Gallien kruger 2001rb.....oppseedaisy!

[nevetslab]

This week I’ve tackled a Gallien Kruger 2001RB Bass Amp that was labeled Low Output/Distorting. In the past couple years, all of the GK amps in our inventory belong to GK, parked here for client visibility as we rent to working professional clientele. And, I had been advised not to run the amps that hit the repair shelves thru the shop, as it was preferred to send them back to GK (however long it takes to send them out). That didn’t bother me, as they can be a real headache to service.

But this one 2001RB WAS sent to my shop, this one being needed, as we’re currently sub-renting 3 of them. So, I opened it up to have a look. Initially, I was getting output signal on both channels, just looking on the scope, and it appeared the problem was in the Overdrive Channel. But, after a bit of probing, it looked like the preamp was working fine, so I connected a speaker and plugged in the shop bass to have a listen. Initially, I was getting output from both channels, and dialed in suitable Clean & Overdrive signals as I switched between the A & B preamp channels. It had been switched to Bridge Mono on the rear, so I switched into that mode to see if I’m getting what I should. Then, I begin having the output cut in and out and distorting badly. I switched back to Dual Mono, and was still having bad intermittent behavior, on both power amp channels.

Sigh…….

Put the bass aside, pulled the amp apart to get the power amp PCB onto the bench for close inspection. As I suspected…..Lead-Free solder, and all the connections on both power amp sections showed de-soldering around the bridge rectifier leads, as well as on the output connectors….the NL4’s and the phone jacks. And, as is usual for GK, the component leads are cropped real close to the PCB with their lead trimmer, and, as I just hate about their PCB’s….all the solder pads are tiny annular rings…or the solder mask where there IS substantial copper, is likewise tiny annular rings. I did the best I could on all the connections, cleaned up the flux residue, and put the amp back together.

Brought it up on the variac, monitoring both power amps for any DC levels that might have changed. I got to around 70VAC mains voltage when the relays pulled in and all hell broke loose, pulling loads of current, as I quickly turned it off and ramped the variac back down. What did I do….it didn’t do this before.

I looked over the connections, and then I saw it. The 7-wire secondary connector on the left channel’s Power Input had been installed one-pin over, while the other channel was fine. I unplugged that connector, left it out, and tried to ramp it up again, leaving the Right Ch plugged in. I got no output, and the AC mains was only 130mA. I unplugged the second channel, ran up the variac again, heard the relays click in, now just powering up the xfmr and the control PCB. Low current.

As I pulled the amp apart again, I noticed the in-rush current limiter Thermistor on the AC mains control board was damaged, missing a chunk of the Thermistor. I pulled the control board out, looked at the Thermistor. Schematic labeled it as a CL-90. 120 ohm/3A part. I knew I didn’t have any, but, after thinking about it, I put in a 100 ohm 3W MO resistor, carefully placing it and insulating the exposed lead, mounted vertically. Temporary fix, mind you.

With the main PCB still in place, I went to power up the amp. Right Ch still drawing low current. I had sensible voltages on the bridge outputs. I connected the Left Ch secondary connector, properly indexed, and found I wasn’t getting the proper readings. I didn’t have the high voltage readings on the two upper tier’s, AND, I was seeing like +/- 80VDC on the low voltage bridge in the middle, instead of around +/- 38VDC. I disconnected the Left Ch, and tried to get output from the Right Ch.

Nothing. Looked at the schematic, and found the Left Ch power supply is the source to the bipolar low voltage. So, pulled the main board back out, and began looking. Then I noticed four pico-fuses, two per channel. Had to go thru a lot of the revisions before I came upon a schematic (Rev A1) that showed them on the +/-38V supply connection following the output of the bridge. 15A fuses. I looked and found I still had some from some prior encounter with an 2001RB. Replaced them and tried again.

Big mistake! Loud racket, sounds of some parts cracking open and smoke flowing out from the pair of 50W tweeter amps….LM3886T power amp IC’s. Sigh………

Pulled it apart again, now feeling really stupid, not further investigating what caused those pico-fuses to open. It seems I’d run into something like this once before, mis-connecting one of the 7-pin connectors, but didn’t end up in a nasty failure like this has turned into.

