picture I posted of above the board from a while ago, it is right above R105 and to the right of C59. Any opinion on that little guy?
Old 22nd December 2016
  #33
Gear Addict
Found it! There's a misprint in the schematic: if you follow the base of Q23 to the left you will find... Another C59! That's actually the right C59, so, where it says C59 in the cap between R112 and ground, it should say C66.

If you removed the actual C59 (a 1u/50V bipolar) just put it back, that wouldn't have affected the tests anyway, and replace C66 with the Panasonic bipolar 10u/50V, if that's what you got. If you got a normal (polar) electrolytic, then don't put it in, the voltage at the base of Q24 could go either way in the tests and we don't want reversed caps exploding on us...
Old 24th December 2016
  #34
Gear Head
 

Alright! So switched those caps and finally got that straightened out. Thanks a lot for that find! Quick question for these pcb boards that were obviously used for several different designs, components left blank on the board, etc. Some of the holes left blank have solder in them and then a choice few that don't. Should it directly correspond to a schematic which ones were filled in or just some are/some aren't? I got really curious in thinking that a left blank hole on the board was 'breaking' the line somewhere in between the resistor and ground.

Spent several hours cleaning up the board(I've since ordered a nice soft tipped flux remover pen for pcb boards from newark). I also ordered replacements for c5 and c8(dreadful leakage), and R4(covered in the leakage from all of those caps in the input section). Also found ANOTHER(2 actually) solder points that weren't perfect and redid those).

SO! did the test again after all that cleaning...

R113: 837mV
R114: 30.14V
R115: 17.5mV
R144L 45.7V
C68: 27.2V

Seems a lot closer! I guess flux can muck things up a bit? Something definitely was changed during the cleaning process! What do you think?
Old 24th December 2016
  #35
Gear Addict
Whether the pads have solder or not shouldn't make a difference if the copper trace is intact, but some fluxes are and/or become conductive by absorbing humidity (you would expect manufacturers to clean those but they don't always do it), so between the solder joints you've redone and the thorough cleaning I think you may have fixed the problem, because those measurements are looking very good indeed: if I simulate the circuit with some suitable Vbe imbalance in the input pair (to be expected), I pretty much nail those values. Great job!

Next step: put the driver transistors back (you could re-test them before just to be sure they're ok), remove the short from the output and measure again the voltages across R113, R114, R115, R144 and C68, and also these (ballpark values in square brackets):

- Across R135 (330R) [0.5-0.6V] and R118 (330R) [0.5-0.6V]
- From ground to the output (point 36) - [up to a few tens of mV]
- Between the bases of the drivers (Q29, Q30) - [~1.8V]

With the drivers in we now have DC feedback that should force the differential pair to (almost) balance, so things will change from the last measurement, the values should be much closer to the ballpark ones I gave you two posts back. If all checks out we can then mod the Vbe multiplier to make it adjustable and more stable, install the new output transistors and see if we can get that channel up and running again!
Old 26th December 2016
  #36
Gear Head
 

Day after Christmas test!

R113: 680 mV
R114: 33.45V
R115: 1.16V
R144: 21.40V
C68: 31.36V
R135: .572V
R118: .560V
Ground to 36: jumps around, around 50mV
Bases of Q29, Q30: 1.81V
Old 26th December 2016
  #37
Gear Addict
That looks great! If all the other readings were stable, the DC offset jumping around a bit shouldn't be a problem, it's a tiny voltage and even changes in probe pressure against the measuring point can affect it. But all the other measurements are where they should be so I think we're in good shape to proceed with installing the new output transistors, but not before modding the Vbe multiplier.

I don’t want to bore you with a full lecture but I do want to explain a bit the mods I propose, so you know why you’re doing it. Attached below is a schematic of the stock and modded circuit and a plot showing the result of the simulation of output voltage vs. current. The mods:

- R145 replaced with 10k trim pot: this is to make it adjustable since the new output transistors will have different parameters than the stock ones and require a different bias voltage to achieve a given quiescent current setting. In fact most hi-fi and pro audio amplifiers have it adjustable because even the differences between individual samples of the same type of transistor can be enough to require some adjustment.

