I have this old Kenwood amp that 2 transistors get super hot as soon as i give it power. tech guy went through it and can;t find what's wrong Maybe you guys are familiar with it.
Electronics guy can't figure out why these transistors are getting super hot on car amp....
- Thread starter daves66nova
- Start date
Without a schematic I'd be guessing, just as I'd be without know which transistors are getting hot.
While the TO-220's could be bad, what could be bad is the bias circuit. Perhaps someone turned up bias too high, that will cause heat in a hurry (and eventual damage). Or the bias circuit is simply bad. I don't see any big caps here so I'm wondering if everything is DC coupled and if so, the driver stage could be bad too, as it likely could impact bias on the final stage.
I've never liked reverse engineering boards, chasing where things go. Have you tried to find a schematic?
Lastly: those plastic cased TO-220's should be fully plastic around them. I don't think I've seen any with exposed metal... but you could do a quick check to make sure that there is no exposed metal in the screw hole and on the backside (just gently bend upward). I would not use a mica on them, although a small amount of heatsink compound is advisable (but if they are getting super hot immediately, not the problem).
While the TO-220's could be bad, what could be bad is the bias circuit. Perhaps someone turned up bias too high, that will cause heat in a hurry (and eventual damage). Or the bias circuit is simply bad. I don't see any big caps here so I'm wondering if everything is DC coupled and if so, the driver stage could be bad too, as it likely could impact bias on the final stage.
I've never liked reverse engineering boards, chasing where things go. Have you tried to find a schematic?
Lastly: those plastic cased TO-220's should be fully plastic around them. I don't think I've seen any with exposed metal... but you could do a quick check to make sure that there is no exposed metal in the screw hole and on the backside (just gently bend upward). I would not use a mica on them, although a small amount of heatsink compound is advisable (but if they are getting super hot immediately, not the problem).
No idea what any of this means but I'm eager to hear from people who do understand this if ChatGPT is on point or spitting out crazy talk.
If the transistors get very hot immediately on power-up with no signal, that almost always indicates that they're being driven into heavy conduction rather than simply dissipating normal idle power. The fact that a competent tech has already looked at it makes me wonder if the fault is something intermittent or subtle.
A few questions would help narrow it down:
If you can provide:
If the transistors get very hot immediately on power-up with no signal, that almost always indicates that they're being driven into heavy conduction rather than simply dissipating normal idle power. The fact that a competent tech has already looked at it makes me wonder if the fault is something intermittent or subtle.
A few questions would help narrow it down:
- What is the Kenwood model number?
- Which two transistors are heating up? (Output transistors on the heatsink? Driver transistors? Voltage regulator transistors?)
- Do they get hot on both channels or just one?
- Does the amplifier come out of protection or stay in protection?
- Bias circuit failure (most likely)
- A bad bias transistor (often mounted to the heatsink)
- Bad bias diode/string (STV diode)
- Open or drifted bias adjustment trimmer
- If the bias voltage is too high, the outputs conduct heavily at idle and get scorching hot.
- Shorted or leaky driver transistor
- Even if it tests okay with a multimeter, it may leak under operating voltage.
- I've seen this several times in vintage Kenwoods.
- Failed emitter resistor
- One of the low-value (0.22–0.47 Ω) white cement resistors opens or changes value.
- This upsets current sharing and can overheat one side.
- Differential input stage imbalance
- Mismatched or leaky input transistors can force the output stage to conduct continuously.
- Oscillation
- The amp may actually be oscillating at several hundred kHz or MHz.
- Everything looks "normal" with a meter, but the outputs get extremely hot.
- This requires an oscilloscope to diagnose.
- Power supply issue
- One rail missing or incorrect.
- Bad grounding.
- Incorrect repair from a previous owner.
Things I'd measure
If it were on my bench, I'd immediately check:- DC offset at the speaker outputs
- Idle bias current
- Voltage across the emitter resistors
- Base-emitter voltage of the hot transistors
- Rail voltages
- Look for oscillation on an oscilloscope
One thing that gets overlooked
Many Kenwoods from the 1970s and 1980s use 2SA726, 2SC1451, 2SC1345, or similar small-signal transistors that become noisy or leaky with age. They can test "good" with a handheld meter yet fail under operating conditions, causing exactly this type of excessive bias. Experienced vintage-audio restorers often replace these proactively.If the tech can't find it...
If the amp has been powered repeatedly while the transistors are overheating, I would stop doing that until the cause is found. Each power-up risks damaging the outputs or other components.If you can provide:
- the exact Kenwood model (e.g., KA-7100, KA-8100, KA-7300, KR-9600, etc.),
- a photo of the board showing the hot transistors,
- or the transistor part numbers,
Do you have a good short and solid ground no more than 18" same size as the power cable. that's not a new amp could be anything at this point bad resistors or capacitors. are they getting hot with no load no speakers hooked to the channels? could be an impedance mismatch trying to drive 2 ohm speakers and amp is rated at 4 ohms output?
maybe you have the wrong impedance speakers and they in the emitter circuit.
It's actually fairly on point. I can't speak to historical failure modes and I'm no repair tech but it's not a bad list to run through.
Similar threads