After pulling everything out, looking at the circuits, and, seeing a 40 ohm short across the low voltage bridge outputs in both directions….that pointing towards both the Tweeter amps and perhaps the LM317T/LM337T bipolar regulators, I set the power amp aside, and broke out an HP dual tracking supply to check out the preamp assembly. After finding a place to hook into, I slowly raised the supply voltage, watching the current draw of the supply on one of the meters….drawing loads of current as I increased the supply voltage, and, began smelling something cooking, as well as feeling some of the IC’s getting hot. Shut it down, now further defeated.

Great. Mis-connected one xfmr connection by one pin, and have done MAJOR damage, just trying to solve this intermittent output.

I cut the two Tweeter amps off the PCB, then lead by lead, removed them from the top so I didn’t destroy the tiny solder pads. Then looking over the +/- 15V supply regulators, I found the control xstr Q49 and it’s diode D38 had failed. Probably meant the LM337T is no longer working. I set up the amp PCB assy to see if either of those two supplies worked, after I replace the two parts. Sure enough, the Neg supply isn’t working, though the Pos supply is.

GK’s construction is such that there’s a full sheet of Bergquist K6 insulator between all of the power xstrs, drivers and regulator, and after years of operation of the amp, Thermoset HAS set in on the insulator sheet. Only way to get at the LM337T is to remove the heat sink, and ripping the insulator sheet to pieces in the process. Sigh…..

This has been an interesting day. I can’t blame it on my eyesight, it now having been restored recently. I was sure I had properly connected both xfmr wire connectors to their headers…but…..I DIDN’T DOUBLE CHECK THAT! I usually have such high success on repairs. But this one.....!!!!###?!!

The Power Amp schematic is Rev A1, it having the pico fuses on the +/- 38V supplies

2001RB_Power_Amp.zip
2001RB_Pre-Amp.zip
2001RB_Switch_Board.zip

[glebert]

Seems like everyone is having crappy luck with repairs these days. Must be the heat.

[nevetslab]

Failure mode of gk 2001rb explained

Refer to the GK 2001RB schematics provided in the previous thread of 8/8/18.

When the power transformer secondary 7-pin connector was mis-connected to the power supply header J16, off by one pin (Header Pin 1 open, mating connector P16 pin 7 open), it placed the upper 24VAC coil between an AC input to BR6 and an AC input to BR5, then placed the 48VAC C/T coil between the GND connection between C81 and C66, the C/T to the other AC input of BR5, and the other end of the 48VAC coil to one of the AC inputs of BR4. The lower 24VAC coil harmlessly connected to the other AC input of BR4 and it’s other side of the coil was open circuit.

When the AC mains relays closed (AC mains ramped up via metered Variac), there was an instant high current path thru both 15A pico fuses F4 and F3, which opened immediately.

That left 48VAC connected between Power Supply GND and an AC input to BR4, where it’s (+) terminal was reverse-charging C66 every half-cycle with around 67V positive peaks thru BR4 bridge rectifier. That was also feeding the input to the LM337T –15V regulator, which failed. It was also feeding both of the LM3886T power amp IC’s, the Tweeter Amps of the 2-channel power amp section.

There was no current flow thru BR5 since the two pico fuses F3 & F4 had opened.

I hadn’t assessed this damage yet, when I finally saw the cause of the sudden high current flow during ramping up the AC mains with the variac. I unplugged and re-connected the secondary plug so it was now properly connected, and tried to power up again. Nothing happened with the amp at this point, other than I had no amplifier functions, no DC offset from the power amp, though didn’t yet know there was no lower supply rails present. I found that out after removing the amp assembly again.

I also discovered the in-rush current limiter on the AC mains PCB assembly had cracked, discharging a chunk of the CL-90 NTC Thermistor, which is in series with the primary of the power xfmr until the relays pull in, shorting it out and applying full current to the primary. I replaced the broken CL-90 with a 100 ohm 3W MO power resistor for a temporary fix.

When I found the two pico-fuses, both 15A parts open, I removed and replaced them with the same value, having some of those on hand from previous encounters with the GK 2001RB. I put the amp back together to try again. Having had the AC mains all the way up before finding the open fuses, and it was only drawing around 150mA from the wall, I powered up the ‘repaired’ amp at full line voltage.

