This is a delightful understatement.
Recall that these parts will still require matching and the manufacturer has discontinued them.
I am trying to think of something that is not down right rude, insulting and patronising to say here... nah, I give in
A variac is a real tool and infinitely more useful than the light bulb hack. It is a bit like asking a surgeon's opinion about the efficacy of using a bread knife to do abdominal surgery - "because I already have one in the kitchen drawer, Doc".
Check out Anatech's signature line.
Oops, appears I forgot the link in my previous post:
http://www.diyaudio.com/forums/solid-state/74321-counterpoint-sa-100-sa-220-schematics.html
Funny Alan! BTW, ordering a variac today, and placed an order for the 1 ohm resistors Friday, so should be ready to proceed late this week.
http://www.diyaudio.com/forums/solid-state/74321-counterpoint-sa-100-sa-220-schematics.html
Funny Alan! BTW, ordering a variac today, and placed an order for the 1 ohm resistors Friday, so should be ready to proceed late this week.
I mentioned early on, but did not reiterate the point, the need for a dummy load. This is required to give the output stage a decent ground reference while you test it and check/adjust bias & offset voltage.
A couple of 8R2 10W wire wounds would do fine - see 8.2W-10-ND at Digikey as an example.
I think I have some 8R0 20W resistors from RatShack around, I'll check them. I think I even put spades on them
8-Ohm Non-Inductive Resistor - RadioShack.com
The variac will arrive today, and the 1 ohm (and a couple of 0.1 ohm) resistors showed up yesterday.
As for the resistor soldered to a fuse, I assume the fuse should be open (blown)? I don't normally use this form factor of fuse, so I'll need to pick some up today. Probably get some low-amp ones to make them easier to blow, or should I get 1A fuses as mentioned in the biasing instructions (link below)?
Once I have checked the variac out, what are my next steps? Should I follow the biasing procedure posted on altavistaaudio? Biasing Counterpoint SA Amplifiers
Also, the rails currently have 10A fuses in (I didn't look to see if they are fast or slow blow). The article recommends 6A fast blow. Should I replace with these?
As for the resistor soldered to a fuse, I assume the fuse should be open (blown)? I don't normally use this form factor of fuse, so I'll need to pick some up today. Probably get some low-amp ones to make them easier to blow
, or should I get 1A fuses as mentioned in the biasing instructions (link below)?Once I have checked the variac out, what are my next steps? Should I follow the biasing procedure posted on altavistaaudio? Biasing Counterpoint SA Amplifiers
Also, the rails currently have 10A fuses in (I didn't look to see if they are fast or slow blow). The article recommends 6A fast blow. Should I replace with these?
Correct. An easy way to 'blow' the fuse is to attack it with a soldering iron which will quickly detach/melt the copper wire fuse element.
Yes. But remember the steps I laid out in my post of 18 Oct (can't see the post numbers in the edit mode). I place the '1R0 fuse' in both rail fuse holders. That way you can check that both rails have the same quiescent current. They should be the same, but I prefer to not make assumptions.
Yes. I use 5A fast blow but that reflects what I can get my hands on as much as anything. Get a handful (fast blow) and use two to make that resistor/fuse thingy.
I would also replace the speaker fuses - again the factory fuse is a higher rating than I'm comfortable with. I typically use 2A fast blow - I can live with the occasional fuse replacement and this way there is a chance that the fuse will do something useful. But this does mean lifting the cover on the amp so you may choice something higher and closer to the factory setting. Your taste in music, speaker efficiency, deafness status etc will have an impact on the 'right' value for you. You could even get some audiophile approved fuses... I can hear Chris spinning from here
anatech said:
I believe you were replying to Chris' post, so I included both of them. Without removing the output board, I can't really clean it up or do much of a post-mortem. The worst of the damage appears to be around VR1/2, and I haven't seen a chip split in half like that in quite some time. I would guess that you are correct Alan, that the +/-15V regs went, or one of them, thereby presenting +/-70V to the APC. None of the fuses blew, so I would guess that there was not much if any large DC offset when this occurred.VivaVee said:
Getting back to the steps, just a couple of questions:
1) for discharging the caps, I have a discharge probe (red plastic probe thingy with an internal resistor, measures 1K, but certainly not 10W). It has a lead (pretty heavy) with an alligator clip on that end. I got it from some tube place, don't recall where. Is this OK to use for discharging the main caps?
2) you said the bias pot is VR101; you recommend turning this 2 turns clockwise to decrease the bias before I apply any voltage? Also, I assume there is a DC offset pot as well?
