Kevin's PL400 Upgrade

Kevin's amp is "burning in" and it sounds awesome. The increase in Q6 does make a difference and I like it.

Voltage at the start of clipping was 49.6 VAC RMS into the 8 Ohm dummy loads, which calculates to 307.52 Watts per channel. It is a PL600! The clipping was a soft flattening at the top and bottom of the waveform peaks at 2kHz.

One thing that was interesting was that the DC Protect board would cycle with the dummy loads connected and the amp at full power (sine wave). These dummy loads are just large 300 Watt WW resistors, so it might be a bit of inductive loading or ringing back to the outputs. Joe, what do you think?

In this picture Kevin's amp is driving the JBL's close to the limit of human endurance (it was loud). And you can see it was working a bit (meter average position). The sound was excellent and so crystal clear. I was holding on to my camera so tightly at this point that I surprised myself.
 

Attachments

  • Kevin\'s Amplifier Test.jpg
    Kevin\'s Amplifier Test.jpg
    1.6 MB · Views: 119
307 ?????? Good lord !!!!! Have the others you did had that much ?
 
Hi Mark
I would have to look at the relay circuit to give you a theory on why that was happening.
 
Kevin,

Is 614 watts going to be enough? Just kidding....I know you... and know you will LOVE this amp. Ill bet you end up with more than one PHASE LINEAR. With your Haflers working...you'll have at least two Plant the time everything is said and done..
 
Joe, I was reading that the induction of a large WW can be helped by a cap in parallel but I don't have a formula or even a starting point.
 
mlucitt said:
Kevin's amp is "burning in" and it sounds awesome. The increase in Q6 does make a difference and I like it.

Voltage at the start of clipping was 49.6 VAC RMS into the 8 Ohm dummy loads, which calculates to 307.52 Watts per channel. It is a PL600! The clipping was a soft flattening at the top and bottom of the waveform peaks at 2kHz.

One thing that was interesting was that the DC Protect board would cycle with the dummy loads connected and the amp at full power (sine wave). These dummy loads are just large 300 Watt WW resistors, so it might be a bit of inductive loading or ringing back to the outputs. Joe, what do you think?

In this picture Kevin's amp is driving the JBL's close to the limit of human endurance (it was loud). And you can see it was working a bit (meter average position). The sound was excellent and so crystal clear. I was holding on to my camera so tightly at this point that I surprised myself.
Click to expand...

I also had cycling of the dc protect boards while driving the PLWO that Lee put together for me. No idea the voltage or wattage, but am guessing around the clipping level. Turn up, no output, turn down, output returns. Must be normal behavior of the protect board circuit, although I do not like it.
 
What speakers were you driving with that Phase Linear amp and what is the impedance? When I connected the amp I just finished to the JBL speakers the DC Protect board did not cycle, those speakers have a DC resistance of 5 Ohms.
 
I am driving a pair of 8 ohm subs.

Not sure what you are saying there. Is it - on 8 ohm dummy loads you saw cycling but with the 5 ohm speakers you did not?
 
I believe so Scott. Seems i had that same behavior on one other one, but could not get it to trip on my system.
 
So I am thinking when you and mlucitt tested with speakers you did not have the amp to clipping as you did with the dummy loads. Typically you would not drive your speakers to clipping. When running full range the sound of clipping and fear of tweeter damage tends to keep you safer. Did either of you monitor the output during connection to speakers with scope to see if you were clipping as you did with the dummy loads?

Or is it clipping at all low frequency below 120 that incites it easier? I hope Gepetto can look at the schematic so we can understand it, address it. Not sure what my next step is other than keeping it turned down.
 
And that would defeat the whole purpose of having a PLWO 1000. I have not scoped the output with speakers connected. Not a bad idea though. I wouild think if it is clip related you wouldn't want to mess with the threshold, but if not, there should be a way to move that point. Are your subs inefficient enough to push Cathy that hard?
 