- 1k5 resistor between base and emitter of Q32: that's the reason why the output voltage vs. current plot is much flatter than the stock one. Without going into too much detail, that resistor draws extra current from the collector of Q33, pulling it more firmly into its active region and making its Vbe, which is what the multiplier multiplies, much more stable (incidentally it also makes the Vbe of Q33 larger, which is why to get the same output voltage for a given current the trimmer needs to be adjuster to a higher value than 4k7, to reduce the multiplying factor). This is specially important in this case where the load of the VAS isn’t a constant current source but a resistor (R140+R144): any increase in supply voltage (e.g. due to a mains surge) will increase the current through the Vbe multiplier, which in the stock circuit will increase its output voltage significantly, which in turn will increase the quiescent current of the output transistors, risking thermal runaway.

- 22n cap between collector and emitter of Q32: it's used in some other versions of the JC120 I've seen and with good reason, since that configuration of Q33 and Q32, called complementary feedback pair (CFP) or Sziklai pair, is prone to oscillation and that cap will prevent it.

To replace R145 with the trimmer pot, it's very important that before installing it you bend the middle lead to touch one of the side leads, solder them together and then turn the pot all the way to the end where you measure the full ~10k value, so that when it's installed and you power up the amp, the quiescent current will start at its minimum value. The 1k5 resistor and 22n cap can be simply soldered directly onto the solder side of the PCB.

If you got the 10R resistors, insert them in series just before the bases of the new output transistors and as close as possible to them. On those little PCBs where the transistors are installed, you can just lift the wire that goes to the base and insert the resistor there. Their purpose is again to prevent oscillation, which could happen because of the different parameters of the new transistors (it’s very unlikely in this circuit but better safe than sorry). Install the transistors on the heatsink (as I said before, if you use a mica insulator, spread a very thin layer of thermal grease on both sides of it), connect a speaker to the output of that channel too and switch on the amp.

To adjust the quiescent current, you need to measure the voltage between the emitter of Q28 and the collector of Q25. You’re going to measure a very small voltage and you need to watch it constantly as you adjust the trimmer, so it’s a good idea to use crocodile clips or even solder a pair of wires to the measuring points with banana plugs on the other that you can plug into your DMM. After a brief initial transient when you switch on the amp, the voltage should be almost zero. Turn the trimmer until you get some reading and then keep going (slowly!) until you measure 11mV. Give it a few minutes to make sure it’s stable and readjust if necessary.

Now it’s finally time to plug in a guitar and see if it works. If it does, you can then proceed to do the same mods to the Vbe and replace the output transistors in the healthy channel too.
Attached Thumbnails
Roland JC120-H Output Transistors-vbem-sch.png   Roland JC120-H Output Transistors-vbem-plot.png  
Old 27th December 2016
  #38
Gear Head
 

SO! did everything and finally plugged in a guitar to the faulty channel. Finally works but is noisy as all get out. Sounds like cap noise I suppose. I can gladly record a video for you to observe. I was just wondering from you looking at the schematic if there was any cap that might be introducing some noise into the right channel? Tried left channel too. Works fine but noise floor does seem a bit high I suppose. I'd have to have my buddy here to judge. My apartment does have some ****ty electricity problems from time to time. C82, C83, and C93 all appear leaky. 92 looks like it is part of the chorus.(still noisy with chorus off), 83 is connected to IC6. Hmmmm. I'm disappointed of course but so glad that it is at least working/not popping fuses/etc!

Edit: Redid the cap/resistor on Q32. Made sure it was perfect, cleaned around it. Still noisy. Guess next I'll try and redo the soldering and clean a lot on the new output transistors. I'll update that progress later today.

Second Edit: Redid solder connections and cleaned around the new output transistors. Noise seems better than last night but definitely still there. It sounds just like cap noise that I've experienced on several tube amps I've worked on. I can't find any change in noise when gently tapping on every single cap/component in the right side. Sometimes it will be dead quiet for a couple of minutes and then start again, also why I feel like it is an electrolytic cap going bad. Guess I should just buy every cap in this thing and replace them all lol. Left side seems very quiet but note that I did experience this noise several times throughout working on the amp and only having the left speaker out plugged into my cab. As I'm typing this, right amp is dead quiet but 2 minutes ago, it was crackling at a medium volume.
Old 29th December 2016
  #39
Gear Addict
Replacing all remaining electrolytics is actually what I would do at this point. There are a few 1u/50V bipolar ones (marked "N.P." in the schematic) that I'd replace with film caps, these are very compact, fit nicely in place of electros and don't go bad with age. For the others, I'd replace all the 16V ones with 25V versions like we did with C6 and C7. Not all of them see the full 15V supply voltage but the difference in price / size is going to be peanuts (size-wise, a modern 25V electro is usually smaller than an old 16V one) so there's no point in figuring out how much voltage each one is seeing and you'll be sure not to have any caps working too close to their max voltage rating. For recapping I use any of Panasonic FC/FM/FR/FK series depending on price / availability, they are all good.
Old 7th January 2017
  #40
Gear Head
 