I got some 250mA fuses and abused 2 of them to make the 1ohm fuse replacements. Variac arrived and works fine, though the meter is off some (as well as the dial). I also got some 6A fast blow for use when (if?) this is fixed. And I will use a load resistor as discussed. I'll have to wait for the HiFi tuning fuses
And yes, I do detect some spinning to the north as well...This sounds fine. Just remember to unclip the 'discharge probe' before you wind the voltage back up. That little faux pas is why I recommend the 10W rating - it gives you a chance to 'smell the roses' and hit the kill switch. Naturally I have no personal experience of such a silly mistake...
Correct. It is a multi-turn trim-pot.
Yes - it is a single-turn trim-pot. But since you have not, at this stage at least, changed the set of output MOSFETs, the DC offset adjustment is likely to be very close to what ti should be. So leave it be.
Another potentially stupid question: Chris mentioned monitoring the supply current. Meaning AC line current in? Or the rail current (voltage across the rail fuse replacement 1 ohm resistors)? Also curious as to why I couldn't just use an ammeter in place of the fuse? Protection of the meter (let the resistor burn up instead in case of high rail currents?)
Variac only has a voltage meter, no ammeter.
Skipping ahead a bit, supposing I do get this working, if I were wanting to run RMAA on the amp to see how the repaired channel does compared to the good one, what sort of protection would I need in place for the sound card (M-Audio box actually) inputs?
Thanks,
Chris (other one)
Variac only has a voltage meter, no ammeter.
Skipping ahead a bit, supposing I do get this working, if I were wanting to run RMAA on the amp to see how the repaired channel does compared to the good one, what sort of protection would I need in place for the sound card (M-Audio box actually) inputs?
Thanks,
Chris (other one)
To quote various teachers - there is no such thing as a stupid question. However, I'm sure that I have tested this assertion on occasions
Rail current is #1 on the list. But a good weekend project is to add an ammeter (a big tough moving iron analogue meter) to your shiny new variac. Seeing that needle climb as you wind up the variac is the first indication that trouble is a brewing.
The resistor will take more abuse and is much less expensive than the fuse in your DMM. Also the resistor/fuse will fit snugly in the fuse holder while your DMM leads may unclip from the fuse holder, which is not pretty.
An attenuator plus a voltage clamp. This can be as simple as a resistive divider with a TVS across the lower arm or you can build a buffered, protected switched frequency-compensated attenuator that emulates the first section of an oscilloscope. See attachment or Pete Millet's contribution elsewhere on this site.
Attachments
Yes, just be careful where you put your tea (auto translator suggests this would be "coffee" in your town)
I assume you mean this thread:
http://www.diyaudio.com/forums/tubes-valves/155405-test-measurement-interface-soundcard.html
Link to Mr. Millet's site:
Soundcard Interface
Looks like he is just using a switched resistor divider and small signal clamping diodes? Is the attachment you provided something from here (diya)? I wouldn't want to blow my M-Audio FW Audiophile box up
http://www.diyaudio.com/forums/tubes-valves/155405-test-measurement-interface-soundcard.html
Link to Mr. Millet's site:
Soundcard Interface
Looks like he is just using a switched resistor divider and small signal clamping diodes? Is the attachment you provided something from here (diya)? I wouldn't want to blow my M-Audio FW Audiophile box up
OK, had time to start looking at this tonight. I went thru all the steps you both had suggested, removing the rectifier tube, installed 1 ohm resistors in the rail positions, decreased the bias by 2 turns CW, etc. I used 3 meters; one measuring DC offset, one measuring bias on the negative rail, and one measuring rail voltage. Brought the rails up to 25V; everything looked good. Very little (as in no) bias voltage/current. DC offset nearly immeasurable. Brought the rails slowly up to 50Vdc (watching the meters). Fine still. Finally, brought them up to 70V. Finally seeing some bias current (0.1V, so 100mA). I adjusted the bias current to .485V or 485mA. Let it sit for awhile. DC offset was fine (12mV or less). Bias current dropped after some time to .473V. I took a meter and checked the variac output and it was a bit high (130Vac), so decreased it to 122Vac. Rails dropped to ~67V, bias current to 443mA. I readjusted the bias to 485mA. I also started adjusting DC offset, which seems touchy (not surprising really), but got to to around 6mV. Bias current seems quite stable, other than it drops some the longer the amp is on. I had the amp on for 30 minutes or so.
At this point it would appear that it is working, so what are my next steps? Bias it per Mike's instructions? Should I reinstall the rectifier and the relay?
Also, I had removed the +/- 15V regulators. The positive 7815 was definitely blown, shorted output to ground pins, but the negative was probably OK, or at least didn't measure bad WRT shorts.
