Update and finalize Rotel RB970BX post

content/electronics/repair-rotel-rb-970bx.md

@@ -1,11 +1,10 @@
 +++
-date = "2017-11-06"
+date = "2017-11-13"
 title = "Repair: Rotel RB-970BX"
 tags = ["rotel", "repair", "audiorepair"]
 categories = [ "Audio Repair" ]
 description = "Another lucky find: Rotel RB-970BX Power Amplifier with one defective channel. Let's fix this!"
 slug = "repair-rotel-rb-970bx"
-draft = true
 +++
 
 ![The Rotel RB-970BX](/img/rotel-rb970bx-front.jpg)
@@ -14,58 +13,70 @@ I did another great find on _Marktplaats_: a Rotel RB-970BX. This is a power amp
 volume or tone controls whatsoever. You'd normally pair a RC-970BX pre-amp with one or two of these
 power amplifiers. 
 
-The amp can run in stereo mode at 60 Wpc or in _bridged mono_. In bridged mono you can use the entire
-amp to amplify _one_ channel and get 180 Wpc. Of course, for stereo, you'd need two of these
-amplifiers in your setup. 
+The amp can operate in _stereo_ at 60W per channel or in _bridged mono_ for up to 180W of power. In bridged 
+mono you can use the entire amp to amplify either the left or right channel to a single speaker. Of course, 
+for stereo, you'd need two of these amplifiers in your setup. 
 
 ![Bridged Mono on the RB-970BX](/img/rotel-rb970bx-bridged-mono.png)
 
 ## Diagnostic
 
-I bought this amp with the remark that one channel did not produce any sound. This could be due
-to a number of reasons, but most likely one or more components in the circuit will have failed. 
-
-Before testing the amp to determine _which_ channel is not working, I opened up the Rotel and 
-checked for any obvious failures: burnt resistors, exploded or leaking caps, and blown
-fuses. Besides some oxidation on the mechanical parts and jumpers I could not find anything 
-out of the ordinary.
+I bought this amp with the remark that one channel did no longer produce any sound. This Rotel
+amp was a complete unknown to me, but my guess was it'd be easy enough to fix. 
 
 ![The insides of the RB-970BX](/img/rotel-rb970bx-pcb.jpg)
 
-After that I did some quick measurements (still with the unit powered off): 
+Before powering up an unknown device I always take of the cover and do a visual inspection. 
+This amp looks really clean and lightweight. There's not a single IC in there! I checked for
+any obvious failures, like buldging capacitors, fried resistors or burn marks on the PCB. 
+Nothing looks out of ordinary, so that means there's a possible issue with any one of the
+transistors on the board.
 
- * Check for shorts on the transistors on the heatsinks
- * Resistance of the bias resistors (the larte white ones)
+The power transistors, those mounted on the heatsinks, are prime candidates for failure due
+to the thermal stress they get. With my multimeter I did a quick continuity test to see if
+any of these were shorted out - all appear to be fine. 
 
-Both of these showed that none of the power transistors had shorts, and that all
-bias resistors were well within spec (0.22Ω).
+Another quick test I performed is verifying the bias resistors (the large white ones). These
+are high power, low resistance. All measured in spec at around 0.22Ω - so no open
+or short circuits there either.
 
 Next I powered up the amp for the first time - with nothing connected to it. I measured
-for DC on the outputs but both were in the mV range, so nothing out of the ordinary 
-there. 
+for DC on the outputs but both were in the mV range. Right was close to 0V whereas the
+left channel had a few mV of DC. Might be nothing, might be something. 
 
-With that I continued to hook up my dummy load and test which channel - _if any_ - was not
-functioning. The right channel was working just fine, but the left channel appeared to be 
-dead as a doornail. 
+Next I hooked up a dummy load (8Ω 200W in power resistors) and applied a 1kHz sine 
+wave to the input. With my oscilloscope I quickly determined that the right channel was
+in working order. Left was out completely. 
 
-Because the amp is pretty bare bones, my next step was to test the transistors on the
-PCB for the left channel. The first one I tested (Q615, a humble 2SC2910 NPN transisor) 
-showed a short between _emitter_ and _base_.
+Since I already checked the major culprits of a dead channel, I continued to check the
+other transistors on the left channel circuit. The first one I tested (Q615, a 2SC2910 
+NPN transisor) showed a short between _emitter_ and _base_.
 
-A quick probe around showed no other damaged parts. So, I ordered a few new 2SC2910's from 
-a local supplier and waited for them to arrive.
+A quick probe around showed no other damaged parts. Resistors showed their correct values
+or I would get values that indicate a capacitor getting charged. All other transistors 
+checked out ok as well. 
+
+After taking out Q615 I put it in my component tester and instead of an NPN transistor, it
+showed up as two diodes. 
+
+![Defective NPN transistor 615](/img/rotel-rb970bx-npn-defect.jpg)
+
+This is what a working 2SC2910 transistors looks like on my tester:
+
+![Replacemenet NPN transistor](/img/rotel-rb970bx-npn-okay.jpg)
 
 ## The Repair
 
-The repair is as simple as replacing the damaged transistor.
+The shorted transistors was quickly replaced (a local supplier has the exact model in stock). 
+This model has a removable bottom panel that gives easy access to the bottom of the PCB. 
 
-Since I'm there, I've also cleaned the oxidation from the jumpers and mechanical parst, 
-cleaned the PCB and replace the four capacitors for good measure.
+![Bottom view during repair](/img/rotel-rb970bx-bottom-repair-annotated.jpg)
 
-## Conclusion
+While I had the unit opened up on my bench I also to the opportunity to clean the insides using
+some rubbing alcohol and compressed air.
 
-A single failed transistor took out the entire left channel. Replacing a €1 part revived 
-this power amplifier to its former glory.
+After verifying correct operation with a dummy load and my scope I hooked up pair of speakers and
+got to enjoy the beautiful sound of this Rotel amp.
 
 ## Links