I know you are just shaking in your boots for an update lol but yeah, finally got some parts in from Newark. I forgot to order the NP replacements because I am an idiot, but I will get them soon enough.

So I had to do a live sound gig last night, and didn't get the package until around 5 o'clock, so I only had time to replace a couple of parts before I had to leave. I chose the 'obviously' old looking rectangle electrolytics in the power supply(c94) and in the right channel(c85). Before I left, I plugged her in and wow, with those two cap replacements alone, noise DRAMATICALLY decreased, but still a faint sound of the ol' bacon frying.

Today I accomplished the task of all the remaining electrolytics(besides the 1uf NP, and all of the PF rating tiny electrolytics) and just tested it out. Stellar improvement if chorus is off. When switched to either chorus mode, some noise present again, so obviously the remaining noise problem is within the chorus circuit.

Thanks for all the help throughout this whole thing. Its been a long and frustrating journey. I also had great success with the TechSpray brand flux remover pen on this pcb board. I snuck some silver mica replacements on an order I did for a buddy, whose bassman I'm working on, for some of those tiny caps, but only the ones in the original faulty circuit.. maybe I'll put those in.

Question, what type of caps are these green ones that are in the jc-120? I've never seen them before. Also, that extra green(ground I believe) wire that has a loop on the end, am I correct from reading the schematic, that it is placed around the 'Low" input jack? I've been doing that since I took the amp apart but honestly didn't take notice until i didn't know what it was connected to.
Old 12th January 2017
  #41
Gear Addict
Hey pedalhoundz, finally back from a late holiday... Those green caps aren't electrolytics, they are polyester film caps and although they don't age like electros, if those two looked bad and replacing them gave you a noticeable improvement, they probably were bad. I don't see any other components in the chorus circuit likely to give you that kind of crackling noise apart from the NP electros, so I'd replace those first and see how it goes. And of course you can always do your chopstick test in that area, both above and below the PCB because a cold solder joint could also be causing trouble.

As for the green wire with a loop, I don't see it in the pictures... Which one you mean?
Old 14th January 2017
  #42
Gear Head
 

Quote:
Originally Posted by Cabirio View Post
Hey pedalhoundz, finally back from a late holiday... Those green caps aren't electrolytics, they are polyester film caps and although they don't age like electros, if those two looked bad and replacing them gave you a noticeable improvement, they probably were bad. I don't see any other components in the chorus circuit likely to give you that kind of crackling noise apart from the NP electros, so I'd replace those first and see how it goes. And of course you can always do your chopstick test in that area, both above and below the PCB because a cold solder joint could also be causing trouble.

As for the green wire with a loop, I don't see it in the pictures... Which one you mean?
I included a picture of the green wire with the loop. It connects to the distortion foot pedal jack on the back of the chassis.

Yeah I replaced the electrolytic c94, the .047uf in the power supply and c85, the .01uf connected to R147, and point 36 on the schematic. Still gotta get the 1uf NP's in.

So I was all excited to have my bandmate check out the mostly done jazz chorus so I took it to the studio last night for practice so he could turn it on and see it not blow fuses/work haha. I hadn't played it anywhere besides my apartment complex.. he immediately was like wow, this has like 1/4 the volume that it normally has! hmmm back to square..2 or 3.. at least not 1 because it stays on and works.. Getting his opinion on the volume(we tried both power amps) he said its normally mega loud, even mono 60w on about 4 on the volume pot.

I checked all solder points all over the thing, including the output transistors today and all seem good.

I have a few questions about what would affect the actual volume of the amp.

