advertisement


Magic Smoke

orangeart

KJF Audio Ltd.
I've managed to let it out again.

I successfully swapped my NAPA 6/3 boards for NCC 200 boards recently, all went well.

Last night I decide to change feedback caps from the standard BC128 to some FFBs, Again all went well. I decided I should re-check the bias. The method I normally use is Malcolms method of setting 7.2mV across both emitter resistors. I unfortunately managed to slip with one of the probes, not for the first time :( and had very brief continuity between the resistor shown in the pic and the heatsink on the collector of TR8. Note the pics are not of my board, i borrowed them from Les' site. Anyway the resistors with circles round them spectacularly set of fire. The cap next to one of them has melted as well, although I think this is collateral damage from the inferno next to it. There was smoke coming from under the board as well, i've not had a chance to look under there yet, but the resistors were so hot that it could have been the solder melting.

ncc200-top.jpg


power-amplifier-circuit-ncc200-z.jpg


Anyway, apart from the obvious resistors and cap where should I start? I've got most spares available although maybe not exact models, I have the NAPA 6/3 board I could steal some parts from to get it going then replace what I need to.

Any help appreciated.

Thanks

Stefan
 
I'd just like to point out that it's Naim's method of setting bias, not mine.

Les will tell you that he only recommends the method detailed on the diagram above.

My advice FWIW is:

  • remove both output transistors and check them with a meter
  • replace TR7 (MJE15030) & TR8 (MJE15031) - n.b. the transistors in these positions on a NAPA board are MJE243/MJE253 and their pins are in reverse order
  • replace all obviously burnt resistors and capacitors
  • replace the two 0R22 resistors in series with the output transistors
  • power up the board with your meter on the 2V d.c. range connected across the two 'bias test points'
  • check that you can set this to 1.6 - 1.7 volts (not millivolts) using the trimmer and that the trimmer allows you to adjust this
  • only when you are successful with the last step should you replace the output transistors
  • When everything is working again, check the dc offset carefully as you may also have damaged the feedback capacitor and/or the capacitor across TR5

We've all done the slipping probe thing so don't feel too bad about it,

malcolm
 
Stefan
Real sorry to hear about that I know just how you feel, I have released the majic blue smoke several times in the past. :(
As a good tip for measuring something like that go to Maplin and get a pair of these mini spring hooks they plug in the end of your test lead
http://www.maplin.co.uk/probe-clip-test-leads-46225
they are realy good for protection against shorting whilst testing.
As for the board I would remove TR6 TR8 TR10 and all the components in the circuit connected to them test them and only put back what you are sure are OK
Little bit worrying that the 100 ohm resistor on the other side of the board has gone as well as that could mean that side of the board is damaged also.

If the tracks are damaged under the board personnaly I would go for a new board and rebuild.

Best of luck

Alan
 
Oops, I didn't notice that the opposite 100R had gone as well, so change TR7 as well with the same caveat that the MJE243 from the NAPA board has its pins in the reverse order to the MJE1030.

malcolm

p.s. as both 100R resistors have gone it may be worth replacing the two 0R22 resistors that are in series with the output transistors
 
I can sympathize and although its around 30 years ago I dropped my screwdriver 'probe' onto my Lindsley-Hood 75w and got a bang and smoke. Took out the drivers and the ouput trannies. Being a perfectionist (then) I bought new trannies and resistors each from the same corresponding batch for both channels...... Can't have any mismatching!

I later sold the JLH 75w to a mate when I 'upgraded' to a Quad 404 current dumping amp. He was still using it when I last made contact a couple of years ago.

Cheers,

DV
 
Awful feeling isn't it?!

Probe clips are the way to go or if you have to use probes you can reduce the chance of contact when slipping by wrapping insulating tape on the probes to the very tip such that a sideways touch on a component doesn't conduct.

Good luck with the fix.
 
I did the same thing with my 140 clones and took out one of the output trannies. The cure was to strip some sheathing off some old cable and slide it over the pins on my test leads leaving just 2mm showing. Those hook type leads are very good though. I have some on my AVO but I just can't be bothered to drag the bulky thing off the shelf sometimes :(
 
Thanks for the advice guys, i'll get some time tommorow eve hopefully.what is the best way to test the trannies once I get them out?

Stefan
 
I can't imagine why anyone would want to risk damaging the amp in this way.

What is so all-fired difficult about installing the probes into the positive lead.?

I give up - I really do......I sometimes feel as if I'm dealing with a bunch of numbskulls
 
Seems a bit harsh there Les, we all have to start somewhere and through mistakes we learn. It does sound like your method is much safer, so why choose a risky option?

Can't answer that, but I for one am not as methodical and measured as I would like to be at all times (and it is especially important with electronics it seems!) and sadly but rarely, mistakes do happen.

Still orangeart, you have been slapped firmly on the wrist! :D
 
I'd just change the two 100R resistors (and the heat damaged cap) and see if it's all OK. You can often do more damage removing components just to test them, and in this case I reckon TR8 & 10 will definitely be OK and TR7 & 9 probably will be too.
 
Hi les, I have tried to use your method but have mever found a secure way of attaching the probes to the lead and the terminal, I'd have thought slipping here could cause as much damage, how do you connect securely? Would a resulting loss of power on the positive rail while the negative rail was powered result in damage? I guess as I'm no expert I follow what seems to be the method most folk on here use to good effect.
 
Hi les, I have tried to use your method but have mever found a secure way of attaching the probes to the lead and the terminal, I'd have thought slipping here could cause as much damage, how do you connect securely? Would a resulting loss of power on the positive rail while the negative rail was powered result in damage? I guess as I'm no expert I follow what seems to be the method most folk on here use to good effect.

The method of connecting a pair of leads to the male and female terminals isn't rocket science old chum. Simply make up a pair of gash leads with 4mm plugs on one end and croc clips on the other.
Now, is that too difficult I wonder.? Perhaps you enjoy blowing up amps.?

The NCCs will tolerate a loss of + rails without too many problems so long as it's not for too long a period.

However, if you prefer to use the method that you report and slip with the probes, it does in some small way reflect on me and the product I make.

Please forgive me if I appear a little caustic in my comments but having pointed out a fairly foolproof way to deal with setting up, if you lot persist in behaving like numpties, don't expect any sympathy from me.
 
I'm sure many of us are numbskulls relative to your level of knowledge and experience Les. I certainly am! I'm sure Stefan will get past this and learn much from the experience. I have to admit I have done far worse :rolleyes: though never had a fry up (yet but I will I promise).

The 38mA lead way is the method I use - I'm well clumsy so attaching things before powering on is more conducive to me not screwing up :). Plus you need to stay there for 20 minutes anyway, so probes isn't really the way to go.

Worth repeating the test after a good warm up too IME or set a little under 38mA just to be sure of some headroom when its all closed up.

Stefan, crocodile clips is what I use, after all we are talking low current while not fully connect up to audio and speakers. To be extra safe you can clip then tape them over and even tape them down to something as a good technician would do with a test setup as you know its easy to catch something with a sleeve etc.

Also, try not to be close by (ie over looking) the board while doing these things... you could have been more unlucky, for example a reverse condition on an electrolytic wouldn't be fun.
 
For setting the bias to LesW method I made up a couple of leads with a crimped insulated male and female spade connector on the test ends so no more slips;)

Alan
 
Please forgive me if I appear a little caustic in my comments but having pointed out a fairly foolproof way to deal with setting up, if you lot persist in behaving like numpties, don't expect any sympathy from me.

It does seem a little caustic to be honest Les, like many I am just trying to get the best amp I can with a little outlay and within my ability. Mistakes do happen and I hadn't had the foolproof method that you outline above pointed out to me. I'm sorry if you feel my mistake has cost your reputation, I'm sure it hasn't, we all know what a great circuit it is, that's why I'm using it and why the hacker nap and various China clones are based on it. It sounds much better than the napa boards I replaced.

As I've already made the mistake and won't again - as the developer of the circuit would you care to add your 2c to trouble shooting now I find my self in this position? I appreciate your support and continued interest in our community. Thanks for looking at my thread.

Stefan
 
Whats this obsession with setting Bias current , It was reported on PFM (by Robert I think) , that any bias setting between 40 and 30 something mA was reported to sound the same.
I tried setting mine to 38mA and was met with a constant wobble/slightly variation ie you are never going to get them the same , so 15 mins warm up ,set to roughly 38mA , job done.
 
Ok, all fixed, sort of......, changed the obviously burnt bits, and took out and tested the drivers and out put transistors, both the out put transistors were blown, replaced them and all is back up and running, there is one curious problem though. I can't adjust the bias with the pot, testing les' way or the naim way (with the new mini retractable probe clips :) ) the bias barely changes over whole resistor range, I normally manage to set 7.2 mV across the emmiter resistor pair but I can't get it above 2.7. I had a quick probe around and compared to the other board which seems mostly the same.

Sounds good though.

If anyone could help with what to measure, I'd appreciate it, I'm scratching my head.

Stefan
 


advertisement


Back
Top