At this point it would appear that it is working, so what are my next steps? Bias it per Mike's instructions? Should I reinstall the rectifier and the relay?
Also, I had removed the +/- 15V regulators. The positive 7815 was definitely blown, shorted output to ground pins, but the negative was probably OK, or at least didn't measure bad WRT shorts.
You have a functioning output stage. I suggest that you set the bias to be the chosen value just after turn on. You have noticed that it deceases over time, but not too much. If you keep adjusting the bias upwards until it settles at, say 0.5A, after running for some time, then it will be much much higher when you turn it on next time. Personally, I would accept a lower bias setting, in the knowledge that the amp runs cooler, with lower quiescent current and might just not go into thermal run away quote so easily...
The dc offset does drift about - anything under 20mV is fine. This normally takes a while - so set the bias, then tweak the offset every 30 minutes over a 4 hour period. Maybe get a good book, watch a movie, or whatever to pass the time. Provided that the rectifier valve is still out and you have that dummy load attached, nothing should happen, while you tweak away.
Once that is done, power down, discharge the big output stage power supply filter caps (BOTH SIDES), and fit yr 1R0 fuse thingies in the fuse holders of the good side and check the bias/offset on that side.
Once you are happy that both output stages are set up, power own and discharge those big caps again (do I sound like a broken record, yet?).
Re-install the rectifier valve and mute relays.
Then connect yr signal generator to both inputs and apply some small signal - say 0.1V 1kHz and check the outputs with your scope after the mute relays have clicked on (the front panel LED goes from red to green, but you have probably already noticed that a few times by now). You see a clean output at increased magnitude. If you get keen you can check the gain fro each channel. But if Left in = Right in and Left out = Right out, then all is good with the world.
Depending on the power rating of your dummy load you can try a bigger signal.
Optional: I like to connect my lab reference speaker at this point - this is a truly staggeringly awesome full range driver that came out of an old Philips tv.
I am sure that it has been anointed by the audio gods it is so good
This is the speaker that gets the amp turn on test - hook it up and cycle the power to the amp. I like to check that there are no thumps, clicks, squeaks or squeals - these all sound so much better through my ref speakers. Probably because if I used 'real speakers' I would be cringing in the corner dreading the moment of impact, sorry turn on. I think you will be getting a feel for my strategies - find a sacrificial something to take the heat rather than your good stuff until your confidence level gets high enough to hook the repaired amp back into a decent system.
Thanks Alan. I assume I can run this without the variac now? I think I will remove one of the 1R0 things and move it to the good side. Do both sides at once for 4 hours. Will switch the fuses to 6A from the 10A that were in there. I could also pick up some 1A which is what Mike suggests using for biasing and use 1 fuse and 1 resistor in each channel (resistor in the negative rail position towards the back of the amp is what he tells you to do).
I don't have a signal generator, but can use my M-Audio box to feed a 1KHz sine and adjust the levels on the computer. I may build some protection circuitry as we discussed and run RMAA on the amp as well when it is biased.
Thanks again for all of your invaluable help! Chris as well!
I don't have a signal generator, but can use my M-Audio box to feed a 1KHz sine and adjust the levels on the computer. I may build some protection circuitry as we discussed and run RMAA on the amp as well when it is biased.
Thanks again for all of your invaluable help! Chris as well!
Currently biasing the amp now, has been on for about an hour, looking good (both channels). The good channel was at 425mA, so I set the repaired channel to match. I increased the bias to about 450mA on both channels (recommended was 450-500mA)... or should I just leave it a bit low at 425mA?
Also, my brother asked if it would be a good idea to disable the APC on the other channel?
As for running RMAA and soundcard protection, I may just build something like this:
http://www.diyaudio.com/forums/tubes-valves/167207-soundcard-protection-z.html
I have some Dale RN60s in 10K and 499 or 511 ohms. I would leave the 8 ohm load resistors in place (I think), and closely watch input levels. Would this be adequate? If I added some diodes, would I use either one or a pair across the ~500 ohm resistor, and put a set in each way (or pairs to give ~1.4V limits)?
Chris
Also, my brother asked if it would be a good idea to disable the APC on the other channel?
As for running RMAA and soundcard protection, I may just build something like this:
http://www.diyaudio.com/forums/tubes-valves/167207-soundcard-protection-z.html
I have some Dale RN60s in 10K and 499 or 511 ohms. I would leave the 8 ohm load resistors in place (I think), and closely watch input levels. Would this be adequate? If I added some diodes, would I use either one or a pair across the ~500 ohm resistor, and put a set in each way (or pairs to give ~1.4V limits)?
Chris