1. Do only output transistors determine the actual output volume?
2. Do driver transistors or VBE multiplier or VAS have any bearing on that?
3. Would the fact that BOTH power amps having a volume drop narrow it down to something more obvious?
4. Can something in the 2 caps that i changed have a bearing on volume? The values were both exactly what the removed component was, physically, and what was listed in the schematic except the .01uf(that does connect to the right speaker out) has a 630v rating instead of the removed 200v rating. Both were film caps instead of electrolytics..

So while I'm typing this I smelled something and i went over to the amp quickly to shut it off! R8 is burnt up! What the heck?? its a 3.3 ohm that is connected to Q3, C8(installed like a week ago, polarity correct, solder points good), and the reverb unit?? My god the problems won't end lol.
Attached Thumbnails
Roland JC120-H Output Transistors-img_0860.jpg   Roland JC120-H Output Transistors-img_0857.jpg  
Old 14th January 2017
  #43
Gear Addict
That's a ground wire and by the looks of it yes, it goes on one of the input jacks, it doesn't matter which one since both are grounded. And as I said C94, C85 and all those green caps aren't electrolytics, they are film caps, at least according to the schematic and their values. Unless the ones you replaced were actually electrolytics, in which case I bet someone who didn't know what they were doing had already been in there replacing stuff, in which case I'd take a good look at anything that doesn't look stock and make sure there aren't any other wrong parts installed...

None of the transistors you replaced have any bearing on the gain of the power amps and neither do those caps, the gain is set entirely by the feedback resistors R50/R40 and R111/R112. In any case since the volume drop affects both channels, I'm pretty sure the problem is somewhere upstream of the power amps. Almost any of the resistors in the preamp, from the input jacks up to IC3b, could change the gain, and also any of the pots if one leg has become disconnected or something like that.

The circuit with R8 is the reverb driver and it's like a little power amp where IC1b works as the differential pair and VAS combined, D3 and D4 is the bias spreader (= Vbe multiplier) and Q3/Q4 are the output transistors (there are no driver transistors, not needed). I would test those transistors, I wouldn't be surprised if one or both are faulty. If necessary you can replace them with KSC2690 and KSA1220. Even if only one is faulty, replace both. Also check (maybe directly replace if you haven't, though it looks from the picture that you already have) all the other electrolytics (C13, C14, C15, C5) and double check again that you replaced C8 correctly. Finally check D3 and D4 and if necessary replace (both) with 1N4148. By the way none of this would have any effect on overall gain either.
Old 4 weeks ago
  #44
Here for the gear
 
SpiderMullum's Avatar
Quote:
Originally Posted by pedalhoundz View Post
I included a picture of the green wire with the loop. It connects to the distortion foot pedal jack on the back of the chassis.

Yeah I replaced the electrolytic c94, the .047uf in the power supply and c85, the .01uf connected to R147, and point 36 on the schematic. Still gotta get the 1uf NP's in.

So I was all excited to have my bandmate check out the mostly done jazz chorus so I took it to the studio last night for practice so he could turn it on and see it not blow fuses/work haha. I hadn't played it anywhere besides my apartment complex.. he immediately was like wow, this has like 1/4 the volume that it normally has! hmmm back to square..2 or 3.. at least not 1 because it stays on and works.. Getting his opinion on the volume(we tried both power amps) he said its normally mega loud, even mono 60w on about 4 on the volume pot.

I checked all solder points all over the thing, including the output transistors today and all seem good.

I have a few questions about what would affect the actual volume of the amp.

1. Do only output transistors determine the actual output volume?
2. Do driver transistors or VBE multiplier or VAS have any bearing on that?
3. Would the fact that BOTH power amps having a volume drop narrow it down to something more obvious?
4. Can something in the 2 caps that i changed have a bearing on volume? The values were both exactly what the removed component was, physically, and what was listed in the schematic except the .01uf(that does connect to the right speaker out) has a 630v rating instead of the removed 200v rating. Both were film caps instead of electrolytics..

So while I'm typing this I smelled something and i went over to the amp quickly to shut it off! R8 is burnt up! What the heck?? its a 3.3 ohm that is connected to Q3, C8(installed like a week ago, polarity correct, solder points good), and the reverb unit?? My god the problems won't end lol.
So what happened??? It is 3 years and 3 days later :-) I am fixing a JC-120 for a mate and also having all sorts of fun with 30 year old parts!!
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump