If those are pass transistors for the phono supply, then the emitter would be the phono section power I think. And if the target voltage at the emitter is 18 which some of your devices are, and if those are the cool ones, and the 17.1V's are the hot ones, then it kind of makes sense. The bad phono section is pulling down the supply a full volt, which is quite a bit. Without a schematic of the phono section itself, could be quite difficult to isolate. Maybe zoom your IR camera on the component side to the phono section and see if any of the components are hotter from one channel to the other. I expect the resolution/sensitivity isn't that good though on the camera.
@mikeAtx
The drawing below shows the 4 transistor pairs at the front of the phono section, and the voltages observed at each of the leads are noted.
The two "hot" pairs are 1 and 2, whereas 3 and 4 run cool.
However, the lower emitter voltages are not found on pairs 1 and 2 (which all run hot) but on the four transistors on the A side (only 2 of which run hot).
Later I'll post more detailed pics of the phono section.
The drawing below shows the 4 transistor pairs at the front of the phono section, and the voltages observed at each of the leads are noted.
The two "hot" pairs are 1 and 2, whereas 3 and 4 run cool.
However, the lower emitter voltages are not found on pairs 1 and 2 (which all run hot) but on the four transistors on the A side (only 2 of which run hot).
Later I'll post more detailed pics of the phono section.
Here are detailed pics of the phono section, top and bottom (bottom is flipped).
I tested all caps for ESR and they're all fine.
There are four trim pots in the phono section. Two are close to and to the right of the heatsinks, and two are to the left of the DIP switches. I tried to adjust the trim pots next to the heatsinks but that did not cause any temperature change at the transistors, nor any change in the voltage drop across the two 5.6 ohm power resistors (375 mV if I recall).
The 2-position toggle switch (between the DIP switches) selects MC or MM cartridges. When MC is selected, impedance can be adjusted separately for each channel via the two DIP switch banks. When MM is selected, all DIP switches must be off.
![phono section [top layer].jpg](https://www.diyaudio.com/community/attachments/phono-section-top-layer-jpg.1444982/ "phono section [top layer].jpg")
![phono section [bottom layer].jpg](https://www.diyaudio.com/community/attachments/phono-section-bottom-layer-jpg.1444983/ "phono section [bottom layer].jpg")
I tested all caps for ESR and they're all fine.
There are four trim pots in the phono section. Two are close to and to the right of the heatsinks, and two are to the left of the DIP switches. I tried to adjust the trim pots next to the heatsinks but that did not cause any temperature change at the transistors, nor any change in the voltage drop across the two 5.6 ohm power resistors (375 mV if I recall).
The 2-position toggle switch (between the DIP switches) selects MC or MM cartridges. When MC is selected, impedance can be adjusted separately for each channel via the two DIP switch banks. When MM is selected, all DIP switches must be off.
Pairs 1 and 2 are the ones running hot, but each transistor pair are mounted back to back to a single heatsink, so that the base and emitter are reversed between A and B sides in the drawing. Hopefully that makes sense, sorry! 🙂
Yes, this PCB has seen heat. My goal right now is to get the preamp to run 100% cool and I'm 70% there.
The only remaining overheating is with four voltage regulating transistors in the phono stage as described above. Tricky to diagnose so far.
I would like to button up the main preamp and go back to rebuilding the power supply. All feedback / advice / pointers are greatly appreciated... Thank you!
The only remaining overheating is with four voltage regulating transistors in the phono stage as described above. Tricky to diagnose so far.
I would like to button up the main preamp and go back to rebuilding the power supply. All feedback / advice / pointers are greatly appreciated... Thank you!
I don't think the regulator transistors but something in the circuit drawing too much. As a wag again, assume the heatsink is 11C/W and the difference between the two heatsinks is 29C (57-28). or 2.6W more on the hot ones. Assume 4V across the pass tran and that gives .65A. I think maybe true too, given the temps of everything else, almost all the juice from the bench supply may be going thru the hot phono channel. There should be some relatively hot things in the phono section besides the pass tran. .6A x 18V gives 10W of heat to be dissipated on the positive and negative halves. If you have a thermistor maybe probe around for a hot spot or even a finger might be able to tell. And of course you can compare voltages between good/bad. Without a schematic that could be difficult to get much info out of. Have you tried running a signal thru the phono section? A 1mv 1KHz signal you could hear if it is badly distorted. Maybe check for Vbe's of < .5V on the bad channel to see if something is cutoff or if 0, shorted. Same for Vce. Any 0's?
My thoughts entirely... but it sure looks like there aren't. I have a thermistor and I must have poked every single component in the phono stage. I also scanned the whole section with the thermal camera (however that camera is garbage and I'm returning it). Anyway, nothing elsewhere in the phono section runs hot. Everything is 22-23 C, a few up to 27-28C but more importantly comparable temps on both channels.
The only exception is the small section highlighted in the pic below. The right half is about 10 C hotter than the left. But I wouldn't exclude the possibility that the right side is merely being heated up by the hot transistors.
I did random-check a number of voltages and all looked good - consistent between L and R channels.
I haven't tried running a signal through the phono stage. I may need a better signal generator than the dodgy handheld I have.
On the bright side, I hooked the Krell to my main system's Classé monoblocks, my DAC to the Krell and darn that thing sounds good! I listened for several hours while taking measurements and trying to get some work done.
No matter what I played - industrial at 100+ dB, acoustic blues, nothing at all - the Krell chugged the exact same amount of current, to the milliamp.
Meanwhile, I worry about my "hot" transistors running 50-52 C but the air coming out my laptop's vent is 10 C hotter and I sure don't worry much about my laptop 🙂
The investigation continues. I can't wait to have this thing working right.
Last edited:
Total guess as I have not traced the board at all, but I think the highlighted section at the bottom is just some zener's that set the 18V reference for the pass trans. And agree with your guess, heat from the pass trans. It does look like the MC pre-preamp is the stuff near the RCA connectors to me based on the MM/MC switch. You could make a very cheap voltage divider of 100/1 with a 5K/50 ohm resistor pair. I think your cheap waveform gen can probably do 100mv output and then easily divided down. I think you mentioned you have a scope, so inject the signal and poke around with the probe to look for differences between channels. You'll very likely need to ac couple the probe to see the signal.