That’s when the unit emmited loud noises….cracks with parts breaking and smoke coming off of both tweeter amp IC’s. The loud crack I had heard was the LM317T breaking into pieces, still bolted to one of the two heat sinks.

With the preamp connected to the bipolar outputs of the two IC regulators Q24 LM317T and Q59 LM337T. I have to presume the LM337T was now passing full -38VDC input voltage thru the part and onto the preamp, and killed most or all of the TL072’s and probably the DG419 Analog Switch IC’s.

After pulling the amp apart again, now licking my wounds, I connected my HP 6227B Dual Tracking power supply to the preamp assembly, to see if I had killed the IC’s or not. The DC current meter climbed immediately as I advanced the current, having pre-set the DC voltages. I smelled the IC’s beginning to cook, as well as felt a couple of them with only a few volts applied, it pulling over half an amp.

All of the amp is soldered with Lead-Free solder. And, as it typical of the 2001RB and similar amps, the solder pads are layout software defaults….60 mil pads. So, that produces very tiny annular rings on both sides of the PCB. Plated thru holes. I’ve had miserable success with lead-free solder joints and my Pace SX-70 & SX-80 Desoldering irons. First, it requires higher tip temperatures, and, it tends to just mush into a solder joint instead of melting it and allowing the vacuum system to suck it up. I haven’t come to a successful solution with the Desoldering and lead-free solder, so I’ve had to resort to using 0.098” wide solder wick.

And, of course, with lead free solder, tiny annular ring solder pads and plated-thru holes, there has been copper left on the IC & component leads during the de-soldering procedure of all the IC’s on the preamp, as well as a handful of lifted solder pads, though none having traces attached on that side. I probably should have just cut the leads off the IC’s and then removed the leads from the component side one by one, heating up the pad, instead of desoldering each IC and fighting them, removing them all one at a time. I didn’t yet know if I had only a few IC’s shorted or all of them.

With the two LM3886T Tweeter Amp IC’s, I cut the leads from the body, removed the body, then one by one, gripped the lead and heated up the solder joint, removing the lead. Then, used the solder wick to get the solder out of the holes. Did the same with the LM337T IC. I did manage to get that IC out off the heat sink without damaging the Bergquist K-6 Insulator sheet. I had removed it's mounting screw (having nylon shoulder washer), and loosened all of the 'local' power xstr mtg screws so there would be less clamping force on the LM337T body) Getting the new part into place was tedious, needing to bend the leads up and over, allowing them to get forced into the solder pads, and then work the IC body forward without breaking the 3 leads, while getting the leads pulled thru. Succeeded the first time, and now have the Bipolar supplies back up and running, verified with the external power supply applied.

I’ve been repairing this since Tuesday afternoon, and finally got the preamp board loaded with nice gold-plated machine-pin sockets for all the 16 IC’s on it, as well as replacing the electrolytic caps, since the preamp was subjected to high-enough voltage to have killed nearly every IC on it.

When I sketched the mis-connection circuit, I saw the lower negative-supply cap C66 had been reverse-charged by BR4 (+) output and the full sinewave applied to the (+) terminal of the cap. I removed both C66 and C81 from the power amp PCB. I had a spare broken GK 2000RB amp module, and extracted it’s four 15,000uF/50V supply caps. Checked those as well as the suspect C66 cap on my GenRad bridges…first on the Digi-Bridge to see if there was anything immediately obvious (wasn’t), then on the 1617A Capacitance Bridge, which has a 0-600VDC supply for biasing caps being measured, as well as checking it for leakage. C66 did reveal high leakage current vs the other cap C 81, and the other caps I rescued from the 2000RB assembly, they measured better in all regards, same OD, though 5mm taller.

So, I’m now in the final phase of repairs on the preamp….swapping out suspect electrolytic caps that would have been over-voltaged during the failure.
So, what have I learned from this exercise? MAKE SURE THE CONNECTIONS ARE CORRECT!! Just being one pin off on the Left Ch’s AC xfmr connection has caused major damage. I still haven’t powered up the system after the repairs, but didn’t find any of the power amp circuits damaged…..thus far. I still have to look further on that.

I’m pretty sure I did this once before, but it must have been on the Right Ch, which did blow the pico fuses, but there wasn’t any further damage, since that channel doesn’t source the system voltages to run the preamp and tweeter amps like the Left Ch does.