These boards are very sensitive to DC, it may be that there is a detectable level of DC (or just low frequency) at the output at high levels and this is causing the relay to open. This does not explain why it would cycle with my dummy loads and not the regular speakers. That is probably a different issue.
 
Can one of you scope the input to your speakers to find out what is being sent to speakers? I'm betting it is complete different than what your sine wave generator is creating in amp output level to see the cycling on the bench.

I verified I do not have a DC problem, and they do not cut out until really pushed - needles are banging to the end of travel at times. If it was DC or low frequency as root cause, I wouldn't have to crank it to make it happen.

The thing is, there are alot of DC protect circuits out there in alot of amps and I've clipped a few in my day but they have never cut out on me like this. Pro audio amps have clip indicators that I have flashed many a time/seen many in live sound venues and they keep outputting. Would be bad for business if they didn't. Have you guys ever bench tested an amp for output into dummy loads with a protect circuit in place and have it cut out at clipping as you have seen with the 400/700 protect boards?

I'm not a fan of clipping and do not plan to make a habit of it but I am less of a fan of losing all sound.
 
Yep, the Spec 2 and it GETS RADICAL when it does. Now this was the current limiter so it may nnot be the same. Trying to replicate a potential amp damaging event is not high on my list of things to do , but it may come down to that. Granted , current limit and DC voltage are separate parameters for the protection circuit, and dealing with them effectively may be tougher than we think. Then again you may be seeing the results of a square wave wounded DC protect board. We will know more late tonight or tomorrow mornin g. When ya leaving the country??
 
I was hoping Don would put in a showing here. He could answer a lot of questions about now.
 
Lee,

Yeah, current limiting is a whole different animal, and the Spec is one of the few amps with this type of protection circuit.

Yes, would love to hear from the designer, Don, where are you tonight?

I am out of the country at the moment, in Monterrey, MX until the 26th.

I really hope I have a damaged board and that is cause, as I really would like to keep my dc protect. I just celebrated recently by removing inline fuses, thought I was done with those babies for good.

I really don't want to go back to fusing speakers
 
Output relay circuits are a compromise. You want to detect a failed amplifier (DC) and ignore audio signals. When you detect DC you need to disconnect as soon as possible. The faster you make the detection, the less time you have with DC on the speakers. Once you detect then it takes about 15ms for the relay to drop out due soley to it's mechanical characteristics. A diode on the coil slows it down more. That's why I use an RC network.

I designed the circuit from the formulas in the data sheet for the UPC1237. The 400 board has a low threshold of around a couple volts DC. It has about a 30ms delay in tripping. I spent months testing it and recording waveforms, one of which you can see on Flickr. I made it as fast as possible. Yesterday I spent the whole afternoon testing boards. At 5hz or above with sinewave testing the board doesn't trip on a stock 400. it does trip at about 3 Hz. My test set uses a channel out of a BGW750 which puts out about 40 volts. I drive the amp well into clipping with an 8 ohm load and verify it trips. I then verify that it will not trip at 5hz.

For the heck of it I put a stock 400 on the bench and ran the same test. then I ran the frequency up to 20khz. The relay will trip at 20khz with the amp well into clipping. It doesn't trip at 10khz.

So if users are having troubles the question is what is the difference between the test set and actual use. Does the signal contain some something that is causing subsonic excursions. Generally if the woofers can be seen moving an appreciable amount that is subsonic. Turntables can be the culprit.

As I reread the posts it appears that both 400 and 700 board are tripping. Due to higher voltages, the 700 circuit is right out of the Phase Linear D500. They used it in the 200 and 300 also. How could I go wrong using the circuit they used?

After all the testing was done on the 400 protoype. I hooked a 6" woofer up to the output of the amplifier. I connected a knife switch
from B+ to the output and closed in the switch. Ten times. Each time the relay tripped and the speaker made a little yelp. Then I played music through the speaker to make sure it wasn't damaged. Sounded fine.

As for me being around on Saturday night, nope, not last night. Asleep by 10:30 on the couch. Ohio State was busy pouring the coal to Nebraska. Boring game.
 
You must log in or register to reply here.
Back
Top