Help Me Solve This Power Amp Riddle

[gbono]

When setting the bias do you input a 1000Hz signal and look at the harmonics on the analyzer or do you do several spot frequencies?

BTW I also purchased a IR thermometer from HF 1-2 years ago. Now when you use the laser "pointer" to locate/target the meter the temperature keeps rising on the display. It still works okay when the laser isn't used. Works better than my finger tip

[Phostenix]

I got a neat toy at Harbor Freight. It's an infrared no-contact thermometer.

I was thinking about getting something like that & wondering how accurate it would be pointed at the cases of transistors.

I ended up with the trim pot in essentially the same place that it was originally. That gives me 1-2mV on the emitter resistors at idle (3-6mA). With the emitter resistors siliconed down, I can't get readings from the top side, so I have to have the PCB out of the chassis, which means no contact with the main heatsink. That limits how hard I could push it, but I did put a signal into the Line In that got me around 20v p-p into the dummy load. That heated it up in 5-10 minutes to a point where the partial heatsink was too hot to comfortably touch. Removing the signal & allowing it too cool down, the bias levels returned to the 1-2mV range. I did this several times and it always returns to the same level. I guess the only other test I could do is with that thermometer.

I got a chance to play through it tonight for about an hour. I really didn't know what to expect since every reviewer or ebay seller I've seen just says that it's "LOUD". I expected it to be a one trick heavy metal pony, but I'd say my initial impression is surprise at how versatile this amp is. The range of tone shaping is VERY broad and the clean channel is great. It's very punchy (thanks to its huge power capabilities). The only thing it doesn't do very well is light overdrive. There's really no setting to get just a little dirt. I know from my scope work that the clean channel will begin to clip when the volume knob gets past about 3 o'clock, but my ears & my 60 watt speaker weren't up for the challenge.... I was also surprised at how usable it is with a single 12" speaker in a small room environment. I was able to play comfortably with both good clean sounds & full gain distortions. The solo channel has great sustain even at low levels. I'm impressed so far.

[gbono]

I'm wondering if it makes sense to monitor the idle current on the primary side of the PT to check and see if the bias is set properly and you are safely away from thermal runaway.

What about using a "kill-a watt" meter? ($28) Though I'm not sure it would be accurate enough for the job - they would be much cheaper than an HP35670 SA

[Phostenix]

I think what I'll do is get another set of these 4 ohm 100W Ohmite resistors that I have so that I can put a heavier load on the fully assembled amp and use a thermometer to make sure the temp stabilizes & trim back some silicon on the emitter resistors so that I can monitor them, too.

If it can be stable with a sine wave to 1/3 power continuously, it will handle guitar playing just fine.

[Euthymia]

I'm wondering if it makes sense to monitor the idle current on the primary side of the PT....

Good idea!

[Phostenix]

I borrowed a Fluke 61 infrared thermometer from work to do some stress tests tonight. The amp is reassembled with the heatsink attached & the PCB screwed down. I trimmed the silicone away from the emitter resistor leads so that I could attach meter leads.

Here's the initial setup:
84 degree room with no air movement
2 4 ohm 100W Ohmite resistors in series as the load
DMM across 1 of the 3 emitter resistors of the NPN power transistors
Anaolg meter across 1 of the 3 emitter resistors of the PNP power transistors
Scope on the load

20v p-p into an 8 ohm dummy load (6 watts) for 15 minutes - temps ranged from 110-120 degress on the output xstrs.
Followed immediately by 10 minutes of 30v p-p (13 Watts) - temps rose to 120-135 degrees. The outside fins of the heatsink rose to 117 degrees. At that point, I felt that the heatsink just wasn't able to keep up. I let it cool down.

Then, I brought in a box fan, a small fan for the load resistors, and turned on the ceiling fan. I ran it for 10 minutes with 60v p-p (55 Watts) and the temps rose to 120-128 degrees. The outside fins of the heatsink were at 100 degress.

Through all the tests, the voltage across the emitter resistors was very controlled, slow increase. On the final 55W test, the voltage stayed constant for at least a minute between .1v increases. It was more stable than the lower wattage tests (fans work wonders). I feel confident that the bias is OK & the amp can handle "normal" guitar service.

Crate didn't put any heatsink compound between the aluminum L bracket that the output xstrs are mounted on & the external heatsink. When I do my final assembly, I will add that.

[Phostenix]

BTW, this amp will deliver very close to 100v p-p into an 8 ohm load, which by my calulations is about 150 watts.

[Phostenix]

I replaced Q16 tonight & was pleased to have it dramatically reduce the power off thump. It's more of a dying pop now. It was brutal without the muting FET. So, that's one way of knowing if Q16 has failed....

The turn on thump is still pretty obnoxious.

[Phostenix]

I also did some more power tests with an analog meter across the load in addition to the scope.
145 watts before clipping into an 8 ohm resistive load.
210 watts before clipping into a 4 ohm resistive load.
Those resistors were singing!

[Phostenix]

I was able to borrow a Potomac Industries AG-51 signal generator & AA-51 audio analyzer from work and set the bias for lowest output distortion. The idle current ends up around 16mV across the .33 ohm emitter resistors.

[R.G.]

I think you've done a good job on this one.

Kudos.

[Phostenix]

Thanks for all your help!

The bias adjust worked just like you siad it would. As I raised the bias, THD went down to a certain point and then started to raise again.