I still don’t know if I’ve cured the intermittent output, which I was about to try and verify, before discovering this mistake of being one pin off.

[nevetslab]

I hate gk's pcb layout defaults & lead-free solder!!!!

This is the 5th day in recovering from the calamity that awaits any other careless well-meaning soul who mis-connects the transformer secondary 7-wire power connection to the Left Ch. Normally, replacing radial lead electrolytic caps is simple and straight forward. When they're NOT in tiny annular 0.060" dia rings around the lead-free soldered leads. Can't use my Pace SX70 desoldering tool for it, and the success I've had last Friday and today with getting the IC's removed, and now finally the remaining caps that got subject to excess DCV.....it just doesn't seem to matter how much care I take in trying to not damage the solder pads. I've had to patch the circuits back together from the bloody solder-side pads lifting off the board, breaking the plate-thru connection to the other side, break away from the trace on the solder side pads, etc. It takes an aged GK2001RB like this, which failed enough to leave you questioning your mechanical ability to do such repair work. I'm rapidly reaching the point to where if it comes back up, and is still behaving with the intermittent output......then THAT'S WHAT IT IS MEANT TO DO AT THIS STAGE IN THE PRODUCTS' LIFE!!! GRRRRRRR!!!

[nevetslab]

Powers up, goes into mute. Pwr amp front end turns off

After restoring the preamp, having to replace every IC, since the neg 15V regulator (LM337T) failed/shorted, producing -38V or so, it took out every IC in the system running on +/- 15V. The +15V regulator survived (LM317T). Took out JFET's in the preamp as well, and stressed enough electrolytic caps to warrant replacing all of those. I now have the preamp back up and running, and the amp does power up without blowing smoke and fuming. And, just to tickle my fancy, for the briefest moment, I'll get output...200-300mS before the amp mutes, and the front end of the power amp turns off.

I've looked at the controller IC on the AC mains PCB, which directs operating logic for the power amps....if any DC offset above their selected threshold (haven't found that), it can shut down the amp via the mute line. I did replace the xstr Q3 on the AC mains board, as I was seeing the output on the cable go high. That line connects to Q16 (MPS A56) on the power amp board, whose PNP emitter ties to +15V, and the collector ties to the two channels' constant current source biasing/mute line.

While no published voltages are shown on the schematic, normal current for the front end of the power amp stage is set by the B-E junction of Q43, across R87 (200 ohms). If that voltage is 0.6V, then the front end is 3mA, split in half at the diff pair Q39 (2SC3381 dual NPN).

Watching the voltages in the switched current source for this amp (Q42, Q40, Q43 CC source, and Q50 which controls the muting of the tweeter amp, not yet installed), I see the current source charge up, briefly letting signal pass thru the amp before something shuts it all down, and the 3mA current goes away. I see the same results looking at the bias test points on the lower emitter's of the 2-tier output stage.

The Left Channel was the one I made the connection mistake on, being one-pin off connecting the three secondary coils of the power xfmr to. The Right Ch was connected correctly, but thus far, it too isn't passing signal. I find that odd, since both channels' Mute/Turn-on lines are controlled the same, and that line remains high @ 13.3V to both 200k resistors of the respective turn-on circuits.

When the amps start up, watching the output terminals, I see the Left Ch's DC offset go as high as 40mVDC, then drifts back down, settling in the 10-15mV range, though the amp is turned off quiescently. The Right Ch doesn't go that high, stays in the 10mV offset range. I never see any signal pass thru it.

I've replaced the IC U1 that has one cell U1B functioning as the unity gain inverter to provide signal for bridge mono, and U1A that serves as the DC Sense circuit, that drives Q25 thru a diode network over to the controller IC.

The Mute signal from the controller IC, driving Q3 thru a 10k resistor, shows up at the base of Q16 that is the mute control line/ turn-on source for the two channel's constant current sources. It remains turned on, pulling up the two 200k resistors R85 and R49, that source level being +13.3VDC. So, the controller isn't issuing a shut-down command.

I haven't yet installed the two Tweeter amps (LM3886T), thinking I'd hold off until I have the main power amp restored. I would have thought, prior to this morning, that the Right Channel, in spite this Left Ch only showing signal momentarily, that it would be operating. Not so.

