As you said, the 2 meg ohm reading on a 48 k resistor was perhaps you didn't press down hard enough with the meter probe to push through the dirt. Generally checking resistors to see if they are nominal reading, or lower since current can leak out through attached semiconductors, is good use of time. Higher resistance than nominal is bad. While carbon comp resistors can go high value due to moisture (if not Allen Bradley or Sprague brands post 1961) I only replace those carbon comp resistors that measure right 100 kohm and over. At 220 kohm the hiss of carbon ccmp was pretty obvious, below hundred k I can't hear it.
I don't ever buy carbon film resistors. Why save $.004 over metal film if the freight on the box was $12? Maybe if metal film or oxide wasn't available in 1 meg or higher, but more likely I would try another vendor with metal film or oxide in stock. When I buy resistors, I buy $1 worth minimum and put the extras in stock. Ordering a lot of $.02 lines is a good way to earn a handling fee.
I use wirewound resistors both for emitter resistors on the output transistors, also the 3 to 10 watt resistor parallel the output inductor, and the zobel resistor series the .1 uf cap on the output. The output inductor prevents radio leaking in the feedback line, the zobel limits the tendency to oscillate with no load or capacitive loads like planar speakers.
Removing every transistor is as Mr Fahey said, dangerous because it can lift the land off the pcb. But after the output transistors shorted, it is likely the drivers, predrivers, and VI limiter transistors & diodes are shorted or open. I had one amp every semiconductor back through the high freq protection op amp and jfet were shorted. Shorted does not mean zero ohms at 2 volts of the DVM, shorted means collapses under reasonable voltage, say 5 or higher. Hence the Iceo<10microamps at 12 v test. Buying every transistor as a poster on the other thread suggested is time consuming because of all the old obsolete part numbers you have to cross to something available. The crossover tables from RCA, GE, NTE suggest a lot of garbage parts with all the specs left off the datasheet. TO5 transistors (ebc pattern mid power) are no longer available and D150 may have some. You can achieve same result as the >2 volt c-e leakage test by plugging in the amp with a light bulb limiter in the AC line and checking c-e voltage of every transistor & diode. In class AB amp, Vce 0 volts is wrong except for lamp drivers etc. Vbe not ~0.6 volts is wrong too except if a lamp driver is off. With light bulb in AC line you can also check capacitors to make sure they have a dc voltage across them. 0 volts on a capacitor, why did they install it in the first place? You cannot catch open capacitors with a limited power test. I've had capacitors open up in the VI limiter, also 50 v rated ceramics protecting the op amp power supplies in the front end.
Happy hunting.
I don't ever buy carbon film resistors. Why save $.004 over metal film if the freight on the box was $12? Maybe if metal film or oxide wasn't available in 1 meg or higher, but more likely I would try another vendor with metal film or oxide in stock. When I buy resistors, I buy $1 worth minimum and put the extras in stock. Ordering a lot of $.02 lines is a good way to earn a handling fee.
I use wirewound resistors both for emitter resistors on the output transistors, also the 3 to 10 watt resistor parallel the output inductor, and the zobel resistor series the .1 uf cap on the output. The output inductor prevents radio leaking in the feedback line, the zobel limits the tendency to oscillate with no load or capacitive loads like planar speakers.
Removing every transistor is as Mr Fahey said, dangerous because it can lift the land off the pcb. But after the output transistors shorted, it is likely the drivers, predrivers, and VI limiter transistors & diodes are shorted or open. I had one amp every semiconductor back through the high freq protection op amp and jfet were shorted. Shorted does not mean zero ohms at 2 volts of the DVM, shorted means collapses under reasonable voltage, say 5 or higher. Hence the Iceo<10microamps at 12 v test. Buying every transistor as a poster on the other thread suggested is time consuming because of all the old obsolete part numbers you have to cross to something available. The crossover tables from RCA, GE, NTE suggest a lot of garbage parts with all the specs left off the datasheet. TO5 transistors (ebc pattern mid power) are no longer available and D150 may have some. You can achieve same result as the >2 volt c-e leakage test by plugging in the amp with a light bulb limiter in the AC line and checking c-e voltage of every transistor & diode. In class AB amp, Vce 0 volts is wrong except for lamp drivers etc. Vbe not ~0.6 volts is wrong too except if a lamp driver is off. With light bulb in AC line you can also check capacitors to make sure they have a dc voltage across them. 0 volts on a capacitor, why did they install it in the first place? You cannot catch open capacitors with a limited power test. I've had capacitors open up in the VI limiter, also 50 v rated ceramics protecting the op amp power supplies in the front end.
Happy hunting.
Last edited:
Hey, I had that same amp where everything from the outputs back to the input stage got taken out! Direct coupled designs are great until they aren't. Plus, when the signal looks bad anywhere, it probably looks bad everywhere. That's why I do so much troubleshooting by passive measurements before I power up.
Yeah. That's fair. The only place I use wire wound resistors is in my resistive load bank. They see 20 kHz at the most ... or maybe a little bit of 80 kHz if you count the harmonics too.
LOL <snort> Otherwise known as carbon composite. Yep. Charcoal belongs in a BBQ.
Tom