When I went home last night, still defeated, though encouraged not seeing it belch smoke & fire any more, I looked at the turn-on circuits for the power amp, following the traces on the PCB layout (Rev A1 in the docs, though the board I have is a Rev B, no doc's for it on hand). I discovered I'd been plugging the 4-wire connector from the I/O board into J11 instead of J9. Having seen signal thru the I/O board (replacing it's IC's as well), that plugging the cable into J9 would result in a working amp. Wishful thinking.

So, I'm puzzled. I haven't yet replaced CC source Q43, switches Q42 & Q40 that feed the diff pair Q39, nor Q50, which does have a 10uF/50VNP cap at the top of R121 that connects to the MUTE terminal, nominally pulling around 500uA out of Pin 8. I saw both LM3886T's smoke when the amp went down hard, and have since been clipped off the PCB and the remaining pins removed. All the resistors for the LM3886T are underneath the two heat sinks.

I've also run out of MPS A-06's, about to place an order for more, but do have 2N5551's on hand that would suffice as replacements.

Any ideas?

[glebert]

Any ideas?

Do they sell replacement boards for this? Are you working with GK?

[nevetslab]

I haven't contacted GK on this thus far. In fact, I haven't even checked to see if the 2001RB is still a current product. Just looked, yes, it is a current product, so they would be supporting it.

[nevetslab]

Removed Q50 & Q26, power amps now working

Puzzling over the brief time the Left Ch turned on and then muted....I decided to remove Q50 and Q26, both of which are used as the Mute/Turn-on control path for the Tweeter amps. I left their associated 10uF/50V caps in place, which have 20k resistors between the cap and the collectors of those two xstrs now removed. When I re-installed the amp PCB and connected everything up, I monitored the voltage across the Constant Current resistor R87. It lit up and remained @ 560mV/2.8mA, as did the Right Ch R51, and now seeing the quiescent current/power on the power analyzer pulling 560mA/37W @ 120VAC. Turned it off and on several times, seeing the same results....time-delay turn-on and quick shut-down.

Plugged in my signal, and now have full output on both channels....thus far, just into the audio analyzer/scope, not yet connected to speaker with the bass as source, to see if I've made any change in the intermittent output behavior that began this whole service episode.

I also noticed one other thing while I had the main PCB out. On the output phone jacks, there is a tiny metal protrusion that's tied in to one of the Ground terminals. I noticed it wasn't really protruding out past the plastic mounting surface of the jack body. In re-soldering the terminal on the four jacks, I forced that terminal to stand proud of the surface.

I had forgotten about those little hidden features, but seem to recall having intermittent behavior before on one of the RB series amps, that was associated with those ground pins.

So now, I'm about to tempt fate, tear it down again, remove the amp PCB assy, replace those caps C26 & C63, and the xstrs Q50 & Q26, and then install the pair of Tweeter amps U2 & U3 (LM3886T), and see what happens.

[Enzo]

I went off to look for your part numbers, and see they change from revision to revision. SO I can't imagine why we haven't called GK yet to get the correct schematic for your version.

[nevetslab]

I went off to look for your part numbers, and see they change from revision to revision. SO I can't imagine why we haven't called GK yet to get the correct schematic for your version.

You're right.....I do need to contact them to see if I can get the current revision doc's. Not being an authorized GK service center, I don't know what their policy is.

While I had the amp idling for a while, I checked the bias on their test points, and bumped it up to their 10mV figure across the lower tier emitters of the output stage (was sitting at 6.7mV). That made minimal change in AC mains current.

Anyway, having gone ahead and pulled the power amp PCB assy back out, removed the former caps C63 and C26 that are at the Pin 8 terminals for the Tweeter Amp LM3886T, installed replacement current source xstrs for that same pin (using 2N5551's having run out of MPS A-06's), and then installed the new LM3886T Tweeter amp IC's. When I put it all back together and powered up the amp, I did see an increase in AC mains current draw from 560mA to 760mA. When I went to verify output from the amp, I no longer had output from either channel. Hmmmm.... Checked the voltage across the constant current source of the two power amp channels at R87 and R51. Still have the same 2.8mA (560mV across 200 ohms), holding solid, so no hint that the amp is in mute.

I had to leave to go catch my bus heading home, so I didn't get a chance to see if something else has occurred, such as loss of the +/- 40V supply, which sources the +/- 15V regulators that run the preamp. I didn't see any abnormal DC output from the two amp channels....just no longer see output signal.

I've been using Rev A1 schematics in working thru this Rev B amp, that most closely resembling the build I have on the bench. It sure has been an interesting adventure, though I have better things to do than be abused.

[Enzo]

I suspect the rev A1 would work fine for thought experiments, but the part numbers probably won;t agree with your board. But the boards ought to be very similar, so with the A1 layout drawing you can determine A1's Q1234 is this one on your board,, even though your B board is numbered Q2345.

I don't know if they still make them, but the 2001RB came along over 15 years ago. In my experience, GK has never been tight with their documents, especially on older stuff. They will either send it to you or not, might as well find out as soon as possible.

This circuit is really pretty similar to the one I explored recently on the 700RB. There are various mutes.

[Leo_Gnardo]

You're right.....I do need to contact them to see if I can get the current revision doc's. Not being an authorized GK service center, I don't know what their policy is.

Let'see if I have this right. GK provides amps to Center Staging for rental, in order to promote their product? Then it only stands to reason they should set you up as an authorized repair center with fast access to all parts & schematics, or be responsible enough to fix or replace their amps pronto when they go bad.

Given your expertise in fixing amps of all sorts it makes more sense for GK to be cooperative in having you be an authorized service center. But the way companies run things these days, who knows if they'll have enough sense to do the right thing. And if not, pile the busted amps on THEIR doorstep and let them sweat over 'em.

[Enzo]

Leo, be fair, read the first paragraph of post #1. Sounds like GK was being supportive. I know if my shop had a factory keeping my shelves full and not tying up my bench, I;d be happy as a clam. Apparently until this one GK has been doing all the repairs. They can do that and see what sorts of failures the product endures, they can see what happens later in product life, not just new ones. ANy repairs will include any circuit updates and ECOs. Send them my duds, they send me working amps.

[nevetslab]

On my way home, puzzled over NOT finding output signal again, after all I've gone thru, and only added the last tidbits to complete the surgery, having restored the preamp, power supplies, front end constant current source to run the amps, AND, finding the proper voltages on the front end of both channels. That led to conclude the +15V regulator was ok, the low voltage supplies and high voltage supplies were ok in order to have the current sources reading correctly. Had to be something else....possibly Jack normals on the I/O board?

I re-patched the amp and powered it back up this morning, first verifying I still had the preamp working. Yes....solid signal on the Effects Send jack. And, had solid signal thru the Chain output. Moved the cable to the output of the left channel, and like magic, I NOW had output. Great news, but troubling, as it doesn't explain why I didn't see this at the end of yesterday. Verified I had signal on both Left & Right Channels, Dual Mono mode. I hadn't yet checked the tweeter amps, and did that next. Both tweeter amps work now. When I began this a week ago Tuesday, I only had tweeter output from the Left Ch, but not the right, and the Left Ch had a DC output level of around 0.35V. Now, both are in the 10mVDC range, typical.

I connected the test speaker and my bass to see if the intermittent & distorted behavior was gone....also where I started. Now solid as a rock, no intermittent nonsense. Connected the horn tweeter in my rig, dialed that in, and had the usual string noise/snap it ads, from both channels, so that's great news. Disconnected the tweeter cable, switched to Bridged Mono, which was where I first found all the problems when I began. Now solid as a rock. Thank Heavens! The fans both work...hadn't had them connected/installed since I had first pulled it apart.

Now all that's needed are a correct pair of front panel handles. It showed up with no handles, stripped lower front panel screws, so all of the front end mechanically was all wanky. I'll have to order handles....tried one from the boneyard GK 2000RB, but those are different mtg centers. So, truss head screws and keps nuts holding the panel to the chassis, meaning NO Rack Mounting of this amp until there are handles. Least of my worries today. Now, everybody is pleading for their Vox AC30 & Ampeg SVT's, so FINALLY I can get back to work. This bloody amp has slammed me for....seven days running. Will I get paid for all of this education and adventure-land work? The education was for free, at the expense of nothing else getting done; NO....as I was the one who mis-connected the transformer connections. I will get paid for 5 hrs labor + all the parts cost.

[glebert]

Way to stick with it!